# Future Human Wiki — full corpus

> A reference work on longevity, gene editing, cybernetics, and the technologies changing what a human being is — with the evidence, the timelines, and the objections.

This file is the entire wiki: all 205 articles as markdown source, concatenated.
The article text below is 2,852,650 bytes — about 2.7 MB — plus this header.

It is large on purpose, and it is probably not what you want first:

- https://futurehumanwiki.com/llms.txt — the index: every article as a title, a summary and its classification.
- https://futurehumanwiki.com/corpus.json — every article's classification and the link graph, without any body text.
- https://futurehumanwiki.com/taxonomy.json — what the classification values mean.
- https://futurehumanwiki.com/raw/<slug> — one article's markdown source on its own.
- https://futurehumanwiki.com/data — documentation for all of the above.

## How to read this file

Articles are separated by a line of 78 equals signs, in the order they appear in the
A–Z index. Each is preceded by its slug, title, portal, and URLs, and then reproduced
verbatim: YAML frontmatter first, then the markdown body.

Nothing is rendered, which is the point. Footnote definitions keep their leading source-type
token (`paper`, `preprint`, `trial`, `regulator`, `report`, `book`, `news`, `statement`,
`law`), the optional {…} note on a citation says why that source is the one cited, fenced
```infobox / ```keyfacts / ```timeline / ```compare / ```figure blocks carry JSON, and
[[wikilinks]] name article slugs. A wikilink whose target has no article is a page the wiki
has asked for and not written; those are listed at https://futurehumanwiki.com/wanted.

## What the frontmatter fields mean

- type: concept, technology, intervention, organization, person, event, or risk
- status: established, emerging, experimental, speculative, contested, or historical
- horizon: when the subject plausibly matters at scale
- trl: technology readiness level 1-9, on technologies and interventions
- humanEvidence: what has been demonstrated in people, as opposed to in animals or cells
- access: who can obtain it today and on what terms
- reversibility: reversible, partly-reversible, context, difficult, or irreversible
- issues: known gaps in the article itself, written by its editor
- updated: an editorial claim about when this article's text last meaningfully changed. It is
  not the same as the save time, which the revision log records separately: every article has a
  full revision history at https://futurehumanwiki.com/wiki/<slug>/history.

Full definitions, with the reasoning behind each value: https://futurehumanwiki.com/taxonomy.json

## Editorial rules these articles are written under

Claims about the future are attributed to their proponents, never asserted by this site.
Results in animals are always distinguished from results in people. Articles state where
their own evidence runs out, in [!caution] and [!debate] callouts collected at
https://futurehumanwiki.com/open-questions. Nothing here is medical advice.

Text is CC BY-SA 4.0 — https://creativecommons.org/licenses/by-sa/4.0/
Cite as: Future Human Wiki (futurehumanwiki.com), full corpus, 205 articles, CC BY-SA 4.0.

==============================================================================
ARTICLE: aav-vectors
TITLE:   AAV vectors
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/aav-vectors
SOURCE:  https://futurehumanwiki.com/raw/aav-vectors
==============================================================================

---
title: "AAV vectors"
slug: "aav-vectors"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["genetics", "bodies"]
tags: ["gene therapy", "delivery", "viral vectors", "immunology", "manufacturing", "rare disease"]
summary: "Engineered adeno-associated viruses used to carry therapeutic genes into human cells, the workhorse of in vivo gene therapy and the source of most of its limits."
updated: "2026-07-27"
humanEvidence: "Licensed AAV products treat inherited retinal dystrophy, spinal muscular atrophy and haemophilia in people; the same clinical record shows capsid immunity blocking redosing and fatal liver injury at high systemic doses."
access: "Obtainable only as specific licensed gene therapies given in specialist centres, at one-time list prices in the millions of dollars; patients with pre-existing anti-capsid antibodies are screened out."
reversibility: "context"
issues: ["The 2025 Duchenne liver-failure deaths and the regulatory response are stated without a source.", "Manufacturing cost is described qualitatively; no sourced cost-of-goods figure is given."]
---

```infobox
{
  "caption": "Gene delivery vector",
  "rows": [
    { "label": "Basis", "value": "Adeno-associated virus" },
    { "label": "Genome", "value": "~4.7 kb single-stranded DNA" },
    { "label": "Cargo limit", "value": "About 4.7 kb" },
    { "label": "Persistence", "value": "Mostly episomal" },
    { "label": "First identified", "value": "1965" },
    { "label": "First approved product", "value": "Glybera, 2012" },
    { "label": "Redosable", "value": "No, in practice" },
    { "label": "Readiness", "value": "TRL 9 (multiple licensed products)" }
  ]
}
```

**AAV vectors** are gene-delivery vehicles built from adeno-associated virus, a small, non-enveloped parvovirus that infects humans without causing known disease. Nearly every licensed *in vivo* gene therapy uses one, including treatments for inherited retinal dystrophy, spinal muscular atrophy, and haemophilia. The vector's virtues and its constraints come from the same source: a genome barely large enough to hold one gene, a capsid the immune system recognises, and a payload that mostly sits outside the chromosomes rather than joining them.

## How it works

Wild-type AAV carries about 4.7 kilobases of single-stranded DNA between two inverted terminal repeats, encoding two gene cassettes — *rep*, for replication, and *cap*, for the capsid. It cannot complete its own life cycle without a helper virus such as adenovirus or herpes simplex, which is why it was first noticed in 1965 as a contaminant in adenovirus preparations rather than as a pathogen in its own right.

To make a vector, everything between the terminal repeats is deleted and replaced with the therapeutic cassette: a promoter, the gene, a polyadenylation signal. The *rep* and *cap* functions are supplied separately during manufacturing, so the resulting particle can enter a cell and deliver its cargo but cannot replicate. Once inside, the single-stranded genome is converted to double-stranded DNA and circularises into episomes that persist in the nucleus, transcribing for years in non-dividing tissue. A small fraction integrates at semi-random sites; wild-type AAV's preference for the AAVS1 locus depends on Rep protein, which vectors lack.

Cell entry requires glycan attachment factors that vary by serotype plus a shared protein receptor, AAVR, identified in 2016.[^pillay2016] The discovery explained why some cell types resist transduction regardless of capsid choice.

Because the vector reaches only body cells and does not enter the germ line at any appreciable rate, AAV therapy sits squarely on the somatic side of the boundary that defines [[germline-editing]]. Outside medicine, AAV is the standard delivery route for [[optogenetics]] constructs in neuroscience and, less legitimately, the vector class most often discussed in connection with [[gene-doping]].

## Serotypes and tropism

The natural AAV serotypes differ chiefly in their capsid surface, and therefore in which tissues they enter. AAV2 was the first characterised and remains common in the eye. AAV8 and AAV5 favour the liver. AAV9 crosses the blood-brain barrier in newborns well enough to underpin an approved systemic therapy for spinal muscular atrophy. AAV1, AAV6, and AAVrh74 are used for muscle.

Capsid engineering aims to do better than nature. Directed evolution, DNA shuffling, and peptide-display libraries have produced capsids with sharply improved central nervous system delivery in mice.[^deverman2016] Machine-learning models fitted to library-screening data now propose viable capsid variants directly, an application of [[ai-protein-design]] to the delivery problem rather than to the therapeutic protein. The best-known cautionary tale is AAV-PHP.B, which crosses the murine blood-brain barrier with startling efficiency and then turned out to depend on a receptor, Ly6a, present in some inbred mouse strains and absent in primates.[^hordeaux2018] Its failure to translate is the standard example of why capsid performance in one species predicts little about another — the same species-gap problem that recurs throughout [[targeted-drug-delivery]].

> [!caution] Mouse capsids are not human capsids
> Every claim about a new capsid's tropism should be read with the question: in which species, and does the receptor it exploits exist in people? Several widely cited enhanced capsids do not work in non-human primates at all.

## Development history

```timeline
[
  { "year": "1965", "title": "Discovery", "text": "AAV is identified as a small contaminating particle in adenovirus preparations and shown to require a helper virus." },
  { "year": "1984", "title": "First recombinant vector", "text": "Researchers demonstrate that the viral genes can be removed and replaced with a transgene between the inverted terminal repeats." },
  { "year": "2006", "title": "The immune barrier appears", "text": "In a haemophilia B trial, factor IX expression rises then falls as capsid-specific T cells destroy transduced hepatocytes, defining the field's central obstacle." },
  { "year": "2012–2017", "title": "First approvals", "text": "Glybera is licensed in Europe and later withdrawn; Luxturna is approved in the United States for RPE65 retinal dystrophy." },
  { "year": "2019", "title": "Systemic dosing at scale", "text": "Zolgensma is approved for spinal muscular atrophy, delivering AAV9 intravenously at doses orders of magnitude above earlier local injections." },
  { "year": "2020–2025", "title": "Dose-limiting toxicity", "text": "Deaths from acute liver failure and thrombotic microangiopathy in high-dose systemic trials, including for X-linked myotubular myopathy and Duchenne muscular dystrophy, force reassessment of dose ceilings." }
]
```

## Limitations

**Cargo.** The 4.7 kb ceiling excludes many important genes outright — dystrophin, most [[crispr-cas9]] constructs once guides and regulatory elements are added, and the still larger dCas9 fusions used for [[epigenome-editing]]. Workarounds include micro-genes such as the truncated dystrophins used in Duchenne programmes, dual-vector systems that reconstitute a protein from two halves via split inteins, and pairing AAV with smaller nucleases. Each adds inefficiency. Muscle-directed constructs expressing follistatin, discussed under [[myostatin-inhibition]], are small enough to fit and have been used in both clinical and unsanctioned settings.

**Pre-existing immunity.** Because AAV circulates naturally, a substantial fraction of adults — depending on serotype and region, roughly a third to more than half — carry neutralising antibodies that inactivate the vector before it reaches its target. Those patients are screened out of trials, which distorts both eligibility and evidence.

**Redosing.** A treated patient develops high antibody titres against the capsid, so a second dose of the same serotype is ineffective. Switching serotypes helps only partially because of cross-reactivity. Strategies under investigation include plasmapheresis, IgG-cleaving bacterial proteases, and B-cell-depleting or tolerising regimens; none is routine practice as of 2026.

**Durability.** Episomal genomes are not replicated with the chromosome, so dividing cells dilute them out. This matters most in children, whose livers and muscles grow substantially after treatment, and it underlies the declining expression seen in some haemophilia programmes. Editing the genome rather than adding an episome avoids the problem, which is one reason [[base-editing]] and [[prime-editing]] payloads are attractive despite the packaging squeeze.

**Manufacturing.** Vector is produced by transient transfection of HEK293 cells or in insect cells with baculovirus, then purified. Yields are modest, a large fraction of particles are empty capsids that contribute immunogenicity without therapeutic effect, and cost of goods for a systemic dose is a meaningful share of the price of the finished therapy discussed under [[somatic-gene-therapy]].

## Risks

The immunology cuts several ways. Innate sensing of vector DNA drives complement activation; high systemic doses have produced complement-mediated thrombotic microangiopathy, acute kidney injury, and hepatotoxicity. Adaptive responses against the capsid destroy transduced cells, a pattern first documented in a 2006 haemophilia B trial in which factor IX expression rose and then fell as capsid-specific T cells cleared the transduced hepatocytes.[^manno2006] Responses against the transgene product itself can occur when a patient has never made the protein and does not recognise it as self. Where a vector delivers a nuclease rather than a gene, [[crispr-off-target-effects|editing errors]] add a second, independent risk channel.

Deaths have occurred. Several patients died of hepatobiliary failure in a trial of a high-dose AAV therapy for X-linked myotubular myopathy, and in 2025 acute liver failure in recipients of a Duchenne muscular dystrophy therapy prompted regulatory action and a re-examination of systemic dose ceilings. Prophylactic corticosteroids, complement inhibitors, and weight-capped dosing are all responses to this pattern, and none of them is a solution.

Insertional mutagenesis is a lower but non-zero concern. Hepatocellular carcinoma driven by vector integration has been reported in mice; the human evidence remains equivocal, and long-term follow-up requirements reflect that uncertainty. Comparison with other delivery routes clarifies the trade-offs:

```compare
{
  "columns": ["AAV vector", "Lipid nanoparticle", "Lentivirus (ex vivo)"],
  "rows": [
    { "label": "Cargo capacity", "values": ["About 4.7 kb", "Large, mRNA or RNP", "Up to about 8 kb"] },
    { "label": "Expression duration", "values": ["Years, episomal", "Days", "Permanent, integrated"] },
    { "label": "Redosing", "values": ["Blocked by antibodies", "Practical", "Not applicable"] },
    { "label": "Main safety issue", "values": ["Capsid immunity, liver toxicity", "Infusion reactions", "Insertional mutagenesis"] },
    { "label": "Typical use", "values": ["Gene addition in vivo", "In vivo editing", "Blood stem cell therapy"] }
  ]
}
```

## Outlook

AAV is unlikely to be displaced soon for durable gene addition to non-dividing tissue: nothing else delivers a gene to photoreceptors or motor neurons and keeps it expressing for years. Its share of new programmes is nonetheless shrinking, because [[lipid-nanoparticles]] are better suited to the hit-and-run logic of genome editing, can be redosed, and are cheaper to make. The most consequential open questions are whether capsid engineering can produce a vector that reaches muscle or brain at doses low enough to avoid complement activation, and whether any of the antibody-clearing strategies makes redosing routine. A negative answer to both would leave AAV where it is now — extraordinarily effective for a small number of tissues and diseases, and a poor fit for anything requiring a second dose or a large payload. Delivery, not editing chemistry, has been the binding constraint on genetic medicine for thirty years, and it remains one for both [[casgevy|ex vivo cell products]] and any future application to [[gene-therapy-for-aging]].

## See also

- [[somatic-gene-therapy]]
- [[lipid-nanoparticles]]
- [[crispr-cas9]]
- [[base-editing]]
- [[gene-therapy-for-aging]]
- [[casgevy]]
- [[optogenetics]]
- [[crispr-off-target-effects]]

## References

[^pillay2016]: `paper` Pillay, S. et al. "An essential receptor for adeno-associated virus infection." *Nature*, 2016.
[^manno2006]: `paper` Manno, C.S. et al. "Successful transduction of liver in hemophilia by AAV-Factor IX and limitations imposed by the host immune response." *Nature Medicine*, 2006. {The first human trial to show the capsid T-cell response; factor IX expression rose and then fell, so it records a failure mode rather than a durable treatment effect.}
[^deverman2016]: `paper` Deverman, B.E. et al. "Cre-dependent selection yields AAV variants for widespread gene transfer to the adult brain." *Nature Biotechnology*, 2016.
[^hordeaux2018]: `paper` Hordeaux, J. et al. "The Neurotropic Properties of AAV-PHP.B Are Limited to C57BL/6J Mice." *Molecular Therapy*, 2018. {The negative result behind the species caveat: as its title states, the capsid's brain delivery is confined to one inbred mouse strain.}

==============================================================================
ARTICLE: accelerating-change
TITLE:   Accelerating change
PORTAL:  Foundational Concepts
URL:     https://futurehumanwiki.com/wiki/accelerating-change
SOURCE:  https://futurehumanwiki.com/raw/accelerating-change
==============================================================================

---
title: "Accelerating change"
slug: "accelerating-change"
type: "concept"
status: "contested"
horizon: "present"
categories: ["foundations"]
tags: ["forecasting", "moore's law", "exponential", "technology", "economics", "singularity"]
summary: "The claim that technological capability grows exponentially or faster, so that the rate of change itself increases over historical time."
updated: "2026-07-27"
issues: ["The DNA sequencing cost curve is described without a source.", "Fisher and Pry's substitution model is named but not cited."]
---

```infobox
{
  "caption": "Claim in technology forecasting",
  "rows": [
    { "label": "Best-known formulation", "value": "Law of accelerating returns" },
    { "label": "Associated with", "value": "Ray Kurzweil", "link": "/wiki/ray-kurzweil" },
    { "label": "Canonical evidence", "value": "Moore's law, 1965–2010s" },
    { "label": "Canonical counterevidence", "value": "Eroom's law in pharma" },
    { "label": "Main rival model", "value": "Sequential S-curves" },
    { "label": "Status", "value": "Domain-dependent; disputed as a general law" }
  ]
}
```

**Accelerating change** is the claim that technological capability does not merely improve but improves at an increasing rate, so that intervals between comparable advances shorten over time. In its strong form it treats acceleration as a property of technological evolution as such, extrapolable across domains and forward in time. That strong form is the engine of most long-horizon forecasting in this wiki's subject matter, and it is also the weakest link in most of it.

```keyfacts
[
  { "value": "~2 years", "label": "Transistor-count doubling time", "note": "Moore's 1975 revision; density gains have since slowed" },
  { "value": "~9 years", "label": "Halving time of new drugs per R&D dollar", "note": "Scannell et al.'s 'Eroom's law', 1950–2010" },
  { "value": "~5%/yr", "label": "Estimated decline in research productivity", "note": "Bloom et al., across US industries including semiconductors" }
]
```

## The claim

Three distinct assertions travel under this heading, and conflating them causes most of the confusion.

The **descriptive** claim is that particular technical metrics have followed exponential curves over decades. This is true of several and false of many.

The **general** claim is that acceleration is a feature of technology as a whole, because each advance provides tools for the next. [[ray-kurzweil]]'s law of accelerating returns is the best-known version: evolutionary processes build on their own products, so the returns — speed, capability, cost-effectiveness — grow exponentially, and the exponent itself grows.[^kurzweil2001]

The **predictive** claim is that these curves can be extrapolated to date future capabilities, including [[artificial-general-intelligence]], [[whole-brain-emulation]] or [[longevity-escape-velocity]]. This is the contested step, and the descriptive claim does not entail it.

## Origins

Henry Adams noticed the pattern early in the twentieth century, plotting coal output and later energy use and concluding in *The Education of Henry Adams* that thought itself was accelerating on a trajectory that would carry the coming century somewhere unrecognisable. Similar observations recur through the mid-century in the work of Buckminster Fuller and in Stanislaw Ulam's report of a conversation with John von Neumann about an approaching "essential singularity" in the history of the race beyond which human affairs could not continue as known.

The modern version dates to Gordon Moore's 1965 observation that the number of components on an integrated circuit at minimum cost per component had doubled annually and would continue to do so, revised in 1975 to roughly two years.[^moore1965] Moore was describing a manufacturing economics trend over a decade of data; it became a coordinating expectation for the semiconductor industry, and then a metaphor for technological progress in general — a transfer he did not endorse.

## The evidence usually cited

Several curves genuinely have been exponential for long stretches.

Transistor density followed Moore's law for roughly half a century. Computing energy efficiency improved on a similar cadence for decades, a relationship documented by Jonathan Koomey and colleagues, with the doubling time lengthening after around 2000.[^koomey2011] Training compute for frontier machine-learning systems doubled roughly every six months through the 2010s, far faster than hardware improvement alone, because spending grew as well.

The steepest curve in biology is DNA sequencing cost. From 2007 the introduction of massively parallel sequencing pushed cost per genome down faster than semiconductor trends for several years, before flattening into a slower decline — a pattern that makes it an instructive case rather than a clean confirmation. Where cheap sequencing has mattered most on this wiki is downstream: [[polygenic-embryo-screening]], [[epigenetic-clock|epigenetic clocks]] and [[connectomics]] all depend on data volumes that were unaffordable two decades ago.

## Moore's law and what replaced it

Dennard scaling — the property that shrinking transistors also reduced their power density, so clock speeds could rise for free — broke down in the mid-2000s. Clock rates stopped climbing, and performance gains shifted to parallelism: multiple cores, then graphics processors, then domain-specific accelerators. Feature-size shrinks have continued but more slowly and at much higher capital cost, with extreme-ultraviolet lithography and three-dimensional packaging substituting for straightforward miniaturisation. Several analyses argue that cost per transistor stopped falling around the early 2010s, which if correct means the economically relevant version of Moore's law ended over a decade before the physical one.

The important structural point is that the exponential was sustained by increasing effort. Bloom, Jones, Van Reenen and Webb found that maintaining the constant doubling rate of Moore's law required a research workforce many times larger by the 2010s than in the early 1970s, and that research productivity has fallen at roughly five per cent a year across the US industries they examined.[^bloom2020] An exponential output produced by a super-exponential input is not evidence for a law of accelerating returns. It is evidence of an expensive and possibly unsustainable subsidy.

## Where the trend fails

The clearest counterexample is drug development. Scannell and colleagues documented that the number of new drugs approved per billion dollars of inflation-adjusted R&D spending halved roughly every nine years from 1950, a decline they named "Eroom's law" — Moore's law backwards.[^scannell2012] Half a century of molecular biology, structural methods and automation improved every input to the process while the output per dollar fell, for reasons the authors attribute mainly to the rising bar set by existing treatments and to regulatory risk aversion rather than to scientific difficulty alone. Whether machine learning bends that curve back is the claim assessed under [[ai-drug-discovery]], and no drug originating from those methods has yet been approved.

Other domains show flat or reversed trends. Passenger aircraft cruise speeds peaked in the 1960s and fell after supersonic service ended. Energy consumption per capita in developed economies stopped rising decades ago. Measured "disruptiveness" of papers and patents, on one widely discussed index, declined across all major fields between the mid-twentieth century and the 2010s.[^park2023]

Within this wiki's own subject matter the record is mixed and mostly slow. [[tissue-engineering]] promised engineered organs within a generation of its 1990s founding and has delivered a handful of simple structures. [[medical-nanorobots]] as described in the 1980s remain design studies. Gene therapy took roughly three decades from first clinical trial to durable approved products. Against that, [[crispr-cas9]] moved from a 2012 mechanism paper to an approved therapy in about eleven years, which is fast by any historical standard, and [[ai-protein-design]] brought single-chain structure prediction to near-experimental accuracy within roughly a decade of deep learning being applied to it.

> [!caution] Selection effects in the evidence
> Exponential curves are easy to find retrospectively because the technologies that failed to accelerate are not the ones anyone plots. Any argument from a set of curves must state how the curves were chosen, and most popular presentations do not.

## The S-curve counterargument

The standard alternative model is that individual technologies follow logistic curves: slow start, rapid middle, saturation as physical or economic limits bind. Substitution between competing technologies has been modelled this way since Fisher and Pry's 1971 work, and the fit is generally good.

On this view an apparent long-run exponential is a chain of overlapping S-curves — vacuum tubes to transistors to integrated circuits — and the appearance of a smooth law is an artefact of aggregation. The distinction matters because the two models make different predictions at exactly the point where forecasting is most valuable. An exponential model says the next doubling is like the last. An S-curve model says a saturating technology's replacement may or may not arrive, and that the historical record contains cases where it did not.

Theodore Modis's critique of the strong version argues that the chosen milestones in canonical accelerating-change datasets are selected and weighted subjectively, and that the underlying pattern is better described by a single large logistic than by a runaway exponential.[^modis2006]

## Forecasting base rates

The final objection is empirical and unflattering. Philip Tetlock's long-running studies of expert political and geopolitical forecasting found that credentialed experts performed poorly against simple extrapolation, that confidence was inversely related to accuracy, and that the most accurate forecasters were those who updated frequently and avoided single grand theories.[^tetlock2005]

Technology-specific reviews find a characteristic distortion. Surveys of published predictions about machine intelligence over six decades found that they cluster fifteen to twenty-five years ahead of whenever the prediction was made, by experts and non-experts alike, which is the signature of a forecast anchored on career horizons rather than on evidence.[^armstrong2014]

This is why the wiki's articles on [[technological-singularity]], [[singularitarianism]] and [[effective-accelerationism]] treat timelines separately from mechanisms. The mechanism arguments can be assessed on their merits; the dates attached to them have a track record, and it is bad.

## Outlook

The defensible position is domain-specific and cheap to state: some metrics have followed exponential curves for decades, the curves are sustained by rising inputs, they saturate, and cross-domain extrapolation has no established basis. What that leaves unresolved is the question the strong claim was invented to answer. If capability in one domain — machine learning is the current candidate — begins to compound the productivity of research itself, the pattern of independent saturating curves would no longer hold, and the argument for [[differential-technological-development|deliberately sequencing which capabilities arrive first]] becomes urgent rather than theoretical. Whether that is happening is an empirical question about research productivity, and it will be answered by measurement rather than by extrapolation.

## See also

- [[technological-singularity]]
- [[ray-kurzweil]]
- [[artificial-general-intelligence]]
- [[technology-readiness-level]]
- [[singularitarianism]]
- [[longevity-escape-velocity]]
- [[future-of-humanity]]
- [[effective-accelerationism]]

## References

[^kurzweil2001]: `statement` Kurzweil, R. "The Law of Accelerating Returns." Essay, 2001; expanded in *The Singularity Is Near*, Viking, 2005. {An advocate's essay rather than a reviewed analysis; the milestones it plots are selected by the author, which is the basis of Modis's objection.}
[^moore1965]: `paper` Moore, G. E. "Cramming More Components onto Integrated Circuits." *Electronics*, 1965.
[^koomey2011]: `paper` Koomey, J. et al. "Implications of Historical Trends in the Electrical Efficiency of Computing." *IEEE Annals of the History of Computing*, 2011.
[^bloom2020]: `paper` Bloom, N., Jones, C. I., Van Reenen, J. and Webb, M. "Are Ideas Getting Harder to Find?" *American Economic Review*, 2020. {Measures output per researcher across US industries and crops; semiconductors are one case among several, not the whole finding.}
[^scannell2012]: `paper` Scannell, J. W., Blanckley, A., Boldon, H. and Warrington, B. "Diagnosing the Decline in Pharmaceutical R&D Efficiency." *Nature Reviews Drug Discovery*, 2012.
[^park2023]: `paper` Park, M., Leahey, E. and Funk, R. J. "Papers and Patents Are Becoming Less Disruptive over Time." *Nature*, 2023.
[^modis2006]: `paper` Modis, T. "The Singularity Myth." *Technological Forecasting and Social Change*, 2006.
[^tetlock2005]: `book` Tetlock, P. E. *Expert Political Judgment: How Good Is It? How Can We Know?* Princeton University Press, 2005.
[^armstrong2014]: `paper` Armstrong, S., Sotala, K. and Ó hÉigeartaigh, S. "The Errors, Insights and Lessons of Famous AI Predictions—and What They Mean for the Future." *Journal of Experimental & Theoretical Artificial Intelligence*, 2014.

==============================================================================
ARTICLE: access-and-inequality
TITLE:   Access and inequality
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/access-and-inequality
SOURCE:  https://futurehumanwiki.com/raw/access-and-inequality
==============================================================================

---
title: "Access and inequality"
slug: "access-and-inequality"
type: "concept"
status: "established"
horizon: "present"
categories: ["society"]
tags: ["justice", "pricing", "policy", "gene therapy", "global health", "equity"]
summary: "The problem of who can obtain expensive biomedical interventions, and whether biotechnology will convert economic inequality into biological inequality."
updated: "2026-07-27"
issues: ["The gene therapy list prices in the opening section are given without citations.", "The 2024 CMS outcomes-based purchasing model for sickle cell is stated without a source."]
---

```infobox
{
  "caption": "Problem in health policy and distributive justice",
  "rows": [
    { "label": "Domain", "value": "Cell and gene therapy, enhancement" },
    { "label": "Highest listed price", "value": "Over $4m per course" },
    { "label": "Canonical worry", "value": "Genetic stratification" },
    { "label": "Named in", "value": "Remaking Eden, 1997" },
    { "label": "Main policy levers", "value": "Pricing, licensing, procurement" },
    { "label": "Status", "value": "Worsening for one-time cures" }
  ]
}
```

**Access and inequality** names the question of who actually receives a biomedical intervention once it works, and the associated worry that technologies capable of altering bodies and minds will translate existing economic stratification into biological stratification. The concern is old in the literature and recent in practice: until the 2010s no medicine cost enough to make it acute. As of 2026 several licensed therapies are priced above two million dollars per patient, and the mismatch between where the diseases are and where the treatments are delivered is the central unsolved problem of the genetic-medicine era.

```keyfacts
[
  { "value": "$2m+", "label": "List price of several one-time gene therapies", "note": "United States, per patient" },
  { "value": ">75%", "label": "Share of sickle cell births in sub-Saharan Africa", "note": "where no approved gene therapy is delivered" },
  { "value": "~99%", "label": "Fall in first-line HIV antiretroviral prices", "note": "from 2000 to the 2010s, via generics and volume purchasing" }
]
```

## The arithmetic of a one-time cure

Conventional pharmaceutical pricing spreads cost over years of dosing. A cure delivered once has to recover its development cost in a single transaction, and the manufacturer's incentive is to price against the lifetime cost of the disease it displaces rather than against the cost of goods. That logic produced the current price ladder: onasemnogene abeparvovec for spinal muscular atrophy at roughly $2.1 million on US launch in 2019, etranacogene dezaparvovec for haemophilia B at roughly $3.5 million in 2022, and atidarsagene autotemcel for metachromatic leukodystrophy at roughly $4.25 million in 2024 — the highest list price ever set for a medicine. [[casgevy]] and a competing lentiviral product for sickle cell disease were approved in the United States in December 2023 at roughly $2.2 million and $3.1 million respectively.

Health economists have generally judged several of these prices defensible on cost-effectiveness grounds, because the alternative is decades of transfusions, factor concentrate, or intensive care. That defence does not answer the budget-impact problem. A public payer facing a hundred eligible patients in one year confronts a bill it cannot smooth, and the treatments are concentrated in exactly the populations — children with rare inherited disease, adults with sickle cell — least able to negotiate.

The commercial results have been poor for everyone. The first gene therapy licensed in Europe was withdrawn in 2017 after a single commercial course was sold. Two later products were pulled from European markets in 2021 when their maker could not agree prices with national payers, leaving patients in countries where the therapy was licensed with no route to receiving it. Approval and access are separate events, and the gap between them has widened.

> [!key] Why price is not the only barrier
> An autologous CRISPR therapy for sickle cell requires apheresis, a manufacturing round trip, and myeloablative conditioning with busulfan followed by weeks of inpatient care. Even at zero drug cost it needs a transplant-capable centre, a cell-processing facility, and a blood bank. Most of the world's sickle cell patients live where none of these exist.

## Where the diseases are

Sickle cell disease is the clearest case because the mismatch is total. The overwhelming majority of affected births occur in sub-Saharan Africa, with India accounting for much of the remainder;[^piel2017] the approved [[somatic-gene-therapy]] products are administered almost entirely in North America, Europe, and the Gulf. A curative technology has been developed for a disease of the global South and deployed in the global North, and the constraint is not patents alone but the absence of the delivery infrastructure the therapy assumes.

The same structure recurs across the wiki's subject matter. [[cochlear-implant]] surgery is routine in high-income countries and rare in most low-income ones, where the majority of children with profound congenital deafness live. [[organ-shortage]] is managed by transplant systems that only some states can operate. The devices described in [[brain-computer-interface]] research require neurosurgical centres and years of technical support per patient. The largest instance by volume is not a one-time cure at all: [[glp-1-receptor-agonists|incretin drugs]] for obesity and diabetes carry US list prices on the order of a thousand dollars a month, and obesity prevalence is rising fastest in the middle-income countries where they are furthest out of reach.

## The historical record on diffusion

The optimistic case rests on precedent, and the precedent is real but partial.

Antiretroviral therapy for HIV cost well over ten thousand dollars per patient per year in 2000 and was, on standard analysis, permanently out of reach for African health systems. Within roughly a decade generic fixed-dose combinations manufactured in India, mass procurement through the Global Fund and PEPFAR, and voluntary licensing had brought first-line regimens to a small fraction of that figure, and tens of millions of people were on treatment.[^msf] Direct-acting antivirals for hepatitis C followed a compressed version of the same path: a US list price of tens of thousands of dollars per course alongside voluntary licences that put courses in some low- and middle-income countries at a few hundred dollars.

Three features made those cases tractable, and all three are absent for cell and gene therapy. The products were small molecules that generic manufacturers could copy at scale. Administration required only a clinic and a pill. And the epidemic's size created political pressure sufficient to override ordinary intellectual-property arrangements.

Autologous cell therapy has none of those properties. Each dose is manufactured for one patient, so unit cost does not fall the way chemical synthesis does. The bottleneck is process and facility, not molecule. And the affected populations, though large in aggregate, are politically fragmented across dozens of rare conditions.

Against this, the cost of reading a genome fell from roughly a hundred million dollars in 2001 to a few hundred, one of the steepest cost declines recorded in any technology.[^nhgri] That curve is often cited as evidence that therapy costs will follow. It is a poor analogy: sequencing is an instrument-and-reagent problem subject to semiconductor-like scaling, while manufacturing a cell product is closer to running a small hospital.

## What could change the curve

Three technical shifts would alter the economics rather than the pricing negotiation.

**In vivo editing.** Delivering an editor directly to the patient removes cell collection, manufacturing, and conditioning at a stroke. [[lipid-nanoparticles]] carrying CRISPR components have been used systemically in humans, and the first bespoke therapy for a single infant with a urea-cycle disorder was designed and dosed within months in 2025 — a demonstration that the regulatory and manufacturing path can be compressed, though at a cost per patient that no system could generalise. Whether in vivo editing reaches the tissues that matter for sickle cell is unresolved; the target is the haematopoietic stem cell, and reaching it in the marrow is harder than reaching the liver.

**Allogeneic and off-the-shelf products.** A cell therapy manufactured in batches from donor material amortises across many patients. This is the main cost lever in the CAR-T field and applies in principle to regenerative products built on [[induced-pluripotent-stem-cells]].

**Cheap geroprotection.** If the interventions that come out of the [[geroscience-hypothesis]] turn out to be repurposed generics, the distributive picture inverts. [[metformin]] costs pennies. [[exercise-and-aging]] describes the best-evidenced intervention available, which costs nothing and is nonetheless distributed with a steep social gradient — a reminder that price is not the binding constraint on health inequality, and never has been.

## Policy responses

Payers have experimented with instruments that spread or condition payment. Outcomes-based agreements make part of the price contingent on durable benefit, which fits one-time therapies well in principle and requires long-term follow-up data that health systems are poorly equipped to collect. Subscription or "Netflix" arrangements, in which a payer buys unlimited access for a fixed annual fee, were adopted by two US states for hepatitis C treatment in 2019 and sharply increased the number of people treated in their prison and Medicaid populations. In 2024 the US Centers for Medicare and Medicaid Services announced a multi-state outcomes-based purchasing model specifically for sickle cell gene therapy, with participation beginning in 2025 — the first attempt to negotiate a gene therapy price on behalf of a public programme at national scale.

Compulsory licensing under the TRIPS agreement and the 2001 Doha declaration remains available and is almost never used for biologics, partly because a licence to the patent does not transfer the manufacturing know-how, which is the actual barrier. Regional manufacturing initiatives — mRNA technology transfer hubs, African vaccine production capacity — address that barrier directly and operate on a decadal timescale.

## The stratification argument

The specifically futurist worry is not about access to cures but about heritable advantage. Lee Silver's 1997 scenario of a society splitting into genetically enhanced and unenhanced lineages remains the reference version, and it depends on premises that have aged unevenly.[^silver1997]

The premise that has held is that expensive medicine reaches the wealthy first. The premise that has not held is that genetic enhancement would deliver large, heritable, purchasable advantage. [[genetic-enhancement-of-intelligence]] explains why: the traits people would pay for are highly polygenic, and the achievable shift from [[polygenic-embryo-screening]] is on the order of a fraction of a standard deviation under favourable assumptions, smaller across ancestries not represented in the training data, and not obviously larger than the effect of the environment money already buys. [[germline-editing]] could in principle do more, but not for polygenic traits, and its clinical use remains prohibited in most jurisdictions.

> [!debate] Which inequality is the live one
> Some argue the genetic-divide scenario distracts from the inequality that exists now: a difference of a decade or more in life expectancy between rich and poor districts of the same city, produced by ordinary causes and requiring no new technology. Others reply that a stratification which becomes heritable is categorically different from one that resets each generation, and is worth preventing before it is cheap.

Both readings agree on the near-term implication. The interventions plausibly capable of stratifying a population — screening, [[embryo-selection]], [[in-vitro-gametogenesis]] if it matures — are lightly regulated in most countries precisely because they involve selection rather than modification, and therefore fall outside the instruments described in [[governance-of-genome-editing]]. The technologies that ethics has watched most closely are not the ones distributing advantage.

## Open questions

No health system has yet solved the problem of paying for a large cohort of one-time cures arriving at once, and the cohort is growing: dozens of gene therapy programmes are in late-stage development for conditions with patient populations in the tens of thousands. Nor has any mechanism been established for delivering a curative therapy in a country without transplant infrastructure. The 2023 international summit on genome editing devoted an unusual share of its attention to this, on the argument that a governance regime concerned only with restraint has answered half the question. Nothing binding followed, and the affected populations still have no route to the treatments licensed for their disease.

## See also

- [[casgevy]]
- [[somatic-gene-therapy]]
- [[longevity-dividend]]
- [[genetic-discrimination]]
- [[enhancement-arms-race]]
- [[bioethics-of-enhancement]]
- [[governance-of-genome-editing]]
- [[disability-rights-and-enhancement]]

## References

[^silver1997]: `book` Silver, L. *Remaking Eden: Cloning and Beyond in a Brave New World*. Avon Books, 1997.
[^piel2017]: `paper` Piel, F.B., Steinberg, M.H., Rees, D.C. "Sickle Cell Disease." *New England Journal of Medicine*, 2017.
[^msf]: `report` Médecins Sans Frontières. *Untangling the Web of Antiretroviral Price Reductions*. MSF Access Campaign, successive editions from 2001. {Compiled by an advocacy organization from prices it and other buyers were quoted, so it records transaction prices rather than an independent audit.}
[^nhgri]: `report` Wetterstrand, K.A. "DNA Sequencing Costs: Data from the NHGRI Genome Sequencing Program." National Human Genome Research Institute. {Tracks production cost at funded sequencing centres, not the price a customer pays and not the cost of interpreting the result.}

==============================================================================
ARTICLE: aging-biomarkers
TITLE:   Aging biomarkers
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/aging-biomarkers
SOURCE:  https://futurehumanwiki.com/raw/aging-biomarkers
==============================================================================

---
title: "Aging biomarkers"
slug: "aging-biomarkers"
type: "concept"
status: "emerging"
horizon: "late 2020s"
categories: ["longevity"]
tags: ["biomarkers", "aging", "clinical trials", "regulation", "measurement"]
summary: "Measurable indicators intended to track the rate or extent of biological aging, sought as surrogate endpoints so that anti-aging trials can read out in years rather than decades."
updated: "2026-07-27"
issues: ["No source for the claim that no aging biomarker has passed FDA biomarker qualification.", "Proteomic and metabolomic panels are named as candidate classes without any citation."]
---

```infobox
{
  "caption": "Measurement problem in geroscience",
  "rows": [
    { "label": "Purpose", "value": "Trial surrogate endpoint" },
    { "label": "Main classes", "value": "Molecular, functional, composite" },
    { "label": "Criteria origin", "value": "AFAR, from the late 1980s" },
    { "label": "Coordinating body", "value": "Biomarkers of Aging Consortium" },
    { "label": "Regulatory status", "value": "None qualified" },
    { "label": "Bottleneck", "value": "Causal mediation evidence" }
  ]
}
```

**Aging biomarkers** are measurable quantities intended to indicate how far or how fast a person is aging, independent of the calendar. The field wants them for a specific and practical reason: without one, a trial testing whether a drug slows aging must wait for participants to become ill or die, which takes decades and enormous numbers of people. A validated biomarker would compress that to a few years. None has been accepted for the purpose.

```keyfacts
[
  { "value": "3", "label": "Validity tiers a candidate must clear", "note": "analytical, clinical, and response to intervention" },
  { "value": "None", "label": "Biomarkers qualified as aging surrogate endpoints", "note": "by any regulator, as of mid-2026" },
  { "value": "Decades", "label": "Trial duration without a surrogate", "note": "the reason TAME uses a composite disease endpoint" }
]
```

## Why the field needs one

The [[geroscience-hypothesis]] holds that slowing aging would delay many chronic diseases at once. Testing it requires an endpoint. Lifespan is unambiguous and unusable in humans. Incidence of a single disease reintroduces the disease-by-disease framing the hypothesis was built to escape. A composite of several age-related diseases plus death is the compromise adopted by the [[metformin|TAME]] design, and it still requires thousands of participants over years.

A surrogate biomarker would change the economics. A six-month readout would let sponsors run dose-ranging studies, compare candidate compounds, and fail fast, which is how drug development works in every field that has one. The absence of a surrogate is a substantial part of why so few geroprotector candidates have entered serious human testing despite a large animal literature.

## What counts as a biomarker of aging

The criteria most often cited descend from work sponsored by the American Federation for Aging Research in the late 1980s and restated since.[^baker1988] A biomarker of aging should predict remaining lifespan or the onset of age-related dysfunction better than chronological age does; it should reflect an underlying process, ideally one of the [[hallmarks-of-aging]], rather than a specific disease; it should be measurable repeatedly without significant harm; and it should work in laboratory animals as well as in humans, so that it can be validated against lifespan in a species where lifespan is observable.

Modern framings add a decomposition into three tiers. Analytical validity asks whether the measurement is reproducible across laboratories, platforms and repeat draws. Clinical validity asks whether it predicts outcomes in independent cohorts. Responsiveness asks whether it moves when an intervention is applied, and, crucially, whether a change in the marker predicts a corresponding change in the outcome. Most candidates have some evidence for the first two tiers and almost none for the third.

## Candidate classes

Candidates are conventionally sorted by what they measure, a catalogue set out under [[biological-age]]. Sorted instead by which validity tier each class is stuck at, the picture is more useful, because the tier a class fails is what determines the experiment that would rescue it.

**Molecular measures are cheap and hard to interpret.** DNA methylation dominates, in the form of [[epigenetic-clock|epigenetic clocks]], with proteomic, metabolomic, glycomic and transcriptomic panels alongside. One blood draw yields tens of thousands of features, and a model can be fitted from them to almost any outcome. That is simultaneously the attraction and the difficulty: an index selected for prediction carries no evidence that moving it moves anything, and that evidence cannot come from the data the index was fitted on.

**Cellular measures are the ones the field most wants and least has.** Senescent-cell burden is the direct target of [[senolytics]], and no accepted method exists for quantifying it in a living person, because no clean marker of [[cellular-senescence]] exists to build the method on. Assays therefore rely on surrogates for a surrogate. Clonal haematopoiesis, a readout of [[stem-cell-exhaustion]], is the exception: measurable by sequencing, predictive of cardiovascular as well as blood outcomes, and confined to one tissue.

**Immune and inflammatory measures track a real axis and move for the wrong reasons.** Composite immune-aging scores, thymic output measures and circulating markers linked to [[inflammaging]] all decline with age. They also shift within days after infection, injury or vaccination, which is close to disqualifying for something meant to read a decades-long process off a single timepoint.

**Functional measures are the only class with real responsiveness data.** Gait speed, grip strength, chair-rise time and cardiorespiratory fitness predict mortality and disability well, and they already serve as endpoints in frailty and sarcopenia trials, so the regulatory path for them is partly mapped. The awkwardness is commercial rather than scientific: they improve reliably with training, so a sponsor whose drug is judged on grip strength is competing against [[exercise-and-aging|exercise]].

## Coordination efforts

The Biomarkers of Aging Consortium, established in the early 2020s, brought together academic groups, companies and funders to standardise definitions, share data and set validation criteria. Its output includes a framework paper setting out what would be required for a biomarker to be used in longevity-intervention trials and subsequent work on validation standards.[^moqri2023] The consortium's contribution is less a new measure than an agreed vocabulary and a shared benchmark, which the field previously lacked.

> [!debate] Surrogates can mislead
> Cardiology provides the cautionary case. Antiarrhythmic drugs suppressed ventricular ectopic beats, a plausible surrogate for sudden cardiac death, and the Cardiac Arrhythmia Suppression Trial found they increased mortality.[^echt1991] A marker can be strongly associated with an outcome and still be the wrong thing to move.

## The regulatory bar

Regulators do not accept surrogates because they correlate with outcomes. They require evidence that an intervention's effect on the surrogate accounts for its effect on the clinical outcome, which is a mediation claim and much harder to establish. In the United States, formal qualification runs through a biomarker qualification programme, and no aging biomarker has been submitted successfully through it.

A second obstacle is categorical. Aging is not recognised as an indication by the major regulators, so even a perfect biomarker would be measuring progress against something a label cannot name. This is why proposals in the field are usually framed around a specific age-related condition or around multimorbidity, and why the ICD classification of aging-related decline has been argued over with unusual intensity for a coding question. A working group convened around [[nir-barzilai|Nir Barzilai]]'s TAME trial published a shortlist of blood-based markers judged closest to usable for geroscience trials, chosen for availability and predictive validity rather than for mechanistic depth.[^justice2018] The economic case for clearing this obstacle is set out under [[longevity-dividend]].

## Validation problems

Validating a biomarker of aging against mortality is partly circular: mortality is the outcome, and a marker that predicts it may be measuring current illness rather than aging. Distinguishing the two requires long follow-up in healthy cohorts, which few datasets support.

Animal validation offers a partial escape. A marker that tracks lifespan across mouse strains and across interventions of known effect, including [[caloric-restriction]] and [[rapamycin]], has passed a test no human dataset can supply. Cross-species methylation clocks, which [[steve-horvath|Horvath]] and collaborators extended across mammalian species, were built partly for this purpose. The translation gap remains: a marker calibrated in mice may not carry the same meaning in a species that lives thirty times longer.

Reliability is the mundane problem that undermines many published intervention effects. Where a measure's test-retest variation is comparable to the effect being claimed, the claim cannot be evaluated, and this has been documented for several widely used clocks and partly remedied by reconstructing them on principal components.[^higgins2022] The same difficulty afflicted [[telomeres-and-telomerase|leukocyte telomere length]], the field's previous favourite marker, which proved too noisy at the individual level to be informative.

## Outlook

Two paths are open. One is to keep improving molecular measures until one clears the mediation bar, which will require intervention trials large enough to demonstrate that the marker carries the treatment effect. The other is to give up on surrogates for now and use function directly, which is the choice made by [[xprize-healthspan]] in defining its target as restored muscle, cognitive and immune performance. The second path is slower per trial and immune to the criticism that has attached to biological-age claims. Which route the field takes will shape whether the first credible human geroprotector result arrives from a biomarker readout or from a decade-long clinical endpoint.

## See also

- [[biological-age]]
- [[epigenetic-clock]]
- [[geroscience-hypothesis]]
- [[healthspan]]
- [[hallmarks-of-aging]]
- [[xprize-healthspan]]
- [[compression-of-morbidity]]

## References

[^baker1988]: `paper` Baker, G.T. & Sprott, R.L. "Biomarkers of aging." *Experimental Gerontology*, 1988.
[^moqri2023]: `paper` Moqri, M. et al. "Biomarkers of aging for the identification and evaluation of longevity interventions." *Cell*, 2023.
[^echt1991]: `paper` Echt, D.S. et al. "Mortality and morbidity in patients receiving encainide, flecainide, or placebo: the Cardiac Arrhythmia Suppression Trial." *New England Journal of Medicine*, 1991.
[^justice2018]: `paper` Justice, J.N. et al. "A framework for selection of blood-based biomarkers for geroscience-guided clinical trials: report from the TAME Biomarkers Workgroup." *GeroScience*, 2018.
[^higgins2022]: `paper` Higgins-Chen, A.T. et al. "A computational solution for bolstering reliability of epigenetic clocks." *Nature Aging*, 2022. {Addresses test-retest noise by rebuilding existing clocks on principal components; it does not establish that any of them measures aging.}

==============================================================================
ARTICLE: ai-drug-discovery
TITLE:   AI drug discovery
PORTAL:  Foundational Concepts
URL:     https://futurehumanwiki.com/wiki/ai-drug-discovery
SOURCE:  https://futurehumanwiki.com/raw/ai-drug-discovery
==============================================================================

---
title: "AI drug discovery"
slug: "ai-drug-discovery"
type: "technology"
status: "emerging"
horizon: "late 2020s"
trl: 5
categories: ["foundations", "longevity"]
tags: ["machine learning", "drug development", "generative chemistry", "target discovery", "clinical trials", "pharmacology"]
summary: "The use of machine learning to select drug targets, generate candidate molecules and predict failure, so far demonstrably faster at making compounds than at making medicines."
updated: "2026-07-27"
humanEvidence: "Several AI-derived candidates have reached phase 1 and phase 2 trials in people; as of mid-2026 none has completed phase 3 or been approved, and the largest clinical success is a repurposed existing drug."
access: "Platforms are proprietary to pharmaceutical companies or licensed to them, and some models and datasets are public; no molecule designed this way is available outside a trial, though AI-suggested uses of approved drugs are."
reversibility: "context"
issues: ["Needs a citable source for the absence of any approved AI-discovered drug.", "The phase 1 to approval rate is uncited; Cook 2014 supports the efficacy-failure point only."]
---

```infobox
{
  "caption": "Computational method in pharmaceutical research",
  "rows": [
    { "label": "Type", "value": "Discovery and design method" },
    { "label": "Main tasks", "value": "Target choice, molecule generation, property prediction" },
    { "label": "First candidate in trials", "value": "2020" },
    { "label": "Furthest clinical stage", "value": "Phase 2" },
    { "label": "Drugs approved from the approach", "value": "None as of 2026" },
    { "label": "Main criticism", "value": "Attrition is a biology problem" },
    { "label": "Readiness", "value": "TRL 5" }
  ]
}
```

**AI drug discovery** is the application of machine learning to the early stages of pharmaceutical
research: choosing which protein to drug, generating molecules that might bind it, predicting which
of those will be soluble, selective and non-toxic, and planning routes to synthesize them. The
chemistry claims are the better supported ones: software now proposes synthesizable, potent-looking
compounds against a defined target at a speed no medicinal-chemistry team matches by hand, though how
much of that speed the models themselves explain is disputed. The clinical claims are not settled. As
of mid-2026 several AI-derived candidates have entered phase 1 and phase 2 trials, and none has
completed phase 3 or been approved.

## What the models actually do

The label bundles at least five tasks that succeed and fail independently.

**Target selection** ranks proteins whose modulation might change a disease, using expression data,
genetic association studies, knowledge graphs assembled from the literature, and increasingly
language models reading that literature directly. The most consequential public case is
BenevolentAI's identification of baricitinib, an approved rheumatoid arthritis drug,
as a plausible COVID-19 treatment in early 2020.[^richardson2020] Baricitinib went on to be
authorized for that use. It is a genuine result and a narrow one: the system reranked known
pharmacology rather than producing a molecule.

**Molecule generation** trains models over SMILES strings, molecular graphs or three-dimensional
poses, then steers them with reinforcement learning or conditions them on a binding pocket.
Insilico Medicine's 2019 report on the kinase DDR1 was the first widely noticed demonstration,
claiming designed inhibitors in weeks rather than months.[^zhavoronkov2019]

**Property prediction** estimates absorption, metabolism, cardiac ion-channel liability and toxicity
before synthesis. This is the least glamorous application and probably the most useful, because it
kills bad series cheaply.

**Screening triage** trains a classifier on a modest set of measured compounds and applies it to
libraries too large to test. A graph neural network trained on a few thousand molecules tested
against *E. coli* nominated halicin, a compound originally investigated for diabetes, which killed
several drug-resistant bacterial species in culture and in mice.[^stokes2020]

**Structure and synthesis.** Protein structure prediction, covered under [[ai-protein-design]],
opened pocket-based design to proteins with no experimental structure, though benchmark studies of
docking have found predicted structures to perform worse than experimental ones. Retrosynthesis
models propose routes, sometimes coupled to automated synthesis platforms so that design, make and
test run as a closed loop.

## Development history

```timeline
[
  { "year": "1964", "title": "QSAR formalized", "text": "Corwin Hansch and Toshio Fujita relate biological activity to physicochemical descriptors, establishing that chemical structure can be regressed against effect." },
  { "year": "2012", "title": "Deep learning wins a pharma benchmark", "text": "A team from Geoffrey Hinton's group wins Merck's Molecular Activity Challenge with multi-task neural networks, drawing pharmaceutical attention to the method." },
  { "year": "2019", "title": "Generative chemistry demonstrated", "text": "Insilico Medicine reports designing DDR1 kinase inhibitors with a generative model and validating them in mice within weeks." },
  { "year": "2020", "title": "First AI-designed molecule in the clinic", "text": "DSP-1181, from Exscientia and Sumitomo Dainippon Pharma, enters a phase 1 trial in Japan for obsessive-compulsive disorder. It was later discontinued." },
  { "year": "2020", "title": "Antibiotic found by screening model", "text": "The Collins laboratory identifies halicin using a graph neural network, the first widely cited antibiotic hit attributed to deep learning." },
  { "year": "2021", "title": "AlphaFold2 released", "text": "Predicted structures for most of the human proteome are made public, changing what structure-based design can be attempted." },
  { "year": "2021–2022", "title": "AI-derived target and molecule reach humans", "text": "Insilico's ISM001-055, a TNIK inhibitor for idiopathic pulmonary fibrosis, enters first-in-human testing after both target and compound were nominated computationally." },
  { "year": "2024", "title": "Consolidation", "text": "Exscientia and Recursion merge, combining two of the largest AI-first drug discovery companies after neither had produced an approved medicine." }
]
```

## What has reached patients

The most advanced case is Insilico's ISM001-055, later named rentosertib, an inhibitor of the kinase
TNIK for idiopathic pulmonary fibrosis. Both the target and the molecule came out of the company's
own platforms, which makes it the cleanest available test of the full pipeline. It cleared phase 1,
and results from a twelve-week randomized placebo-controlled phase 2a trial in China were published
in 2025: one dose arm showed a mean improvement in forced vital capacity where the placebo arm
declined.[^xu2025] That trial was not sized to establish clinical benefit, and the compound had not
entered phase 3 as of mid-2026.

The rest of the first wave has behaved like ordinary pharmacology. DSP-1181 was announced in 2020 as
the first AI-designed molecule to enter a clinical trial and did not progress beyond phase 1; a later
Exscientia oncology candidate was also dropped after early data. In 2024 Exscientia merged with
Recursion, whose approach centred on high-throughput cellular imaging rather than generative
chemistry. Neither company had an approved drug at the time.

The neighbouring field of computational protein design has produced no approved medicine of its own
either. A designed nanoparticle scaffold is a component of a COVID-19 vaccine approved in South
Korea, but as of 2026 no de novo designed protein has been approved anywhere as a therapeutic drug.

> [!caution] What "AI-discovered" means
> No regulator recognizes the category, and companies apply the label to everything from a molecule
> drawn by a generative model to a conventional campaign in which a classifier filtered one plate.
> The US Food and Drug Administration's 2025 draft guidance addresses AI used to produce evidence
> submitted in support of a decision, not the provenance of a molecule.[^fda2025] There is no
> registry, no definition, and therefore no denominator against which to count successes.

## Why faster chemistry may not help

The mainstream objection is structural rather than technical. Roughly one compound in ten that
enters phase 1 is eventually approved, and the largest single cause of failure is lack of efficacy in
phase 2: the drug engages its target and the patients do not improve. AstraZeneca's published
post-mortem on its own pipeline reached that conclusion and rebuilt the company's project criteria
around target and patient selection rather than chemistry.[^cook2014] That is a statement about
biology: the target was wrong, or the animal model misled, or the disease is
heterogeneous in ways the trial did not stratify. Generating better molecules faster addresses a
step that was not the binding constraint.

The strongest evidence on what does move the number points the same way. Drug targets supported by
human genetic evidence are substantially more likely to reach approval than those without it, which
is an argument for better target biology rather than better chemistry.[^nelson2015] Target-selection
models are trained largely on the published literature that produced the current failure rate, so it
is unclear whether they can do more than re-rank the field's existing beliefs efficiently.

> [!debate] The disagreement
> Proponents argue the objection misses compounding effects: cheaper cycles mean more targets tested,
> cleaner selectivity means fewer toxicity failures, and structure prediction opens proteins that had
> no tractable starting point. Critics answer that none of this has yet changed a phase 2 outcome,
> that the industry has absorbed several such tool revolutions without moving approval rates, and
> that the case rests on an argument rather than on a result. Both sides accept that the question is
> empirically open until a candidate designed this way succeeds or fails in phase 3.

## Data, benchmarks and the missing counterfactual

Public bioactivity data is a record of what people chose to measure and publish. Negative results are
under-reported, assays are not comparable across laboratories, and chemical space is sampled around
scaffolds that were already interesting. Models trained on it inherit that shape.

Retrospective benchmarks compound the problem. Widely used virtual-screening benchmark sets contain
biases that allow a model to score well by learning artefacts of how decoy molecules were
assembled rather than anything about binding.[^chen2019] Because the counterfactual campaign is never
run, a company cannot demonstrate that its molecule arrived faster than it would have otherwise, and
almost every published speed claim is a comparison against an industry average rather than a control.

## Aging as a test case

Aging biology is where the approach is most enthusiastically applied and least easily judged. Insilico
began as an aging-focused company and uses [[epigenetic-clock|deep aging clocks]] internally;
[[calico]], [[altos-labs]] and others run computational groups of their own. Screening models have
produced real hits: a classifier trained on known senolytic compounds identified three natural
products that killed [[cellular-senescence|senescent human cells]] in culture, at a small fraction of
the cost of the physical screen.[^smer2023] Those hits then join every other [[senolytics|senolytic
candidate]] in the queue for human evidence that does not yet exist.

The constraint here is sharper than in oncology. A trial of a [[geroscience-hypothesis|geroprotector]]
has no accepted endpoint, no validated [[aging-biomarkers|surrogate measure]], and a natural readout
measured in decades. Compressing discovery from four years to one changes little when the limiting
step is a trial nobody knows how to design. Better preclinical models, whether [[organoids]] or
simulation approaches such as [[human-digital-twins]], plausibly matter more to
[[healthspan|healthspan research]] than faster chemistry does — though the simulation side is
least developed in exactly the systems aging research needs, metabolism and immunity, and no
model of a whole person exists.

## Risks

The clearest demonstrated risk is inversion. Researchers who rewarded a toxicity-prediction model for
lethality rather than penalizing it generated tens of thousands of candidate toxic molecules,
including analogues of known nerve agents, in under a day of computing.[^urbina2022] The molecules
were never synthesized, and the barrier to weaponization remains chemistry and delivery rather than
design, but the episode is the standard reference in [[dual-use-research]] debates about model
release and sits inside the broader [[existential-risk|catastrophic-risk]] literature on biological
misuse.

Less dramatic risks are more likely to bite. Mining the same public data pushes many groups toward
the same targets, which narrows rather than broadens the pipeline. Discovery costs are a small
fraction of total development spending, so savings there do little for
[[access-and-inequality|drug pricing and access]]. And a claim of computational provenance is a
marketing asset, which creates pressure to describe conventional programmes in those terms.

## Outlook

The reversibility of anything produced this way is a property of the molecule, not the method: a
small molecule with a short half-life clears, a covalent inhibitor does not, and the same tools are
being applied to [[gene-therapy-for-aging|gene therapies]] whose effects are permanent. The design
route carries no distinct risk profile of its own, which is one reason regulators treat AI-derived
candidates like any others.

Judgment will take years. The decisive test is whether AI-nominated targets fail in phase 2 at the
historical rate or better, which requires dozens of shots and roughly a decade of readouts, not the
handful of trials available in 2026. Meanwhile the most defensible gains are in places that attract
no headlines: assay triage, guide design that reduces
[[crispr-off-target-effects|off-target editing]], formulation work on
[[lipid-nanoparticles|delivery vehicles]] and [[targeted-drug-delivery|targeting ligands]], and
routine property prediction. Nothing about the field's [[technology-readiness-level|readiness]]
supports the [[accelerating-change|compressed-timeline]] claims made for it, or the assumption that
progress toward [[artificial-general-intelligence|more general AI]] transfers cleanly into biology.
The drug class most often named as the past decade's largest pharmacological advance, the
[[glp-1-receptor-agonists|GLP-1 receptor agonists]], came from decades of peptide endocrinology with
no machine learning involved.

## See also

- [[ai-protein-design]]
- [[human-digital-twins]]
- [[geroscience-hypothesis]]
- [[senolytics]]
- [[aging-biomarkers]]
- [[dual-use-research]]
- [[artificial-general-intelligence]]
- [[organoids]]

## References

[^richardson2020]: `paper` Richardson, P. et al. "Baricitinib as potential treatment for 2019-nCoV acute respiratory disease." *The Lancet*, 2020. {Baricitinib was already approved for rheumatoid arthritis, so the system reranked known pharmacology rather than proposing a new molecule.}
[^zhavoronkov2019]: `paper` Zhavoronkov, A. et al. "Deep learning enables rapid identification of potent DDR1 kinase inhibitors." *Nature Biotechnology*, 2019. {DDR1 was a well-precedented target and the compounds resembled known inhibitors; chemists disputed how much of the speed the model explained.}
[^stokes2020]: `paper` Stokes, J.M. et al. "A Deep Learning Approach to Antibiotic Discovery." *Cell*, 2020. {Halicin's activity was confirmed in mouse infection models; it has not entered human trials.}
[^xu2025]: `paper` Xu, Z. et al. "A generative AI-discovered TNIK inhibitor for idiopathic pulmonary fibrosis: a randomized phase 2a trial." *Nature Medicine*, 2025. {Seventy-one patients across three dose arms and placebo, treated for twelve weeks; a trial that size and length cannot establish clinical benefit in fibrosis.}
[^cook2014]: `paper` Cook, D. et al. "Lessons learned from the fate of AstraZeneca's drug pipeline: a five-dimensional framework." *Nature Reviews Drug Discovery*, 2014.
[^nelson2015]: `paper` Nelson, M.R. et al. "The support of human genetic evidence for approved drug indications." *Nature Genetics*, 2015. {The association is between genetic support and eventual approval; it does not show that adding genetics to a programme causes success.}
[^chen2019]: `paper` Chen, L. et al. "Hidden bias in the DUD-E dataset leads to misleading performance of deep learning in structure-based virtual screening." *PLoS ONE*, 2019. {The bias sits in the benchmark rather than in any one model, which makes published virtual-screening gains hard to compare.}
[^smer2023]: `paper` Smer-Barreto, V. et al. "Discovery of senolytics using machine learning." *Nature Communications*, 2023.
[^urbina2022]: `paper` Urbina, F. et al. "Dual use of artificial-intelligence-powered drug discovery." *Nature Machine Intelligence*, 2022. {The molecules were generated in silico and never synthesized; the paper reports a capability, not a released agent.}
[^fda2025]: `regulator` US Food and Drug Administration. *Considerations for the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products*. Draft guidance, 2025.

==============================================================================
ARTICLE: ai-protein-design
TITLE:   AI protein design
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/ai-protein-design
SOURCE:  https://futurehumanwiki.com/raw/ai-protein-design
==============================================================================

---
title: "AI protein design"
slug: "ai-protein-design"
type: "technology"
status: "emerging"
horizon: "present"
trl: 5
categories: ["genetics", "foundations"]
tags: ["machine learning", "protein design", "structural biology", "alphafold", "biosecurity", "drug discovery"]
summary: "Machine-learning systems that predict what shape a protein sequence folds into and that generate new proteins with no natural counterpart."
updated: "2026-07-27"
humanEvidence: "A computationally designed nanoparticle scaffold is a component of a COVID-19 vaccine approved for use in people in South Korea, and a designed cytokine mimic reached early-phase human trials; no de novo designed protein is approved as a therapeutic."
access: "Most prediction and design code is open source and the AlphaFold structure database is public, so the models are free to run; the gene synthesis and wet-lab testing that turn a design into a molecule are not."
reversibility: "context"
issues: ["Clinical-stage designed proteins need a fuller survey as trials report", "Complex and antibody prediction accuracy stated qualitatively for want of a stable benchmark figure"]
---

```infobox
{
  "caption": "Machine learning applied to proteins",
  "rows": [
    { "label": "Two problems", "value": "Structure prediction, de novo design" },
    { "label": "Prediction systems", "value": "AlphaFold, RoseTTAFold, ESMFold" },
    { "label": "Design systems", "value": "RFdiffusion, ProteinMPNN" },
    { "label": "Prediction milestone", "value": "CASP14, 2020" },
    { "label": "Nobel Prize", "value": "Chemistry, 2024" },
    { "label": "Approved designed therapeutic", "value": "None as of 2026" },
    { "label": "Readiness", "value": "TRL 5" }
  ]
}
```

**AI protein design** covers two distinct uses of machine learning on proteins: predicting the
three-dimensional structure a given amino acid sequence folds into, and generating new sequences
that fold into a shape chosen in advance. The first was largely solved for single chains between
2018 and 2021, mostly by DeepMind's AlphaFold. The second is being solved unevenly, by generative
models from David Baker's laboratory and others, and it is the half that produces molecules which
have never existed. AlphaFold predicts; it does not design.

## Two problems, not one

```compare
{
  "columns": ["Structure prediction", "De novo design"],
  "rows": [
    { "label": "Question asked", "values": ["What shape does this sequence take?", "What sequence takes the shape I want?"] },
    { "label": "Representative systems", "values": ["AlphaFold2 and 3, RoseTTAFold, ESMFold", "RFdiffusion, ProteinMPNN, Rosetta"] },
    { "label": "Input", "values": ["An amino acid sequence", "A target, a fold, or a functional site"] },
    { "label": "Output", "values": ["Atomic coordinates and a confidence score", "A backbone and a sequence to build it"] },
    { "label": "Landmark demonstration", "values": ["CASP14 blind assessment, 2020", "Top7, 2003; RFdiffusion binders, 2023"] },
    { "label": "How failure shows up", "values": ["A confidently wrong model nobody checks", "A protein that does not express or fold"] }
  ]
}
```

Prediction rests on Christian Anfinsen's finding that a small protein's sequence contains the
information needed to specify its folded structure. Reading that information out resisted fifty
years of physics-based attack. From 1994 the field measured its progress through CASP, a biennial
blind assessment run by John Moult in which groups predict structures already solved
experimentally but not yet published. AlphaFold2's entry at CASP14 in 2020 returned a median score
above 90 on the assessment's main accuracy metric, close to the level at which two experimental
determinations of the same protein differ from each other.[^jumper2021]

Design is the inverse problem, with no Anfinsen result to guarantee it is solvable. The Baker
laboratory's Top7, reported in 2003, was the first protein designed to a fold not found in nature
whose crystal structure matched the design.[^kuhlman2003] For the next fifteen years design meant
energy-function optimisation in Rosetta, at success rates low enough that each published design
stood on much failed protein.

## How the systems work

AlphaFold2 takes a sequence, retrieves a multiple sequence alignment of its evolutionary relatives,
and reasons jointly over that alignment and over a matrix of residue pairs before predicting atomic
coordinates directly. The alignment does much of the work: residues that mutate in a correlated way
across species tend to touch in three dimensions. AlphaFold3, published in 2024, replaced the final
geometric module with a diffusion-based generator and extended the model to complexes of proteins
with DNA, RNA, small molecules and ions.[^abramson2024] Protein language models trained on sequence
alone drop the alignment, trading some accuracy for speed and for coverage of sequences with no
known relatives.

Design runs in two stages. A generative model such as RFdiffusion, adapted from a structure
predictor by training it to denoise corrupted backbones, produces a protein shape to order: a binder
against a specified epitope, a symmetric assembly, or a scaffold holding catalytic residues in
position.[^watson2023] A second model, ProteinMPNN, then reads that
backbone and proposes sequences likely to fold into it, raising experimental success rates well
above what Rosetta's energy functions achieved.[^dauparas2022] Designs are filtered in silico by
running a structure predictor on the proposed sequence and asking whether it folds back to the
intended shape; survivors are ordered as synthetic genes, expressed, and tested. That filter is
partly circular, since a design that fools the predictor is selected for alongside one that works.

> [!note] What "de novo" means here
> A de novo designed protein is one whose sequence was chosen computationally rather than copied
> from a natural one. It is not a synonym for "new drug": most engineered biologics, including
> nearly all therapeutic antibodies, are natural scaffolds modified at a few positions.

## Development history

```timeline
[
  { "year": "1972", "title": "Sequence determines structure", "text": "Christian Anfinsen shares the Nobel Prize in Chemistry for work showing that a denatured enzyme refolds spontaneously, implying the fold is encoded in the sequence." },
  { "year": "1994", "title": "CASP begins", "text": "John Moult launches a biennial blind assessment in which groups predict structures already solved but not yet released, giving the field an honest scoreboard." },
  { "year": "2003", "title": "Top7", "text": "Kuhlman and colleagues in the Baker laboratory design a protein with a fold not seen in nature and confirm it crystallographically." },
  { "year": "2018", "title": "AlphaFold at CASP13", "text": "DeepMind's first entry places first among predictors while remaining well short of experimental accuracy." },
  { "year": "2020", "title": "CASP14", "text": "AlphaFold2 predicts most single-chain targets at near-experimental accuracy, and the assessors describe the problem as largely solved for such targets." },
  { "year": "2021", "title": "Methods published and opened", "text": "AlphaFold2 is published with its code released; the Baker laboratory publishes RoseTTAFold, a smaller three-track network, within weeks." },
  { "year": "2022", "title": "Prediction becomes infrastructure", "text": "The AlphaFold Protein Structure Database expands to over 200 million predicted structures, and ProteinMPNN makes sequence design routine." },
  { "year": "2023", "title": "Generative backbones", "text": "RFdiffusion generates protein backbones to order, including binders against chosen targets and scaffolds for catalytic sites." },
  { "year": "2024", "title": "AlphaFold3 and the Nobel", "text": "AlphaFold3 extends prediction to biomolecular complexes; the Nobel Prize in Chemistry goes half to David Baker for computational protein design and half jointly to Demis Hassabis and John Jumper for structure prediction." }
]
```

The 2024 prize is the clearest official statement that these are two achievements rather than one.
Half went to David Baker for computational protein design; the other half was shared by Demis
Hassabis and John Jumper for structure prediction.[^nobel2024]

## Current state

Prediction has become laboratory infrastructure. The AlphaFold database offers a predicted structure
for essentially every sequence in the major reference databases, and models are used routinely to
generate hypotheses about uncharacterised proteins and to help solve experimental structures. That
the code and the database are free is unusual for a result of this weight, and helps explain why it
diffused in months rather than years.

Design has produced working molecules in narrower domains. Miniprotein binders against viral surface
proteins were designed and shown to bind and neutralise in laboratory assays within months of a
target sequence becoming available. Computationally designed self-assembling nanoparticles, of a
kind that also interests workers in [[dna-nanotechnology]], serve as vaccine scaffolds: a
two-component designed particle displaying a coronavirus receptor-binding domain became a vaccine
approved in South Korea in 2022, the clearest case of a designed protein reaching people inside an
approved product.[^walls2020] Designed enzymes exist, including luciferases built around a chosen chemistry,
though catalytic efficiencies for de novo designed enzymes generally remain well below those of
their natural counterparts.

The methods are also being turned on the proteins other biotechnologies depend on: variants of the
nucleases behind [[crispr-cas9]] and [[base-editing]], generated by sequence models rather than
found in bacteria; engineered capsids that redirect [[aav-vectors]] used in [[somatic-gene-therapy]]; and
designed binding domains for [[epigenome-editing]]. [[newlimit]] applies related machine learning to
a different question, which transcription factor combinations to deliver.

As of 2026 no de novo designed protein has been approved anywhere as a therapeutic drug. The South
Korean vaccine is the nearest case, and even there the designed part is the scaffold rather than the
antigen the immune system is being trained on. A designed mimic of interleukin-2, reported in 2019
and meant to separate the cytokine's antitumour signalling from its toxicity, reached early-phase
clinical testing in patients before its developer discontinued the programme.[^silva2019]

## Limitations

A predicted structure is not a predicted function. The models return coordinates, not a mechanism,
a binding partner, or a catalytic rate, and the shape of an uncharacterised protein often leaves its
biology as obscure as before. They also return one static conformation, when what matters is
frequently the movement between conformations or the response to a ligand.

Three failure modes are well documented. Intrinsically disordered regions, a large fraction of the
human proteome, have no single structure to predict; the models flag them with low confidence, which
is useful, but a confidence score is not a description. Prediction of complexes, including
antibody–antigen pairs, remains substantially worse than prediction of single chains, which is one
reason CASP has kept running. Most importantly for medicine, the models are largely insensitive to single
amino acid substitutions: a sequence carrying a destabilising disease mutation usually returns
almost the same confident structure as the healthy one.[^buel2022] Benchmark studies of drug docking
have likewise found predicted structures to perform worse than experimental ones.

> [!caution] Structure is not evidence
> Confidence scores estimate how closely a model matches what an experiment would find, not whether
> the protein does what a paper claims. Predicted structures have already entered the literature as
> though they were determinations, and a confidently wrong model is harder to catch than an
> obviously wrong one.

Design has its own ceiling. Success rates for de novo binders against a new target remain low and
vary sharply from target to target, so the method depends on high-throughput screening rather than
replacing it. Function beyond binding is much harder than shape. And because designed sequences do not
occur in nature, their behaviour in a human immune system is hard to predict in advance: anti-drug
antibodies are a familiar problem even for biologics built from human proteins.

## Risks

Reversibility depends on delivery. A designed protein given as a drug clears from the body like any
other biologic; the same protein expressed from a delivered gene, or built into an organism that
replicates, does not.

The biosecurity concern is specific. Orders for synthetic DNA are screened by many providers against
databases of sequences of concern, and that screening is one of the few practical chokepoints in
[[dual-use-research]] on pathogens and toxins. A 2025 study led by researchers at Microsoft showed
that open design tools could generate variants of toxin proteins that evaded existing screening
software; patches were distributed to screening providers before publication.[^wittmann2025] It is
read both as evidence that the risk is real and that the field can respond, and it sits alongside
questions about [[synthetic-genomes]], [[recoded-organisms]] and [[mirror-life]] on the agenda of
biosecurity bodies.

Researchers in the field signed voluntary commitments in 2024 to synthesis screening and to
restraint in releasing certain capabilities. The comparison usually drawn is with the
[[asilomar-conference]] of 1975, and it carries the same objection: self-governance worked when the
community controlled the reagents, and open model weights are not controlled that way.
[[existential-risk|Catastrophic-risk]] analysts generally rank engineered pathogens above designed
proteins as a near-term concern, though the tools overlap.

## Outlook

Two questions decide how much the technology matters. The first is whether prediction can be pushed
from static structures to the quantities that determine whether a molecule works: binding affinity,
conformational ensembles, and the effect of a mutation. The second is whether design can deliver
function rather than form, since a de novo enzyme matching a natural one would be a stronger result
than any binder.

The deflationary case deserves stating. Producing candidate molecules was never the slowest step in
developing a drug; recruitment and clinical failure are, and the largest single cause of that
failure is lack of efficacy rather than any property of the molecule, which better structures do
not address. The same argument shapes debate over [[ai-drug-discovery]] and over whether machine-designed
[[targeted-drug-delivery|delivery vehicles]] and [[lipid-nanoparticles|nanoparticle formulations]]
shorten timelines in practice. Demis Hassabis has presented AlphaFold as an early demonstration of
what progress toward [[artificial-general-intelligence]] would buy for science. The counter-argument
is not that the achievement is small but that the problem was unusually well posed: a discrete input,
a checkable output, and fifty years of curated experimental structures to learn from. Most of
biology offers none of those.

## See also

- [[ai-drug-discovery]]
- [[crispr-cas9]]
- [[synthetic-genomes]]
- [[dual-use-research]]
- [[mirror-life]]
- [[aav-vectors]]
- [[dna-nanotechnology]]
- [[artificial-general-intelligence]]

## References

[^jumper2021]: `paper` Jumper, J. et al. "Highly accurate protein structure prediction with AlphaFold." *Nature*, 2021. {The paper describes the system entered blind at CASP14 in 2020; the accuracy claim rests on that assessment rather than on the authors' own benchmarks.}
[^kuhlman2003]: `paper` Kuhlman, B. et al. "Design of a novel globular protein fold with atomic-level accuracy." *Science*, 2003.
[^dauparas2022]: `paper` Dauparas, J. et al. "Robust deep learning-based protein sequence design using ProteinMPNN." *Science*, 2022.
[^watson2023]: `paper` Watson, J. L. et al. "De novo design of protein structure and function with RFdiffusion." *Nature*, 2023.
[^abramson2024]: `paper` Abramson, J. et al. "Accurate structure prediction of biomolecular interactions with AlphaFold 3." *Nature*, 2024. {Published without a code release, which drew objections about reproducibility; access for academic use followed later that year.}
[^walls2020]: `paper` Walls, A. C. et al. "Elicitation of potent neutralizing antibody responses by designed protein nanoparticle vaccines for SARS-CoV-2." *Cell*, 2020. {The designed component is the nanoparticle scaffold; the antigen it displays is a natural viral domain.}
[^silva2019]: `paper` Silva, D.-A. et al. "De novo design of potent and selective mimics of IL-2 and IL-15." *Nature*, 2019. {The paper reports the designed molecule itself, not the clinical candidate derived from it, which its sponsor later discontinued.}
[^buel2022]: `paper` Buel, G. R. and Walters, K. J. "Can AlphaFold2 predict the impact of missense mutations on structure?" *Nature Structural and Molecular Biology*, 2022.
[^wittmann2025]: `paper` Wittmann, B. J. et al. "Strengthening nucleic acid biosecurity screening against generative protein design tools." *Science*, 2025.
[^nobel2024]: `statement` Royal Swedish Academy of Sciences. "The Nobel Prize in Chemistry 2024." Press release, 2024.

==============================================================================
ARTICLE: alcor
TITLE:   Alcor Life Extension Foundation
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/alcor
SOURCE:  https://futurehumanwiki.com/raw/alcor
==============================================================================

---
title: "Alcor Life Extension Foundation"
slug: "alcor"
type: "organization"
status: "established"
horizon: "indefinite"
categories: ["organizations", "longevity"]
tags: ["cryonics", "cryopreservation", "nonprofit", "death", "trusts", "vitrification"]
summary: "A United States cryonics organization, founded in 1972, that preserves legally dead members at cryogenic temperatures under a trust structure meant to fund storage indefinitely."
updated: "2026-07-27"
issues: ["Patient and member counts are given as approximations with no source.", "The 1979 collapse of a Californian cryonics competitor is described without a source."]
---

```infobox
{
  "caption": "Cryonics nonprofit",
  "rows": [
    { "label": "Founded", "value": "1972, California" },
    { "label": "Founders", "value": "Fred and Linda Chamberlain" },
    { "label": "Headquarters", "value": "Scottsdale, Arizona" },
    { "label": "Patients", "value": "Roughly 250" },
    { "label": "Members", "value": "Over 1,000" },
    { "label": "Minimum funding", "value": "$200,000 whole body; $80,000 neuro" },
    { "label": "Endowment", "value": "Patient Care Trust" }
  ]
}
```

**Alcor Life Extension Foundation** is a nonprofit organization in Scottsdale, Arizona that performs and maintains human [[cryonics|cryopreservation]]. Founded in California in 1972, it is among the largest such organizations by number of patients in storage and has done more than most to formalize the practice — standby teams, vitrification protocols, and a segregated trust intended to fund liquid nitrogen indefinitely. It makes no claim that revival is currently possible, and its own literature describes the arrangement as a bet.

```keyfacts
[
  { "value": "1972", "label": "Founded", "note": "as the Alcor Society for Solid State Hypothermia" },
  { "value": "~250", "label": "Patients in storage", "note": "whole-body and neuropreservation combined" },
  { "value": "$200k / $80k", "label": "Minimum funding", "note": "whole body and neuro respectively, usually paid by life insurance" }
]
```

## Overview

Alcor's function is administrative as much as technical. Preservation must begin immediately after a physician pronounces death, which requires knowing when a member is dying, having people and equipment in place, and having a legal instrument that survives family objection. The organization's procedures are built around that timing problem, and its documented failures have generally been failures of logistics rather than of cryobiology.

The rule cuts the other way as well: nothing may be done before pronouncement, which guarantees an interval of warm ischemia and rules out preserving a member whose brain is being destroyed by a progressive disease while they remain legally alive. That constraint sits awkwardly beside [[right-to-die-and-right-to-live|assisted-dying law]] and has been challenged unsuccessfully in court.

Members sign a contract and fund it, most commonly by naming Alcor as beneficiary of a life insurance policy, which converts a large capital cost into a manageable premium and makes the arrangement cheap for the already insurable — one narrow instance of the [[access-and-inequality]] problem that runs through life-extension technology. On death, a standby team stabilizes the body, perfusion with cryoprotectant is performed at the Scottsdale facility, and the patient is cooled to liquid nitrogen temperature and stored in a dewar. Two options exist: whole-body preservation, and neuropreservation of the head alone, on the reasoning that any technology capable of revival could also supply a body. That reasoning leans on [[lab-grown-organs]] and on [[whole-brain-emulation]] as the alternative routes to a functioning person.

## History

Fred and Linda Chamberlain founded the organization in 1972 as the Alcor Society for Solid State Hypothermia, taking the name from a star in Ursa Major. It performed its first cryopreservation in 1976. The organization relocated from California to Scottsdale in 1994, partly for seismic and regulatory reasons.

Alcor also holds James Bedford, the first person ever cryopreserved, in 1967. His body was maintained privately by his family for decades and transferred to Alcor in 1991, where an inspection found the preservation had held despite decades of unstandardized storage.

```timeline
[
  { "year": "1972", "title": "Founded", "text": "Fred and Linda Chamberlain incorporate the Alcor Society for Solid State Hypothermia in California." },
  { "year": "1987–1990", "title": "The Dora Kent case", "text": "A coroner rules the death of a neuropreserved member a homicide; Alcor's facility is raided and staff are detained, but no charges result and the organization prevails in subsequent litigation." },
  { "year": "1991", "title": "Bedford transferred", "text": "The body of James Bedford, cryopreserved in 1967, is moved to Alcor and inspected." },
  { "year": "1994", "title": "Move to Arizona", "text": "Operations relocate to Scottsdale, where the organization remains." },
  { "year": "c. 2001", "title": "Vitrification adopted", "text": "Alcor switches from freezing to vitrification using low-toxicity cryoprotectant solutions, displacing ice formation as the dominant damage mechanism in well-perfused tissue." },
  { "year": "2002", "title": "The Ted Williams case", "text": "The baseball player's cryopreservation becomes a public dispute among his heirs and generates lasting negative coverage." },
  { "year": "2011–2020", "title": "The More years", "text": "The philosopher Max More serves as chief executive, professionalizing operations and raising the organization's public profile." }
]
```

## Procedures

Alcor's operational sequence has three phases, and the first is where most of the variance lies.

**Standby and stabilization.** For a member dying in a hospital or hospice with warning, a team deploys in advance. After pronouncement, circulation is restored mechanically, cooling begins with ice, and drugs intended to limit reperfusion injury and clotting are administered. For an unwitnessed death, none of this happens, and warm ischemia proceeds for hours or days. The difference between the best and worst case is the single largest determinant of preservation quality, and Alcor's own case reports say so.

**Cryoprotective perfusion.** Blood is replaced by a vitrification solution, ramped up in concentration over hours. Alcor uses solutions of the class developed at 21st Century Medicine, engineered to suppress ice at concentrations that are less toxic than earlier formulations. Perfusion quality is assessed by measures including cryoprotectant concentration in venous return and, for neuro cases, by observation of the brain surface.

**Cooldown and storage.** Patients are cooled below the glass transition and then to −196 °C and stored inverted in dewars of liquid nitrogen, so that a failure of resupply exposes the feet before the head. Cooling below the glass transition produces thermal stress and fracturing, a documented and unresolved problem; intermediate-temperature storage has been discussed as a remedy and is more complex to maintain.

> [!caution] What Alcor does not claim
> The organization does not assert that its patients can be revived, that current methods preserve a person intact, or that any timeline exists. Its published position is that preservation is worth attempting because the alternative is certain, and because information may survive that present medicine cannot use. Cryobiology societies do not accept even this weaker claim as established.

## Controversies

**Dora Kent (1987).** Kent, the mother of an Alcor member, was neuropreserved shortly after her death at the Alcor facility. The Riverside County coroner concluded that barbiturates had been administered before death and issued a homicide finding; the facility was raided, staff were detained, and the organization fought the case for years. No charges were ultimately brought, and courts found substantially in Alcor's favour. The episode established, in practice, that the organization would litigate.

**Ted Williams (2002).** The cryopreservation of the baseball player, contested between his children, generated intense press coverage and a subsequent book by a former Alcor employee alleging mishandling of remains. Alcor disputed the account. Whatever the merits, the case fixed a public image of the organization that it has not escaped.

**Institutional longevity.** The most serious criticism is structural rather than scandalous. Patients must be maintained for an unknown period by an organization that receives no further revenue from them. Alcor's answer is the Patient Care Trust, a segregated fund invested to cover maintenance from returns, with the operating organization legally distinct from it. The 1979 collapse of the Cryonics Society of California, in which stored bodies were found thawed after the organization ran out of money, is the field's worked example of what the structure is meant to prevent. Whether any private institution can hold that arrangement for two centuries is untested and untestable in advance.

## Membership and finances

Alcor has over a thousand members and roughly 250 patients in storage as of the mid-2020s, with growth that has been slow and approximately linear for decades. Minimum funding is $200,000 for whole-body and $80,000 for neuropreservation, with a portion allocated to the Patient Care Trust and the remainder to the cost of the procedure and standby.

The membership is not representative. It skews technical, male, American, and drawn disproportionately from the [[transhumanism|transhumanist]] and software communities — the same population that produced [[extropianism]], with which Alcor's leadership has overlapped. Notable patients include the futurist [[fm-2030]] and the cryptographer Hal Finney. [[max-more]], who wrote the extropian principles, led the organization from 2011 to 2020.

## Reception

The Society for Cryobiology has distanced itself from cryonics for decades and at one point barred practitioners from membership. Its objection is not that low-temperature storage fails to arrest decay, which is uncontroversial physics, but that the ischemic and toxic injury sustained before and during preservation is unquantified, that no mammal has been revived from cryogenic temperature, and that describing the practice as medicine misleads the public.

Supporters make a narrower argument. The relevant threshold is information-theoretic death, formalized by Ralph Merkle: whether the structures encoding memory and personality remain inferable, not whether tissue remains alive.[^merkle1994] The routes from there — cell-by-cell repair by hypothetical [[medical-nanorobots]], or scanning the [[brain-preservation|preserved brain]] and running it as an emulation on the [[substrate-independence]] premise — and the objections to each are set out under [[cryonics]]. Alcor's own literature endorses neither route and no timeline; critics respond that an unfalsifiable bet with an unbounded payoff is a poor guide to action, and that the resemblance to Pascal's wager is not accidental.

## Outlook

The developments that would most change Alcor's position are happening outside it. Vitrification and rewarming of whole mammalian organs, driven by the [[organ-shortage]] and organ-banking field rather than by cryonics, has produced a vitrified rabbit kidney that supported an animal after transplantation[^fahy2009] and, more recently, rat kidneys rewarmed by inductive heating of infused magnetic nanoparticles and transplanted with restored function.[^han2023] Scaling that to larger organs would supply the first direct evidence that a large vitrified mammalian structure can be recovered. Separately, progress in [[connectomics]] toward reconstructing circuitry from preserved tissue bears on whether a preserved brain contains what the practice assumes it contains.

Neither line of work will answer the question Alcor exists to answer, which is not technical. The organization is a wager that an institution can outlast the problem it was created to survive, and the only test of that is time.

## See also

- [[cryonics]]
- [[brain-preservation]]
- [[max-more]]
- [[fm-2030]]
- [[whole-brain-emulation]]
- [[personal-identity-and-continuity]]
- [[mind-uploading]]
- [[extropianism]]

## References

[^merkle1994]: `paper` Merkle, R.C. "The Molecular Repair of the Brain." *Cryonics*, 1994. {Cryonics is the magazine published by the organization this article describes, so the argument appears in a house venue rather than a reviewed one.}
[^fahy2009]: `paper` Fahy, G.M., Wowk, B., Pagotan, R. et al. "Physical and biological aspects of renal vitrification." *Organogenesis*, 2009.
[^han2023]: `paper` Han, Z. et al. "Vitrification and nanowarming enable long-term organ cryopreservation and life-sustaining kidney transplantation in a rat model." *Nature Communications*, 2023. {The organ was a rat kidney rewarmed by heating infused magnetic nanoparticles; nothing of human-organ scale has been recovered by this route.}

==============================================================================
ARTICLE: altos-labs
TITLE:   Altos Labs
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/altos-labs
SOURCE:  https://futurehumanwiki.com/raw/altos-labs
==============================================================================

---
title: "Altos Labs"
slug: "altos-labs"
type: "organization"
status: "emerging"
horizon: "2030s"
categories: ["organizations", "longevity"]
tags: ["reprogramming", "rejuvenation", "biotech", "aging", "venture funding", "cell biology"]
summary: "A biotechnology company launched in 2022 with about three billion dollars in committed funding to develop cellular rejuvenation programming as a medicine."
updated: "2026-07-27"
issues: ["The $3 billion figure and the investor names rest on a single 2021 news report.", "The research programme section lists strands without citing the company's own publications."]
---

```infobox
{
  "caption": "Private biotechnology company",
  "rows": [
    { "label": "Launched", "value": "January 2022" },
    { "label": "Reported funding", "value": "About $3 billion" },
    { "label": "Chief executive", "value": "Hal Barron" },
    { "label": "Chief scientist", "value": "Rick Klausner" },
    { "label": "Sites", "value": "Bay Area, San Diego, Cambridge UK" },
    { "label": "Focus", "value": "Cellular rejuvenation programming", "link": "/wiki/partial-reprogramming" },
    { "label": "Clinical programmes", "value": "None disclosed as of 2026" }
  ]
}
```

**Altos Labs** is a privately held biotechnology company, launched publicly in January 2022 with roughly three billion dollars in committed capital, that aims to turn [[epigenetic-reprogramming|cellular reprogramming]] into a treatment for age-related disease. It is among the largest single financial commitments made to rejuvenation biology, and as of 2026 it has published research but has not publicly named a clinical candidate. Its existence is the clearest evidence that reprogramming moved from an academic curiosity to an industrial programme; whether it can be made into a drug remains open.

```keyfacts
[
  { "value": "$3bn", "label": "Reported launch capital", "note": "committed at founding, an unusually large sum for a preclinical company" },
  { "value": "2022", "label": "Public launch", "note": "after roughly a year of confidential recruiting" },
  { "value": "None", "label": "Clinical candidates named", "note": "none publicly disclosed as of mid-2026" }
]
```

## Overview

The company's stated goal is "cellular rejuvenation programming": restoring the function of aged or damaged cells by resetting their epigenetic state, without converting them into a different cell type and without producing tumours. Its scientific premise is that aging involves a substantial component of reversible epigenetic change rather than only irreversible damage, and that the [[yamanaka-factors]] provide a handle on it.

Altos was assembled unusually. Rather than licensing a technology and building a pipeline, it recruited senior academic laboratories more or less intact, paid salaries reported to be several times academic norms, and gave principal investigators freedom to publish. The structure resembles a privately funded research institute more than a conventional startup, and the company has said this is deliberate: the underlying biology is not considered ready for a development programme.

## History

The company was incorporated quietly in 2021 and reported before its public launch by MIT Technology Review, which described a recruitment campaign among leading aging and reprogramming researchers and named Jeff Bezos and Yuri Milner among the backers.[^regalado2021] It launched formally in January 2022 with Hal Barron, previously chief scientific officer at GSK and before that at Genentech, as chief executive; Rick Klausner, a former director of the US National Cancer Institute, as chief scientist; and Hans Bishop, a founder of Juno Therapeutics, as president.

Juan Carlos Izpisua Belmonte, whose Salk Institute laboratory produced the founding demonstration of [[partial-reprogramming]] in living mice, moved to Altos with much of his group. Other recruits included Manuel Serrano from Barcelona, Wolf Reik from the Babraham Institute in Cambridge, and [[steve-horvath|Steve Horvath]], whose work established the [[epigenetic-clock]] as a measurement tool. [[shinya-yamanaka]] agreed to serve as a senior scientific adviser while remaining at Kyoto University.

```timeline
[
  { "year": "2006", "title": "The enabling discovery", "text": "Takahashi and Yamanaka show that four transcription factors can return an adult cell to a pluripotent state." },
  { "year": "2016", "title": "Partial reprogramming in vivo", "text": "Izpisua Belmonte's Salk group reports that cyclic, short-duration expression of the factors extends lifespan in a progeria mouse model and improves tissue repair after injury in normal mice." },
  { "year": "2021", "title": "Quiet assembly", "text": "Reports emerge of a well-funded new company recruiting senior reprogramming and aging researchers under confidentiality." },
  { "year": "2022", "title": "Public launch", "text": "Altos Labs announces itself with about $3 billion committed, three research institutes, and a leadership drawn from large-pharma R&D." },
  { "year": "2022–2026", "title": "Publication without clinic", "text": "The company's laboratories publish on reprogramming mechanism, cell identity and epigenetic ageing; no clinical programme is announced." }
]
```

## Research programme

The technical problem Altos works on is narrow and well defined. The four factors identified by Takahashi and Yamanaka return an adult cell to pluripotency when expressed continuously,[^takahashi2006] which in a living animal erases cell identity and causes teratomas. Transient or cyclic expression appears instead to restore youthful gene-expression patterns while leaving identity intact, a result first shown in vivo by Ocampo and colleagues in 2016.[^ocampo2016] A separate line of work from [[david-sinclair]]'s laboratory reported that expressing three of the factors in retinal ganglion cells restored vision in mice after optic-nerve crush and in aged animals, and argued that the recovered information is epigenetic rather than genetic.[^lu2020] The company's work concerns what determines where on that spectrum a cell lands, how to deliver the intervention to a chosen tissue, and how to measure the result in something more meaningful than a clock reading.

Several strands are visible in its published output and public statements:

- **Mechanism of rejuvenation.** Distinguishing the epigenetic changes that carry functional age from those that merely correlate with it, which is the same problem that makes [[biological-age]] measurement contested.
- **Factor and dose engineering.** Alternatives to the canonical four factors, including OSK without Myc and non-Yamanaka transcription-factor sets, and control of expression duration.
- **Delivery.** Reaching a specific tissue in an adult animal, the bottleneck that also constrains [[somatic-gene-therapy]] generally, using [[aav-vectors]] and [[lipid-nanoparticles]] among other approaches.
- **Readouts.** Functional endpoints such as regenerative capacity after injury, rather than [[aging-biomarkers|biomarker]] shifts alone.

> [!caution] What has been shown, and in what
> Partial reprogramming has extended lifespan in a progeria mouse model, improved tissue repair after injury in normal mice, and restored vision after optic-nerve injury in mice. None of this has been shown in a human. Reprogramming has never been given to a person for an aging indication; the first-in-human intentions companies have stated so far concern eye disease rather than aging itself, and none has published clinical data.

## Funding and structure

The three-billion-dollar figure refers to capital committed at launch rather than money spent, and Altos has not published financial statements. Reported backers include Bezos and Milner; the company has not confirmed the full investor list. It operates institutes in the San Francisco Bay Area, San Diego, and Cambridge in the United Kingdom, with a stated intention to build further sites.

The scale matters for a structural reason. Rejuvenation research has historically been funded either by government grants, which are allocated to hypothesis-driven projects on short cycles, or by philanthropic bodies such as the [[sens-research-foundation]] and [[methuselah-foundation]], which had budgets several orders of magnitude smaller. Altos can run a decade of preclinical work without a product, which is a genuinely different experiment in how the field might progress. It is also a concentration of talent that removes senior investigators from university training pipelines, a cost the [[buck-institute]] and other academic centres have noted.

## Reception

Scientific reception has been guarded rather than hostile. The underlying result is real and reproducible, and few biogerontologists dispute that partial reprogramming does something to aged cells. The disputes are about interpretation and translation.

The interpretive question is what "rejuvenation" means when its main evidence is a shift in DNA methylation patterns. A change in a methylation clock reading is a change in a correlate; it is not by itself evidence that an organism is functionally younger, and the field has no surrogate endpoint that a regulator would accept. This is the same gap that constrains the whole [[geroscience-hypothesis]] programme.

The translational question is safety. The factors that rejuvenate are the factors that cause cancer; Myc is an oncogene, and dedifferentiated cells are the raw material of tumours. A therapy given to healthy older people for prevention faces a far higher safety bar than one given to patients with a fatal disease, and no dosing regime has been shown to be safe in a large mammal over years.

A third criticism concerns the company's silence. Altos has published papers but has released little about its pipeline, timelines, or decision criteria, which makes external assessment of progress impossible. Commentators have drawn the comparison with [[calico]], which was launched in 2013 with comparable ambition and produced a modest public output relative to its resources over its first decade.

## Outlook

The informative signals over the next several years will be specific. Does Altos publish a functional rejuvenation result in a large, long-lived mammal rather than a mouse? Does it name a first indication, and is that indication an acute tissue injury — where a short course of reprogramming has a plausible risk-benefit case — rather than aging itself? Does it disclose a delivery method that reaches a solid organ at a therapeutic dose without dosing the whole body?

If reprogramming turns out to be a treatment for particular injuries rather than a general intervention against aging, that would still be a substantial result and a poor return on three billion dollars aimed at the latter. The company's founding wager is that the two are the same problem at different doses, and nothing yet published settles it.

## See also

- [[epigenetic-reprogramming]]
- [[partial-reprogramming]]
- [[yamanaka-factors]]
- [[newlimit]]
- [[calico]]
- [[retro-biosciences]]
- [[epigenetic-clock]]
- [[shinya-yamanaka]]

## References

[^regalado2021]: `news` Regalado, A. "Meet Altos Labs, Silicon Valley's latest wild bet on living forever." *MIT Technology Review*, September 2021. {Published before the company announced itself; the funding figure and investor names come from the reporting, not from Altos.}
[^ocampo2016]: `paper` Ocampo, A. et al. "In Vivo Amelioration of Age-Associated Hallmarks by Partial Reprogramming." *Cell*, 2016. {The lifespan extension was in mice carrying a progeria mutation; effects reported in normally aging mice were smaller and not lifespan.}
[^takahashi2006]: `paper` Takahashi, K. and Yamanaka, S. "Induction of Pluripotent Stem Cells from Mouse Embryonic and Adult Fibroblast Cultures by Defined Factors." *Cell*, 2006.
[^lu2020]: `paper` Lu, Y. et al. "Reprogramming to recover youthful epigenetic information and restore vision." *Nature*, 2020.

==============================================================================
ARTICLE: anders-sandberg
TITLE:   Anders Sandberg
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/anders-sandberg
SOURCE:  https://futurehumanwiki.com/raw/anders-sandberg
==============================================================================

---
title: "Anders Sandberg"
slug: "anders-sandberg"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "minds"]
tags: ["whole brain emulation", "existential risk", "forecasting", "computational neuroscience", "transhumanism"]
summary: "Swedish computational neuroscientist and futures researcher who co-wrote the whole brain emulation roadmap and works on the physical limits of long-term civilization."
updated: "2026-07-27"
issues: ["The anthropic shadow and Converging Cognitive Enhancements papers are discussed without citations."]
---

```infobox
{
  "caption": "Computational neuroscientist and futures researcher",
  "rows": [
    { "label": "Born", "value": "1972, Sweden" },
    { "label": "Nationality", "value": "Swedish" },
    { "label": "Education", "value": "PhD, Stockholm University, 2002" },
    { "label": "Known for", "value": "Whole Brain Emulation: A Roadmap", "link": "/wiki/whole-brain-emulation" },
    { "label": "Field", "value": "Computational neuroscience; futures studies" },
    { "label": "Affiliation", "value": "Institute for Futures Studies, Stockholm" },
    { "label": "Former", "value": "Future of Humanity Institute, Oxford" }
  ]
}
```

**Anders Sandberg** is a Swedish computational neuroscientist and futures researcher whose work
consists largely of putting numbers on questions that are usually discussed without them: how much
compute a brain emulation would need, how long it would take to colonize a galaxy, how much
confidence a risk estimate can carry when the model itself might be wrong. He co-authored the
standard technical roadmap for [[whole-brain-emulation]] and spent nearly two decades at Oxford's
Future of Humanity Institute.

## Career

Sandberg's doctorate, completed at Stockholm University in 2002, was on attractor neural network
models of memory — the mathematics of how a network of units settles into stable states that behave
like stored patterns. That background explains his position on emulation: he approaches the question
as a modeller who has built neural systems, not as a philosopher arguing from the outside.

He chaired the Swedish transhumanist association during the 1990s and co-founded a Stockholm think
tank before joining the Future of Humanity Institute in 2006, where he worked with
[[nick-bostrom]], Toby Ord, [[eric-drexler]] and others until the institute closed in 2024. He
subsequently moved to the Institute for Futures Studies in Stockholm. He is a prolific public
communicator, and much of his output takes the form of estimates published as blog posts, technical
reports and talks rather than as conventional papers.

## Whole brain emulation

The 2008 report written with Bostrom is the reference document for the emulation
question.[^roadmap2008] Its method is to refuse to argue about whether emulation is possible and
instead ask what would have to be true at each level of description, then estimate the requirements
under each assumption.

Its premise, stated as a premise rather than defended, is [[substrate-independence]]: that what
matters about a brain is the organization of its causal structure rather than the material
implementing it.

The roadmap defines a ladder of models: from a coarse brain-region simulation, through spiking
neural networks, compartment models with realistic dendrites, molecular-level channel dynamics, and
down to whole-molecule and quantum descriptions. Each rung carries an estimate of storage, compute
and scanning resolution. The estimates span many orders of magnitude, and the report is explicit
that the correct rung is unknown — this is the central finding, not a hedge. It also identifies
scanning throughput rather than compute as the likely binding constraint, a judgement that
[[connectomics]] has broadly borne out: reconstructing a cubic millimetre of cortex remains a
multi-year effort for a large collaboration. The same analysis makes structural
[[brain-preservation]] a coherent research target, since a preserved connectome is a scannable
object even if nobody can yet scan it at scale.

> [!key] What the roadmap actually establishes
> Not that emulation will work. It establishes that the requirements are finite and estimable *given
> a level of description*, and that the disagreement between optimists and sceptics is almost
> entirely about which level suffices. That reframing is the report's contribution.

Sandberg has separately written on the ethics of emulation — the welfare of a running emulation, the
implications of copying, and the research ethics of experiments on emulated brains, which is
[[machine-consciousness]] with the moral stakes made unavoidable. He is consistently more cautious
in public than the [[mind-uploading]] discourse that cites him.

## Quantifying the long term

A recurring pattern in Sandberg's work is taking a speculative claim and computing its physical
requirements. With Stuart Armstrong he calculated the energy and engineering cost of launching
self-replicating probes to every reachable galaxy, concluding that intergalactic settlement is
within the physical means of a civilization not much more capable than a plausible near-future
one — which sharpens rather than resolves the Fermi paradox.[^eternity2013] With Drexler and Ord he
argued that the paradox partly dissolves under correct handling of uncertainty: multiplying point
estimates in the Drake equation discards the distribution, and propagating the actual uncertainty
places substantial probability mass on humanity being alone.[^dissolving2018]

He has written on the thermodynamic limits of computation, on "Jupiter brains" as the physical
ceiling for a single computational structure, and on the aestivation hypothesis — the proposal that
an advanced civilization might wait for the cosmic background to cool before computing, because
computation is cheaper at low temperature. With Ord and others he formulated the argument that
model uncertainty dominates very low probability estimates: a calculation giving a one-in-a-billion
chance of catastrophe is worthless if the chance the calculation is wrong is one in a
thousand.[^probing2010]

## Enhancement and risk

Sandberg's enhancement work is mostly with Bostrom. "Converging Cognitive Enhancements" surveys the
routes — pharmacological, genetic, external, and interface-based — and argues that they interact
rather than compete. The "wisdom of nature" heuristic asks why evolution has not already made a
proposed improvement, and treats a satisfactory answer as a precondition for expecting the
improvement to be free of hidden costs; the usual acceptable answers are changed environments,
evolutionary constraints, and value discordance between fitness and welfare. It is a rare piece of
pro-enhancement writing that supplies a filter its own conclusions must pass. See
[[human-enhancement]] and [[nootropics]].

On risk, the "anthropic shadow" argument he co-wrote holds that the historical record systematically
understates the frequency of catastrophes severe enough to have prevented observers from existing —
so the empirical base rate for some [[existential-risk|extinction-level events]] is biased downward
by the fact that anyone is around to compute it.

## Reception and legacy

Sandberg is cited across futures studies, AI safety and philosophy, and is unusual in retaining
credibility with mainstream scientists while working on subjects they generally avoid. The reason is
methodological: he states assumptions, publishes the arithmetic, and marks the gap between what
follows from physics and what follows from speculation. Criticism concentrates on the limits of that
method — that computing the physical requirements of a scenario says nothing about its biological or
social plausibility, and that the roadmap's neutral framing gives emulation more credibility than
neuroscience currently warrants.

His long-announced book on grand futures, covering the physical possibilities open to a
civilization over cosmological time, has circulated for years mainly as drafts and lecture
material, and is easier to encounter in talks than in print. Its subject is the same one this
wiki's [[future-of-humanity]] article takes up, approached from the side of physical limits rather
than of technology. And its central
question is the one his quantitative work keeps producing and cannot answer: the physics permits
enormously more than humanity currently does, so the binding constraint is not physical, and
nothing in the calculations says what it is instead.

## See also

- [[whole-brain-emulation]]
- [[mind-uploading]]
- [[connectomics]]
- [[existential-risk]]
- [[nick-bostrom]]
- [[eric-drexler]]
- [[substrate-independence]]
- [[future-of-humanity]]

## References

[^roadmap2008]: `report` Sandberg, A. and Bostrom, N. *Whole Brain Emulation: A Roadmap*. Future of Humanity Institute Technical Report 2008-3, University of Oxford, 2008. {An institute technical report rather than a reviewed paper; every resource estimate in it is conditional on which level of brain description turns out to suffice.}
[^eternity2013]: `paper` Armstrong, S. and Sandberg, A. "Eternity in six hours: Intergalactic spreading of intelligent life and sharpening the Fermi paradox." *Acta Astronautica*, 2013.
[^dissolving2018]: `preprint` Sandberg, A., Drexler, E. and Ord, T. "Dissolving the Fermi Paradox." Preprint, 2018. {Widely cited but not peer reviewed; the argument is about how uncertainty in the Drake parameters is propagated, and rests on no new observation.}
[^probing2010]: `paper` Ord, T., Hillerbrand, R. and Sandberg, A. "Probing the improbable: methodological challenges for risks with low probabilities and high stakes." *Journal of Risk Research*, 2010.

==============================================================================
ARTICLE: artificial-blood
TITLE:   Artificial blood
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/artificial-blood
SOURCE:  https://futurehumanwiki.com/raw/artificial-blood
==============================================================================

---
title: "Artificial blood"
slug: "artificial-blood"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 5
categories: ["bodies", "nanomedicine"]
tags: ["transfusion", "oxygen carriers", "haemoglobin", "trauma", "cell culture", "clinical trials"]
summary: "Engineered substitutes for the oxygen-carrying function of blood, including haemoglobin carriers, perfluorocarbon emulsions and cultured red cells, none in general clinical use."
updated: "2026-07-27"
humanEvidence: "Trials of cell-free haemoglobin carriers in people showed excess mortality and myocardial infarction across sixteen studies; the human data on cultured red cells are small tagged-volume transfusions in volunteers."
access: "No product is licensed in the United States for general transfusion; a polymerized bovine carrier is licensed in South Africa and available in the US only under expanded access, and cultured red cells exist only inside trials."
reversibility: "reversible"
issues: ["Cultured red cells are said to cost many times a donated unit, with no sourced figure.", "No development-history timeline block, unlike comparable technology articles here."]
---

```infobox
{
  "caption": "Transfusion medicine technology",
  "rows": [
    { "label": "Function replaced", "value": "Oxygen transport only" },
    { "label": "Main classes", "value": "HBOCs, perfluorocarbons, cultured cells" },
    { "label": "First product approved", "value": "Fluosol, 1989, later withdrawn" },
    { "label": "Currently licensed in US", "value": "None for general transfusion" },
    { "label": "Key HBOC failure mode", "value": "Nitric oxide scavenging" },
    { "label": "Cells per unit of blood", "value": "About 2 trillion" },
    { "label": "Status", "value": "Trials and expanded access" }
  ]
}
```

**Artificial blood** refers to manufactured substitutes for the oxygen-carrying function of red cells. The term is misleading in a way worth stating at the outset: no product replaces blood. Blood carries oxygen and carbon dioxide, clots, transports immune cells, buffers pH, distributes hormones and regulates temperature. Every substitute developed so far addresses only oxygen transport, and is more accurately called an oxygen therapeutic. Four decades of development have produced one product approved and withdrawn in the United States, several approved in other jurisdictions or for veterinary use, a large trial record of harm, and a recent shift toward growing real red cells in culture instead.

## What a substitute has to do

Donated red cells set a demanding benchmark. They circulate for months, release oxygen in response to local acidity and carbon dioxide, do not leak their contents into tissue, and are cleared by an orderly process when worn out. A substitute must match enough of that to be useful without triggering the vascular and oxidative injury that free haemoglobin causes, and it must do so at a price and shelf life that justify displacing a donation system that already works in wealthy countries. The one property no candidate has ever attempted is the rest of blood's job: clotting, immunity and the transport of everything that is not oxygen. Whole blood, and the [[organ-shortage]]-like logistics of collecting it, remain necessary regardless.

## Haemoglobin-based oxygen carriers

The obvious approach is to use haemoglobin without the cell around it. Cell-free haemoglobin binds oxygen, needs no blood typing, can be sterilized, and can be stored at room temperature for a year or more — the properties that make it attractive for trauma care, military medicine and settings without a cold chain.

Raw haemoglobin outside a red cell behaves badly. The tetramer dissociates into dimers small enough to be filtered by the kidney, causing renal injury. It is not protected by the red cell's reducing enzymes, so it oxidizes to methaemoglobin and releases free haem and iron, which drive oxidative tissue damage. Its oxygen affinity is wrong without the 2,3-diphosphoglycerate that red cells supply. And, most consequentially, it moves freely into the space between endothelial cells and there scavenges nitric oxide, the signalling molecule that keeps arterioles dilated.

Chemistry addressed the first problems and not the last. Cross-linking the tetramer prevented dissociation; polymerization with glutaraldehyde produced larger molecules with longer circulation; conjugation to polyethylene glycol increased size and reduced extravasation. Products developed on these principles included a diaspirin cross-linked human haemoglobin, a polymerized human haemoglobin, and a polymerized bovine haemoglobin. The bovine product remains licensed in South Africa and is available in the United States under expanded access for patients with severe anaemia who cannot receive a transfusion, most often for religious reasons; a veterinary version is approved for dogs.

### Why the trials failed

The clinical record is the reason haemoglobin-based carriers are not in routine use. A trauma trial of the diaspirin cross-linked product was stopped in the 1990s for excess mortality in the treatment arm. A polymerized human haemoglobin was tested in prehospital trauma under regulations permitting enrolment without consent, which generated its own controversy, and did not establish benefit; its manufacturer wound up.

The decisive analysis was a 2008 meta-analysis by Charles Natanson and colleagues covering sixteen trials of five different products across several clinical settings. It found an increase in mortality of roughly thirty percent and a near-tripling of myocardial infarction risk associated with these agents, and the effect was consistent across products, which pointed to a class mechanism rather than a manufacturing defect.[^natanson2008] The mechanism generally accepted is nitric oxide scavenging: an agent designed to deliver oxygen constricts the vessels that deliver it, raising blood pressure and reducing perfusion of the microcirculation, with the heart the most vulnerable organ.

> [!caution] A class effect, not a bad batch
> The consistency of the harm signal across chemically distinct products is what makes the haemoglobin approach hard to rescue. Any successor has to keep haemoglobin away from endothelial nitric oxide, which in practice means putting it back inside something — a vesicle, a polymer shell, a cell.

## Perfluorocarbons

Perfluorocarbons dissolve gases physically rather than binding them chemically. They carry roughly twenty times more oxygen per volume than water, release it readily, and can be manufactured synthetically with no biological source at all. Because they are immiscible with water they must be delivered as emulsions of submicrometre droplets.

Their limitations follow from the physics. Physical dissolution is linear in oxygen partial pressure, so a patient must breathe high concentrations of oxygen for the carrier to be useful, which restricts the setting. Emulsion droplets are cleared by macrophages and stored temporarily in liver and spleen, and the emulsifiers activate complement, causing flu-like reactions and transient reductions in platelet count.

A first-generation emulsion was approved in the United States in 1989 for use during balloon angioplasty and withdrawn a few years later after minimal clinical uptake.[^riess2001] A second-generation perflubron emulsion reached late-stage trials in cardiac surgery, where an excess of neurological events led to the programme being halted. Interest persists for specific niches, including oxygenation during organ preservation and as ultrasound contrast agents, rather than as a general transfusion substitute.

## Encapsulated haemoglobin and cultured red cells

The current field splits along the line the meta-analysis drew.

One branch puts haemoglobin back inside a container. Liposome-encapsulated haemoglobin vesicles, developed principally in Japan using haemoglobin purified from expired donor blood, are in early-phase human testing; the encapsulation is intended to prevent extravasation and nitric oxide scavenging while retaining long shelf life. Polymer-nanoparticle carriers with pH-sensitive oxygen release are in preclinical development under military funding, on the argument that the decisive use case is prehospital haemorrhage where no blood bank exists. These belong to the same family of engineered carriers described in [[targeted-drug-delivery]] and [[lipid-nanoparticles]]. They are also the closest existing objects to the artificial red cell proposed in [[respirocytes]], and the comparison is instructive: the proposed device assumes pressurised diamondoid vessels and onboard control systems of a kind that [[medical-nanorobots]] as a field has not built, while the real vesicles are liposomes filled with purified human protein and perform a fraction of the imagined function.

The other branch grows real red cells. Human haematopoietic stem cells from donated blood, or immortalized erythroid progenitor lines of the kind established at Bristol, can be differentiated in culture into enucleated red cells.[^trakarnsanga2017] The RESTORE trial in the United Kingdom, run by NHS Blood and Transplant with university partners, performed the first in-human transfusion of laboratory-grown red cells in 2022, giving volunteers small volumes of tagged cultured cells alongside standard donor cells to compare survival in circulation. Cultured cells are all newly made, whereas a donated unit contains cells of every age, so a cultured transfusion may last longer and require less frequent transfusion — the trial's actual hypothesis, and one that matters most for patients with thalassaemia or sickle cell disease who are transfused for life and accumulate alloantibodies. See [[casgevy]] for the curative alternative in those diseases.

The obstacle is arithmetic. A single unit of blood contains about two trillion red cells. Culturing that many, reproducibly, at a cost within reach of a health service, is a manufacturing problem several orders of magnitude beyond current bioreactor capacity, and cultured units are presently many times more expensive than donated ones. It is the same scale-up wall that constrains [[tissue-engineering]] generally, without the additional requirement of building a structure. Deriving red cells from [[induced-pluripotent-stem-cells]] would remove the dependence on donor progenitors entirely, and has been demonstrated at laboratory scale with poor enucleation efficiency. The near-term application is therefore not general transfusion but small volumes of rare blood groups that are difficult to source from donors.

A parallel line of work attacks supply rather than substitution, using enzymes discovered in gut bacteria to strip the A and B antigens from donor red cells and convert them to universal group O.[^rahfeld2019] Efficiency has improved and the approach would ease matching rather than replace donation.

## Where a substitute would actually be used

The case for oxygen therapeutics rests less on replacing routine transfusion, which works well in developed health systems, than on situations where banked blood is unavailable: prehospital and battlefield haemorrhage, disaster response, regions without reliable cold chains, patients who refuse transfusion, and patients whose alloantibodies make compatible units hard to find. Long-duration spaceflight adds another, since blood cannot be stockpiled for years and marrow function is altered in microgravity, a problem noted in [[space-medicine]]; proposals to lower metabolic demand instead, discussed in [[human-hibernation]], attack the same constraint from the other side.

Oxygen carriers also have a role outside transfusion. Perfusing a donor organ with an oxygen-carrying solution during machine preservation extends viability, which links this technology to the supply strategies in [[lab-grown-organs]] and to the mechanical circuits of the [[artificial-heart]], where blood is likewise pumped across engineered surfaces.

Meanwhile the alternatives have improved. Whole blood has returned to prehospital trauma care, tranexamic acid reduces bleeding deaths, and restrictive transfusion thresholds reduced demand without harming outcomes. Any new product must beat these, and against the historical background of a class of agents that increased mortality, regulators are unlikely to accept anything short of a well-powered trial with a hard endpoint.

## See also

- [[artificial-heart]]
- [[organ-shortage]]
- [[lab-grown-organs]]
- [[casgevy]]
- [[targeted-drug-delivery]]
- [[space-medicine]]
- [[tissue-engineering]]
- [[respirocytes]]

## References

[^natanson2008]: `paper` Natanson, C., Kern, S. J., Lurie, P., Banks, S. M. and Wolfe, S. M. "Cell-free hemoglobin-based blood substitutes and risk of myocardial infarction and death: a meta-analysis." *JAMA*, 2008. {Pools five chemically distinct products across several clinical settings, which is what makes the harm read as a property of the class rather than of one formulation.}
[^riess2001]: `paper` Riess, J. G. "Oxygen carriers ('blood substitutes') — raison d'être, chemistry, and some physiology." *Chemical Reviews*, 2001.
[^trakarnsanga2017]: `paper` Trakarnsanga, K. et al. "An immortalized adult human erythroid line facilitates sustainable and scalable generation of functional red cells." *Nature Communications*, 2017. {Establishes a cell line that produces red cells in culture; it does not address making a transfusable unit at a cost a health service could meet.}
[^rahfeld2019]: `paper` Rahfeld, P. et al. "An enzymatic pathway in the human gut microbiome that converts A to universal O type blood." *Nature Microbiology*, 2019.

==============================================================================
ARTICLE: artificial-general-intelligence
TITLE:   Artificial general intelligence
PORTAL:  Foundational Concepts
URL:     https://futurehumanwiki.com/wiki/artificial-general-intelligence
SOURCE:  https://futurehumanwiki.com/raw/artificial-general-intelligence
==============================================================================

---
title: "Artificial general intelligence"
slug: "artificial-general-intelligence"
type: "concept"
status: "contested"
horizon: "indefinite"
categories: ["foundations", "minds"]
tags: ["artificial intelligence", "agi", "forecasting", "benchmarks", "machine learning", "risk"]
summary: "A hypothetical artificial system matching or exceeding competent human performance across essentially all cognitive tasks rather than in narrow domains."
updated: "2026-07-27"
issues: ["The 2024 ARC-AGI result and its harder 2025 successor set are described without a source.", "The EU AI Act compute thresholds are described without citing the instrument."]
---

```infobox
{
  "caption": "Concept in artificial intelligence",
  "rows": [
    { "label": "Field", "value": "Computer science, cognitive science" },
    { "label": "Term in use since", "value": "1997" },
    { "label": "Contrast class", "value": "Narrow AI" },
    { "label": "Agreed definition", "value": "None" },
    { "label": "Agreed test", "value": "None" },
    { "label": "Status", "value": "Not demonstrated; timelines disputed" }
  ]
}
```

**Artificial general intelligence** is a hypothetical artificial system whose competence spans essentially the full range of cognitive tasks humans perform, rather than being confined to a domain it was built for. The term functions mainly by contrast with *narrow* AI: a chess engine, a protein-structure predictor and a fraud-detection model are each superhuman within their scope and useless outside it. Whether any existing system has begun to cross that boundary is one of the most consequential open disputes in science, and it is unresolved partly because the boundary has never been defined in a way that would settle it.

```keyfacts
[
  { "value": "1997", "label": "First recorded use of the phrase", "note": "Mark Gubrud, in a paper on nanotechnology and international security" },
  { "value": "~6 months", "label": "Doubling time of frontier training compute", "note": "estimated over 2010–2022 by Sevilla et al." },
  { "value": "6", "label": "Performance levels in DeepMind's AGI framework", "note": "'no AI' through 'superhuman', crossed with an axis of generality" }
]
```

## Defining the target

There is no accepted definition. Legg and Hutter, surveying the problem, assembled more than seventy published definitions of intelligence and extracted a common core: intelligence measures an agent's ability to achieve goals in a wide range of environments.[^legg2007] That formulation is precise enough to be formalised and too abstract to adjudicate any actual system.

Working definitions split into three families. *Task-based* definitions specify jobs a system must do — the Turing test is the ancestor, and modern versions substitute economically valuable work or a benchmark suite. *Capability-based* definitions specify cognitive functions such as transfer learning, long-horizon planning, or acquiring new skills without retraining. *Autonomy-based* definitions require operation over extended periods without human scaffolding. [[nick-bostrom]] sidesteps the threshold in *Superintelligence*, defining superintelligence as intellect greatly exceeding human performance in virtually all domains and treating human-level capability as a waypoint.

A framework from Google DeepMind researchers replaces the binary with a grid, scoring systems on performance (from "emerging" through "competent", "expert", "virtuoso" and "superhuman") crossed with generality.[^morris2024] On that scheme, mid-2020s systems are plausibly "emerging AGI" and clearly superhuman narrow AI in several domains — more informative than either "AGI is here" or "AGI is not here".

> [!note] Terminology
> "AGI" is used inconsistently even within single organisations: human-level competence across tasks, full automation of remote work, or the ability to do novel science. These are different claims with different timelines, and arguments equivocate between them.

## Origins of the term

Alan Turing's 1950 paper posed machine thinking in behavioural terms and proposed the imitation game to sidestep definitional argument.[^turing1950] The 1956 Dartmouth proposal assumed generality as the default goal, expecting machines to use language, form abstractions and improve themselves within a summer.

The narrower ambitions of expert systems and, later, statistical machine learning displaced that goal for decades. The phrase "artificial general intelligence" appears in a 1997 paper by Mark Gubrud on the security implications of advanced technology, and was independently adopted in the early 2000s by Shane Legg, Ben Goertzel and Peter Voss, becoming established with a 2007 edited volume of that title. The phrase then distinguished a small community from a mainstream that regarded the goal as disreputable; several of the largest laboratories now state it as their objective.

## Measuring progress

Benchmark performance is the field's main evidence, and it has an unusual failure mode: benchmarks saturate faster than they can be built. Knowledge tests, grade-school mathematics sets and function-level coding tests that were discriminating in the early 2020s reached ceiling within a few years. Their harder successors — graduate-level science, research mathematics, adversarially collected expert questions, repository-scale software tasks — have compressed in turn.

Saturation is weaker evidence than it appears. Contamination is hard to exclude when training corpora are web-scale and benchmarks are published; benchmarks measure what is easy to score, biasing them toward short checkable answers; and Goodhart's law applies with force when the benchmark is also the optimisation target for the labs reporting the scores.

The most-cited contamination-resistant test is François Chollet's Abstraction and Reasoning Corpus, built around tasks requiring a novel rule to be inferred from a handful of examples.[^chollet2019] It resisted large language models for years. In late 2024 a preview of a model trained to spend far more computation at inference time reported private-set scores in the range of typical human performance, at costs orders of magnitude above ordinary use; a harder successor set was released in 2025 on which scores fell sharply again. The gap closes, then is redefined, and which of those moves is more informative remains unclear.

## The case for nearness

Three arguments carry most of the weight.

**Scaling.** Empirical relationships between model scale, data and loss have held over several orders of magnitude, and frontier training compute grew at roughly six-month doubling times through the 2010s and early 2020s.[^sevilla2022] If capability tracks loss and loss tracks compute, extrapolation implies continued rapid gains. The inferential structure is the one examined in [[accelerating-change]], and its best-known long-range form is [[ray-kurzweil]]'s curve-fitting. The addition of inference-time computation as a second scaling axis from 2024 onward reopened headroom that pretraining scaling alone appeared to be exhausting.

**Generality already observed.** A single model trained on next-token prediction can write code, translate, summarise clinical notes, prove competition mathematics and control a browser. That breadth was not designed in, and it is the strongest evidence against the older view that generality requires explicit architecture — the position Richard Sutton called the bitter lesson.

**Expert opinion has moved.** Large surveys of published AI researchers have shifted median estimates for human-level machine intelligence substantially earlier across successive editions, with the aggregate 50% point falling around mid-century in the most recent.[^grace2024] Such forecasts have a poor track record in both directions, but the direction of movement is itself data.

## The case against

**Jaggedness.** Capability is not a scalar. Systems that solve olympiad problems fail at tasks a child handles, and the failures are not obviously less "general" than the successes. Studies of professional use find a frontier uneven in ways users cannot predict — a different situation from a uniformly sub-human system approaching parity.[^dellacqua2023]

**Reliability rather than capability is the binding constraint.** A system correct 95% of the time is not 95% of the way to automating a task if errors are uncorrelated with confidence and expensive to detect. Long-horizon agentic work compounds this: per-step reliability must be very high for hundred-step tasks to succeed.

**Missing components.** Critics including Gary Marcus and Melanie Mitchell argue that continual learning, causal models, compositional generalisation and grounded understanding are absent rather than underdeveloped, and that benchmark progress measures interpolation over a vast training distribution.[^mitchell2021] Moravec's observation still holds that sensorimotor competence, not abstract reasoning, is the hard part, and robotics has advanced far more slowly than text models.

**The definition problem cuts both ways.** If no test would convince skeptics and no failure would convince proponents, the disagreement is not empirical, and several participants on both sides have said so.

## Overhang and discontinuity

The *capability overhang* argument holds that deployed capability lags latent capability. A trained model's abilities are elicited by prompting, scaffolding, tool access and fine-tuning, none of which requires new compute, and jumps in agentic performance from scaffolding alone support the claim's general shape. Its policy relevance is that it undermines the assumption of a smooth, observable approach to any threshold: if a large gap exists between what systems can do and what they are seen doing, warning time is shorter than deployment curves suggest. That premise underwrites both [[differential-technological-development|proposals to sequence safety work ahead of capability work]] and [[precautionary-principle|precautionary]] pre-deployment evaluation regimes.

The related *compute overhang* argument, prominent in [[existential-risk]] discussion, holds that pausing capability research accumulates hardware enabling a faster jump when it resumes. Both are structural arguments, and neither has been quantified convincingly.

## Why AGI timelines matter here

Almost every long-horizon claim on this wiki has an AI term in it. Rejuvenation biology is bottlenecked on a system with more interacting variables than experiments can resolve, and [[ai-protein-design|protein structure prediction]] has already changed what is tractable there; companies such as [[retro-biosciences]] have made model-assisted protein engineering an explicit programme. The limit of that transfer is visible in [[ai-drug-discovery]], where models now generate candidate molecules far faster than medicinal chemists did and no drug produced that way has been approved. [[connectomics]] and [[whole-brain-emulation]] are limited by segmentation and simulation, both compute-bound. [[neural-decoding]] improves with better sequence models. Conversely, [[dual-use-research]] risk rises if design tools lower the expertise barrier for pathogens.

The [[technological-singularity]] argument depends entirely on this article's subject: an intelligence explosion requires systems capable of improving themselves, which is a strictly stronger condition than AGI. [[intelligence-amplification]] and [[human-ai-merger]] describe the alternative in which capability grows through human–machine teams rather than autonomous systems, and [[mind-uploading]] would require both AGI-scale compute and neuroscience that does not exist. Whether such a system would be a subject as well as an agent is a separate question, treated in [[machine-consciousness]]; the [[substrate-independence]] premise it shares with most [[posthuman]] forecasting is assumed far more often than it is argued.

> [!caution] Forecasts are not findings
> No timeline for AGI is well supported. Survey medians, scaling extrapolations and insider statements are all weak evidence, and the field's forecasting record — including confident predictions of imminence in the 1960s and of impossibility in the 1990s — should discount any specific date, including recent ones.

## Outlook

Governance has moved faster than definition. The EU's AI Act imposes obligations on general-purpose models by capability thresholds measured in training compute, several states have established safety institutes with evaluation mandates, and an international expert report on advanced AI risk was published in 2025. Every one of these instruments must operationalise "general" and "capable" without a scientific definition of either, and they have done so with proxies — training FLOP, benchmark scores, red-team results — that are known to be poor. Whether a capability threshold can be specified precisely enough to regulate before anyone can specify it precisely enough to measure is the practical form of the definitional problem, and it now binds.

## See also

- [[technological-singularity]]
- [[machine-consciousness]]
- [[whole-brain-emulation]]
- [[human-ai-merger]]
- [[existential-risk]]
- [[differential-technological-development]]
- [[accelerating-change]]
- [[future-of-humanity]]

## References

[^legg2007]: `paper` Legg, S. and Hutter, M. "Universal Intelligence: A Definition of Machine Intelligence." *Minds and Machines*, 2007.
[^morris2024]: `paper` Morris, M. R. et al. "Levels of AGI for Operationalizing Progress on the Path to AGI." *Proceedings of the International Conference on Machine Learning*, 2024.
[^turing1950]: `paper` Turing, A. M. "Computing Machinery and Intelligence." *Mind*, 1950.
[^chollet2019]: `preprint` Chollet, F. "On the Measure of Intelligence." arXiv preprint, 2019.
[^sevilla2022]: `paper` Sevilla, J. et al. "Compute Trends Across Three Eras of Machine Learning." *International Joint Conference on Neural Networks*, 2022.
[^grace2024]: `preprint` Grace, K. et al. "Thousands of AI Authors on the Future of AI." arXiv preprint, 2024. {A survey of authors who published at major machine-learning venues; it records opinion, and successive editions of the series have moved sharply.}
[^dellacqua2023]: `preprint` Dell'Acqua, F. et al. "Navigating the Jagged Technological Frontier." Harvard Business School working paper, 2023. {A working paper reporting a field experiment with consultants at one firm, so the frontier it maps is specific to that task set.}
[^mitchell2021]: `preprint` Mitchell, M. "Why AI is Harder Than We Think." arXiv preprint, 2021.

==============================================================================
ARTICLE: artificial-heart
TITLE:   Artificial heart
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/artificial-heart
SOURCE:  https://futurehumanwiki.com/raw/artificial-heart
==============================================================================

---
title: "Artificial heart"
slug: "artificial-heart"
type: "technology"
status: "established"
horizon: "present"
trl: 8
categories: ["bodies"]
tags: ["circulatory support", "implants", "heart failure", "transplantation", "medical devices", "haemocompatibility"]
summary: "Mechanical devices that take over some or all of the heart's pumping function, from routine ventricular assist pumps to total replacements used mainly as a bridge to transplant."
updated: "2026-07-27"
humanEvidence: "Randomized trials in adults with advanced heart failure show left ventricular assist devices improve survival, the magnetically levitated pump holding its advantage at five years; total artificial hearts rest on bridge-to-transplant series."
access: "Implanted in specialist cardiac surgery centres in high-income health systems; a durable assist device and its implantation run to hundreds of thousands of dollars, and total artificial hearts are used in tens of patients a year."
reversibility: "difficult"
issues: ["The 2024-2025 magnetically levitated total artificial heart is described without a source.", "The French bioprosthetic device and the withdrawn competitor pump are unnamed and uncited."]
---

```infobox
{
  "caption": "Mechanical circulatory support",
  "rows": [
    { "label": "Main classes", "value": "Ventricular assist device, total artificial heart" },
    { "label": "First animal implant", "value": "1957" },
    { "label": "First clinical use", "value": "1969, as a bridge" },
    { "label": "Dominant flow type", "value": "Continuous, magnetically levitated" },
    { "label": "Typical indication", "value": "Advanced left heart failure" },
    { "label": "Main complications", "value": "Stroke, GI bleeding, driveline infection" },
    { "label": "Status", "value": "Assist devices routine; total hearts rare" }
  ]
}
```

**Artificial heart** is the general term for a mechanical pump that takes over the circulatory work of a failing heart. The category divides into ventricular assist devices, which work alongside a heart left in place, and total artificial hearts, which replace the ventricles entirely after the native heart is excised. The first are a routine therapy implanted in thousands of patients a year; the second remain rare, used mostly to keep a patient alive until a donor organ appears. The engineering history of both is a record of trading one failure mode for another.

## How it works

An adult heart moves roughly five litres a minute against arterial pressure, adjusts that output over a fivefold range in response to demand, and does so for decades without maintenance in a fluid that clots on contact with foreign surfaces. Any replacement has to approximate all four properties.

**Pulsatile designs** copy the heart's method: a flexible chamber compressed by air or a pusher plate, with inflow and outflow valves. They generate physiological pulse pressure and can be volume-matched to the patient. They are also large, contain flexing membranes that fatigue, and require valves that wear.

**Continuous-flow designs** use a rotating impeller. They have one moving part, are far smaller, and have no flexing components to fail. Early axial-flow pumps rested the rotor on mechanical or hydrodynamic bearings; current designs suspend it magnetically, so nothing touches, wear is negligible, and the gaps through which blood passes can be widened to reduce shear damage. The cost is that the patient often has little or no pulse, which turns out to have physiological consequences of its own.

Control is the third problem. A native heart increases output automatically with venous return. A fixed-speed rotary pump does not, so devices either run at a set speed with limited adaptation or use pressure and flow sensors to autoregulate, an approach used in some total artificial hearts.

## Development history

```timeline
[
  { "year": "1957", "title": "First animal implant", "text": "Willem Kolff and Tetsuzo Akutsu implant a polyvinyl chloride total artificial heart in a dog at the Cleveland Clinic, which survives about ninety minutes." },
  { "year": "1969", "title": "First clinical use", "text": "Denton Cooley implants a device designed by Domingo Liotta in a patient at the Texas Heart Institute as a bridge to transplant, supporting him for roughly three days." },
  { "year": "1982", "title": "Jarvik-7", "text": "William DeVries implants the pneumatic Jarvik-7 in Barney Clark at the University of Utah as permanent therapy. Clark survives 112 days, tethered to a large pneumatic console, and the case triggers lasting debate about experimental surgery." },
  { "year": "2001", "title": "REMATCH", "text": "A randomised trial shows that a left ventricular assist device roughly halves one-year mortality compared with medical therapy in patients ineligible for transplant, converting assist devices from a bridge into a destination therapy." },
  { "year": "2004–2010", "title": "The rotary turn", "text": "Continuous-flow pumps displace pulsatile ones. A pneumatic total artificial heart is approved as a bridge to transplant, and remains the only one in sustained use for two decades." },
  { "year": "2024–2025", "title": "Magnetically levitated total heart", "text": "A titanium rotary total artificial heart with a single levitated rotor is implanted in its first patients, one of whom is supported for more than a hundred days before transplant." }
]
```

## Ventricular assist devices

The success story is the left ventricular assist device. A pump the size of a small fruit is implanted at the apex of the left ventricle, draws blood from it, and returns it to the ascending aorta, unloading the failing ventricle while preserving whatever native function remains. Power and control run through a driveline that exits the abdominal wall to an external controller and batteries.

The randomised evidence dates to 2001, when a trial in patients ineligible for transplant found that an assist device substantially reduced one-year mortality compared with optimal medical management, at the cost of frequent adverse events.[^rose2001] Devices have improved considerably since. The current standard is a centrifugal pump with a fully magnetically levitated rotor and an intermittent speed modulation that produces an artificial pulse; in a large randomised comparison against the previous axial-flow generation, it eliminated pump thrombosis almost entirely and reduced stroke, and the survival advantage held at five years.[^mehra2019] An earlier competing device was withdrawn from the market in 2021 after neurological events and pump-stop failures.

Assist devices are now implanted in patients who will never receive a transplant, and some remain supported for years. They do not restore normal physiology. Patients carry batteries, cannot swim, and take anticoagulants indefinitely.

## Total artificial hearts

Total replacement is a different problem, because both ventricles must be removed and both circulations driven, and because there is no residual native function to fall back on if the device stops.

The pneumatic total artificial heart approved in the United States in 2004 as a bridge to transplant is the descendant of the Jarvik-7 and has been implanted in well over a thousand patients; the pivotal study reported that most recipients survived to transplantation, against poor odds without it.[^copeland2004] It works, and it is bulky, noisy, and tethers the patient to a driver. A fully implantable pulsatile device tested in the early 2000s reached a handful of patients, one supported for well over a year, and was discontinued.

A French bioprosthetic design combines a hybrid membrane of bovine pericardium on the blood-contacting side with electrohydraulic actuation and pressure sensors that adjust output automatically, and received European marking; its manufacturer has struggled financially, and its long-term availability is uncertain as of 2026.

The most recent entrant abandons the anatomical metaphor entirely. It uses a single titanium rotor, magnetically levitated, with impeller surfaces on both faces so that one spinning part drives both the systemic and pulmonary circulations. There are no valves and no flexing membranes, and therefore nothing to fatigue. First implanted in a patient in 2024 as a bridge to transplant, it has since supported a patient for more than a hundred days, with early feasibility testing continuing. Whether a valveless rotary total heart is haemocompatible over years rather than months is the question the device exists to answer.

```compare
{
  "columns": ["LVAD", "Total artificial heart", "Heart transplant"],
  "rows": [
    { "label": "Native heart", "values": ["Retained", "Excised", "Excised"] },
    { "label": "Availability", "values": ["Manufactured on demand", "Manufactured on demand", "Limited by donation"] },
    { "label": "Immunosuppression", "values": ["None", "None", "Lifelong"] },
    { "label": "Typical support duration", "values": ["Years", "Months", "Over a decade"] },
    { "label": "Anticoagulation", "values": ["Required", "Required", "Not routinely"] },
    { "label": "Annual volume", "values": ["Thousands", "Tens", "Thousands"] }
  ]
}
```

## The haemocompatibility problem

Every complication that limits mechanical circulatory support traces to blood meeting an engineered surface at non-physiological shear.

**Thrombosis and stroke.** Blood clots on foreign surfaces and where flow stagnates. Patients require anticoagulation, which shifts risk toward haemorrhage; stroke, both ischaemic and haemorrhagic, remains the most feared complication.

**Acquired von Willebrand syndrome.** High shear in a rotary pump unfolds and cleaves the large von Willebrand factor multimers that mediate platelet adhesion. Combined with reduced pulsatility, which promotes intestinal angiodysplasia, this produces gastrointestinal bleeding in a substantial fraction of patients — a bleeding disorder manufactured by the device in a patient who also needs anticoagulation.

**Haemolysis.** Shear damages red cells directly, releasing free haemoglobin, which scavenges nitric oxide and injures the kidney. The mechanism is the same one that defeated cell-free oxygen carriers described in [[artificial-blood]].

**Right heart failure.** Unloading the left ventricle increases venous return to a right ventricle that is often also diseased, and right-sided failure after implantation is a leading cause of early death.

## Power, infection and the untethered goal

The percutaneous driveline is the single largest source of avoidable morbidity. It is a permanent breach of the skin, and infection along it is common with time. The problem is generic to devices that cross the body surface: bone-anchored limbs face the same skin-implant interface, discussed in [[osseointegration]], and it is the reason most [[neuroprosthetics]] are fully implanted with an inductive link rather than wired through the skin. Transcutaneous energy transfer, in which power is coupled through intact skin by induction, has been demonstrated and has not become standard, because the coupling efficiency, heat generation and internal battery lifetime all impose their own costs. Implanted stimulators such as those used in [[deep-brain-stimulation]] solved a version of this problem by drawing milliwatts; a circulatory pump draws several watts continuously, which is what makes wireless power hard. A fully implanted system also has to solve what happens when the internal battery fails.

> [!key] Why mechanical support is not a bridge to nowhere
> Roughly four and a half thousand heart transplants are performed annually in the United States against a far larger population with advanced heart failure, the imbalance described in [[organ-shortage]]. Devices are the only intervention whose supply is not capped by donation, which is why they persist despite complication rates that would be unacceptable in almost any other field. The alternative supply strategies are edited pig hearts, described in [[xenotransplantation]], and grown tissue, described in [[lab-grown-organs]] — neither of which has supported a human for as long as a pump has.

## Outlook

Two directions are being pursued. One is incremental: better surfaces, smarter speed control, fully implantable power, and reduced anticoagulation requirements, all aimed at making an assist device something a patient can live with for a decade rather than tolerate for a few years. The other is replacement of the pump concept altogether by biological means: a regenerated myocardium, an engineered cardiac patch built by [[tissue-engineering]] from cardiomyocytes differentiated from [[induced-pluripotent-stem-cells]], a construct assembled by [[organ-bioprinting]], or a heart rebuilt on the matrix of a donor organ as described in [[decellularized-scaffolds]]. On the scale set out in [[technology-readiness-level]], mechanical support sits near the top and every biological alternative near the bottom, which is unusual for a field where the biological option is usually the older one.

Cost is the quiet constraint. A durable assist device and its implantation run to hundreds of thousands of dollars, which restricts the therapy to health systems that can absorb it and makes it a standing example of the problems collected under [[access-and-inequality]].

The honest comparison is that mechanical support has a track record measured in decades and hundreds of thousands of patients, while every biological alternative has a track record measured in months and single figures. The question for the next decade is not whether pumps will be replaced but whether a total artificial heart can become a destination therapy rather than a waiting room, which requires a device that neither clots nor bleeds its recipient, and no design has yet demonstrated that over years.

## See also

- [[artificial-blood]]
- [[xenotransplantation]]
- [[organ-shortage]]
- [[lab-grown-organs]]
- [[organ-bioprinting]]
- [[neuroprosthetics]]
- [[decellularized-scaffolds]]
- [[tissue-engineering]]

## References

[^rose2001]: `paper` Rose, E. A. et al. "Long-term use of a left ventricular assist device for end-stage heart failure." *New England Journal of Medicine*, 2001. {Enrolled patients ineligible for transplant and compared the device with medical therapy, so it says nothing about devices against transplantation.}
[^mehra2019]: `paper` Mehra, M. R. et al. "A fully magnetically levitated left ventricular assist device — final report." *New England Journal of Medicine*, 2019.
[^copeland2004]: `paper` Copeland, J. G. et al. "Cardiac replacement with a total artificial heart as a bridge to transplantation." *New England Journal of Medicine*, 2004. {Not a randomized trial: it reports survival to transplantation against a comparison group, which is the general limit of bridge-to-transplant evidence.}

==============================================================================
ARTICLE: artificial-womb
TITLE:   Artificial womb
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/artificial-womb
SOURCE:  https://futurehumanwiki.com/raw/artificial-womb
==============================================================================

---
title: "Artificial womb"
slug: "artificial-womb"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 5
categories: ["reproduction", "bodies"]
tags: ["reproduction", "ectogenesis", "prematurity", "neonatology", "bioethics", "medical devices"]
summary: "A device that sustains fetal development outside a body by perfusing the umbilical circulation while the fetus remains immersed in fluid, tested so far only in animals."
updated: "2026-07-27"
humanEvidence: "No human has been supported by a modern system and no trial has enrolled anyone; the results are from preterm lambs, and 1960s perfusion of previable human fetuses produced no survivors."
access: "Nothing to obtain: the systems exist only as research prototypes in a handful of laboratories, and no regulator has authorised a first-in-human trial."
reversibility: "difficult"
issues: ["The EVE platform and the European perinatal life support consortium are described without sources.", "The 22-week viability figure in the keyfacts block carries no citation."]
---

```infobox
{
  "caption": "Experimental life-support technology",
  "rows": [
    { "label": "Type", "value": "Extracorporeal life support" },
    { "label": "Target use", "value": "Extreme prematurity" },
    { "label": "First lamb study", "value": "2017" },
    { "label": "Longest animal support", "value": "About four weeks" },
    { "label": "Human trials", "value": "None reported as of 2026" },
    { "label": "Key programmes", "value": "EXTEND, EVE, PLS" },
    { "label": "Broader concept", "value": "Ectogenesis", "link": "/wiki/ectogenesis" }
  ]
}
```

**Artificial womb** is the common name for a device that supports a fetus outside a pregnant body by connecting it to an external oxygenator through the umbilical vessels while it remains submerged in a fluid environment. Every system built so far is designed for *partial* ectogenesis: it takes over from a pregnancy already underway, at the edge of viability, and aims to carry the fetus a few more weeks. No device has gestated a mammal from conception, and none has been used in a human. The distinction between continuing a pregnancy and replacing one is the single most important thing about the technology and the thing most often lost in coverage of it.

```keyfacts
[
  { "value": "~22 weeks", "label": "Current limit of human viability", "note": "survival at this gestational age remains low and morbidity high" },
  { "value": "4 weeks", "label": "Longest reported animal support", "note": "preterm lambs, 2017 CHOP study" },
  { "value": "0", "label": "Human subjects treated", "note": "as of 2026, no clinical trial has been publicly reported as begun" }
]
```

## How it works

The design problem is that a fetal lung is not a lung yet. Below roughly 26 weeks of human gestation the alveoli are too few and too thick-walled for gas exchange, and mechanical ventilation of such a lung causes the chronic injury known as bronchopulmonary dysplasia. Conventional neonatal intensive care manages this trade-off; an artificial womb tries to avoid it by never asking the lung to breathe.

Four features distinguish the current systems from an incubator:

- **Umbilical cannulation.** Blood is drawn from and returned to the umbilical vein and arteries, preserving fetal circulation with its right-to-left shunts through the ductus arteriosus and foramen ovale. Neck or chest cannulation, as used in standard extracorporeal membrane oxygenation, disturbs that pattern.
- **A pumpless circuit.** In the Children's Hospital of Philadelphia design, the fetal heart alone drives blood through a low-resistance membrane oxygenator. Removing the pump removes a major source of haemolysis and of pressure injury to a fetal heart that cannot tolerate afterload.
- **Fluid immersion.** The fetus floats in a sterile, circulating fluid resembling amniotic fluid, which it swallows and which fills the lungs. Fluid distension is part of normal lung growth; air is not.
- **A closed, sterile bag.** The enclosure excludes light, handling, and airborne organisms, addressing infection as a leading cause of death in extremely preterm infants.

Oxygenation is titrated to fetal rather than neonatal targets, which are much lower than adult values, and it relies on the fetus's own haemoglobin rather than any [[artificial-blood|oxygen-carrying substitute]]. Nutrition is delivered parenterally. The circuit is, in engineering terms, a carefully tuned variant of extracorporeal life support, closer to a [[artificial-heart|mechanical circulatory support]] problem than to anything in reproductive medicine.

## Development history

```timeline
[
  { "year": "1923", "title": "The word is coined", "text": "J. B. S. Haldane proposes 'ectogenesis' in Daedalus, imagining gestation moved wholly outside the body." },
  { "year": "1950s–1960s", "title": "First perfusion attempts", "text": "Researchers including Robert Goodlin perfuse previable human fetuses in pressurised oxygenated fluid. Survival is measured in hours and the work draws lasting ethical criticism." },
  { "year": "1980s–1990s", "title": "Goat fetuses in Tokyo", "text": "Yoshinori Kuwabara's group maintains goat fetuses in an extrauterine incubation system for up to several weeks, but with poor neurological outcomes and dependence on paralysis and sedation." },
  { "year": "2017", "title": "The biobag", "text": "A Philadelphia team reports eight preterm lambs supported for up to four weeks with normal growth, lung maturation and myelination." },
  { "year": "2019–2024", "title": "Parallel programmes", "text": "The EVE platform in Australia and Japan and a European perinatal life support consortium pursue related designs; CHOP's work is commercialised through Vitara Biomedical." },
  { "year": "2023", "title": "Regulators take it up", "text": "The US Food and Drug Administration convenes its Pediatric Advisory Committee to discuss what evidence a first-in-human trial would require." }
]
```

The modern line begins with the 2017 lamb study from Alan Flake's group at the Children's Hospital of Philadelphia.[^partridge2017] Lambs were delivered at 105 to 120 days of gestation, physiologically comparable to about 23 or 24 human weeks, cannulated through the umbilical cord, and placed in a polyethylene bag of circulating fluid. They grew, laid down wool, opened their eyes, and showed lung and brain development approaching that of age-matched controls. The system was named EXTEND, for Extra-uterine Environment for Neonatal Development.

Two other groups have reached comparable results in sheep. The EVE platform, developed by an Australian and Japanese collaboration, has focused on fetuses compromised by intrauterine inflammation, a common precipitant of preterm birth. A European consortium centred on Eindhoven has worked on a design intended to be filled and used in a delivery room, and has built high-fidelity manikins to train the cannulation, which must be done within minutes of delivery.

Earlier work matters mainly as a warning. Kuwabara's goat experiments in Tokyo achieved long survival but the animals were sedated and paralysed throughout, could not be weaned, and showed signs of neurological injury. The 1960s human perfusion experiments, conducted on previable fetuses obtained after abortion, produced no survivors and are now cited chiefly in discussions of research ethics.[^romanis2018]

## Current state

As of 2026 no human has been supported by an artificial womb, and no clinical trial has been publicly reported as having enrolled a participant. On the [[technology-readiness-level|standard readiness scale]] the leading systems sit at validation in a relevant environment: they work reliably in a large-animal model that resembles the intended use but is not it. The FDA's 2023 advisory meeting did not authorise a trial; it aired the questions a sponsor would have to answer, and those questions turned out to be harder than the engineering.[^fda2023]

The central one is comparison. The intended population is infants born at 22 to 24 weeks, where outcomes vary enormously between centres and where the decision to resuscitate at all is a matter of parental choice and institutional policy. A trial would have to randomise between the device and active neonatal intensive care at a moment of crisis, obtain consent from parents in the middle of a preterm delivery, and define success in terms of long-term neurodevelopment rather than survival to discharge. Committee members also raised the problem of how a device tested in lambs translates to a human fetus whose brain development is on a different schedule, and how a failed cannulation would be handled.

The technology's regulatory identity is itself unsettled. It is a medical device, but one whose subject may not have legal personhood under the law of the jurisdiction where it is used, and whose status is different again if it is delivered by caesarean specifically to be placed in the device. The legal scholar Elizabeth Chloe Romanis has proposed the term "gestateling" for an entity that is neither fetus nor neonate because it is neither gestating in a body nor breathing air; the term has no statutory standing anywhere.

> [!caution] What the lamb studies do not show
> Normal growth and myelination in a lamb over four weeks is not evidence of normal human neurodevelopment over the years in which it would be measured. Sheep are precocial, with a shorter gestation and a differently timed brain growth spurt, and the lambs in these studies were euthanised at the end of support rather than followed to adulthood.

## Complete ectogenesis

Gestating a human from fertilisation to term outside a body — *complete* ectogenesis, in the
vocabulary set out in [[ectogenesis]] — is a different technical problem, not a longer version of the
same one. The unsolved parts are at the beginning, not the end. The only established way for someone without a functioning uterus to carry a pregnancy runs in the opposite direction: [[uterus-transplantation]] moves the organ into the patient rather than the fetus into a machine.

Implantation and placentation involve an invasive, immunologically negotiated interaction between trophoblast and maternal endometrium that no device reproduces. Trophoblast [[organoids|organoids]] grown from placental tissue reproduce part of the fetal side of that exchange in a dish, and endometrial cultures reproduce part of the maternal side, but nobody has assembled a functioning placenta from either. Culture systems have made progress at both ends of the gap and none in the middle. Human embryos are routinely cultured to the blastocyst stage for [[embryo-selection]], and research protocols have extended culture to around the [[synthetic-embryos|fourteen-day]] limit that governs such work in most jurisdictions. At the other end, the artificial womb handles the last trimester in animals. Between roughly two weeks and twenty-two weeks lies the entire process of placental development, organogenesis, and the endocrine dialogue between fetus and pregnant body.

The most notable advance in the middle range is not human. A Weizmann Institute group cultured mouse embryos ex utero from before gastrulation through late organogenesis in a rotating-bottle system with continuous perfusion and controlled gas pressure, reaching roughly half of mouse gestation.[^hanna2021] Extending that to a species with an invasive haemochorial placenta and a nine-month gestation is not a matter of scaling the apparatus.

> [!debate] Whether complete ectogenesis is even a goal
> Some researchers treat partial support as a stepping stone; others argue the two problems share almost no engineering, and that framing neonatal life support as proto-ectogenesis has attracted attention and opposition the clinical programme did not need.

## Limitations and risks

Cannulating a vessel a few millimetres wide, in a fetus weighing under 600 grams, within minutes of delivery, is the immediate technical constraint. Umbilical vessels constrict on exposure and the window is short.

Anticoagulation is the chronic one. Any extracorporeal circuit demands systemic anticoagulation, and extremely preterm infants are already at high risk of intraventricular haemorrhage. The trade-off between circuit thrombosis and brain bleeding has no obviously safe setting and is the failure mode most likely to appear first in humans. Infection of an indwelling circuit, haemolysis, and the unknown long-term effects of a fluid environment without maternal hormonal signalling round out the list.

There is also a scope-creep risk that is social rather than technical. A device that improves outcomes at 22 weeks changes what counts as viable, and viability is a threshold written into abortion law in several countries, most consequentially in the United States before 2022 and in the statutory time limits used elsewhere. Bioethicists on both sides of the abortion debate have noted that the technology decouples "can survive outside the body" from "can survive without a body's support", which is the distinction the legal concept was tracking. Whether courts would follow the technology is unresolved, and it is one of the clearest live cases for the [[precautionary-principle|precautionary reasoning]] applied to medical devices whose second-order effects run through law rather than physiology.

## Social and ethical arguments

Feminist writing on ectogenesis has been split since Shulamith Firestone argued in 1970 that removing gestation from women's bodies would remove the biological basis of their subordination; critics answered that the apparatus would transfer authority over fetal welfare from pregnant people to clinicians and courts.[^firestone1970] That argument concerns a machine nobody has built. The device on the bench raises a smaller and harder set of questions: what parents can meaningfully be told during a preterm delivery, who decides when support is withdrawn from an entity the law has no name for, and what obligation a sponsor incurs to follow the survivors for the decades over which the outcomes of interest appear.

Bioconservative objections of the kind catalogued in [[bioconservatism]] have attached to the technology mainly through its imagined form rather than its actual one, treating a manufactured gestation as a category error about how humans come into the world. Advocates in the tradition of [[procreative-beneficence]] make the opposite move, arguing that if a device produces better outcomes for infants who would otherwise die or survive disabled, the burden of justification falls on refusing it. Disability scholars, whose position is set out in [[disability-rights-and-enhancement]], have raised a third concern: that a technology framed around rescuing infants at 22 weeks will be evaluated on survival statistics rather than on the lives of the survivors.

Distributional questions follow the pattern set out in [[access-and-inequality]]. Extremely preterm birth is unevenly distributed, concentrated among populations with the least access to tertiary neonatal care, and a device requiring a specialised surgical team at the moment of delivery is not obviously going to reach them. The counterargument is that neonatal intensive care itself followed the same diffusion path and is now widespread in high-income systems.

For the wider argument about detaching reproduction from bodies, including its interaction with [[in-vitro-gametogenesis]], the popular imagery collected under [[designer-babies]], and claims about [[morphological-freedom]], see [[ectogenesis]]. The near-term question is narrower and sharper: whether any institutional review board will approve a first-in-human trial in a population that cannot consent, for a device whose comparator is a therapy that already saves some of these infants, on the strength of evidence from sheep.

## See also

- [[ectogenesis]]
- [[in-vitro-gametogenesis]]
- [[synthetic-embryos]]
- [[reproductive-longevity]]
- [[embryo-selection]]
- [[tissue-engineering]]
- [[bioethics-of-enhancement]]
- [[access-and-inequality]]

## References

[^partridge2017]: `paper` Partridge, E. A. et al. "An extra-uterine system to physiologically support the extreme premature lamb." *Nature Communications*, 2017. {Eight preterm lambs supported for up to four weeks; the animals were euthanised at the end of support, so the study measures growth and not later development.}
[^romanis2018]: `paper` Romanis, E. C. "Artificial womb technology and the frontiers of human reproduction: conceptual differences and potential implications." *Journal of Medical Ethics*, 2018.
[^hanna2021]: `paper` Aguilera-Castrejon, A. et al. "Ex utero mouse embryogenesis from pre-gastrulation to late organogenesis." *Nature*, 2021.
[^firestone1970]: `book` Firestone, S. *The Dialectic of Sex: The Case for Feminist Revolution*. Morrow, 1970.
[^fda2023]: `regulator` US Food and Drug Administration, Pediatric Advisory Committee. Meeting on artificial womb technology for extremely preterm infants, September 2023. {An advisory committee discussion of what a first-in-human trial would have to show; it authorised nothing and issued no requirements.}

==============================================================================
ARTICLE: asilomar-conference
TITLE:   Asilomar Conference on Recombinant DNA
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/asilomar-conference
SOURCE:  https://futurehumanwiki.com/raw/asilomar-conference
==============================================================================

---
title: "Asilomar Conference on Recombinant DNA"
slug: "asilomar-conference"
type: "event"
status: "historical"
horizon: "historical"
categories: ["genetics", "society"]
tags: ["governance", "biosafety", "recombinant dna", "self-regulation", "moratorium", "history of science"]
summary: "The 1975 meeting at which molecular biologists lifted their own moratorium on recombinant DNA work by agreeing a system of physical and biological containment."
updated: "2026-07-27"
issues: ["The Cambridge citizens review board and the failed late-1970s US bills are described without sources."]
---

```infobox
{
  "caption": "Scientific self-governance meeting",
  "rows": [
    { "label": "Dates", "value": "24–27 February 1975" },
    { "label": "Location", "value": "Pacific Grove, California" },
    { "label": "Participants", "value": "About 140" },
    { "label": "Chair", "value": "Paul Berg" },
    { "label": "Subject", "value": "Recombinant DNA biohazards" },
    { "label": "Outcome", "value": "Containment framework; moratorium lifted" },
    { "label": "Institutional legacy", "value": "NIH Guidelines, 1976" }
  ]
}
```

The **Asilomar Conference on Recombinant DNA** was a four-day meeting of about 140 scientists, lawyers, and journalists held in February 1975 at a conference ground on the California coast, convened to decide whether and how experiments splicing DNA between species should resume. Its participants had imposed a voluntary moratorium on themselves the previous year; at Asilomar they lifted it, replacing prohibition with a graded system of laboratory containment. The meeting is invoked, half a century later, as the template for scientific self-governance — including in the debates over [[germline-editing]] — and the accuracy of that analogy is itself contested.

## Background

```timeline
[
  { "year": "1971", "title": "An experiment suspended", "text": "Paul Berg's laboratory shelves a plan to propagate SV40 tumour virus DNA in E. coli after a colleague raises the possibility of a biohazard." },
  { "year": "1973", "title": "The Gordon Conference letter", "text": "Attendees at a nucleic acids meeting vote to send a public warning to the National Academy of Sciences about the hazards of the new splicing methods." },
  { "year": "1974", "title": "The moratorium", "text": "A National Academy committee chaired by Berg asks researchers worldwide to defer two classes of experiment voluntarily and calls for an international meeting. Compliance is near-total." },
  { "year": "1975", "title": "Asilomar", "text": "About 140 participants meet for four days in Pacific Grove, agree a framework of physical and biological containment graded to risk, and lift the moratorium." },
  { "year": "1976", "title": "NIH Guidelines", "text": "The framework is codified for federally funded research and enforced through institutional biosafety committees." },
  { "year": "1976–1977", "title": "Cambridge intervenes", "text": "The Cambridge, Massachusetts city council imposes a local moratorium and convenes a citizens' review board, which recommends permitting the work with additional conditions." },
  { "year": "1978–1982", "title": "Relaxation", "text": "The guidelines are progressively loosened as the feared hazards fail to materialise and commercial biotechnology takes hold." }
]
```

The techniques arrived quickly. By the early 1970s it was possible to cut DNA at defined sequences with restriction enzymes, join fragments from different organisms, and propagate the result in bacteria. Paul Berg's laboratory at Stanford planned to insert DNA from SV40, a monkey virus that causes tumours in rodents, into a bacteriophage that grows in *Escherichia coli* — a bacterium that lives in the human gut. A colleague pointed out the obvious hazard, and Berg suspended the experiment.

Concern spread through the field rather than into it from outside. At a 1973 Gordon Research Conference the attendees voted to send a letter to the National Academy of Sciences and to *Science*, warning that the new methods might create biological hazards. The following year a National Academy committee chaired by Berg published a short letter in *Science*, *Nature*, and *PNAS* asking researchers worldwide to defer voluntarily two classes of experiment — those involving antibiotic resistance genes or toxin genes, and those cloning DNA from tumour viruses — until the risks could be assessed, and calling for an international meeting.[^berg1974] Compliance was, by later accounts, essentially complete. That fact is the foundation of Asilomar's reputation: for roughly eight months an entire research community stopped doing something it wanted to do, on the strength of a letter.

## The conference

The organising committee of Berg, David Baltimore, Sydney Brenner, Richard Roblin, and Maxine Singer set a deliberately narrow agenda: the physical risk that recombinant organisms might escape and cause disease. Ethical questions, the eventual application of the techniques to humans, and the entire category now called [[dual-use-research]] — deliberate misuse and biological weapons — were kept off the programme. Journalists were admitted on condition that they publish nothing until the meeting ended.

Progress was slow until the final sessions. What broke the deadlock, by several participants' accounts, was a panel of lawyers who explained the liability that individual investigators and their institutions would face if an accident occurred in the absence of agreed standards. The prospect of personal legal exposure concentrated attention more effectively than the epidemiological arguments had.

The summary statement adopted on the last morning recommended that most work proceed, under containment matched to the estimated risk of each experiment.[^berg1975] A small number of experiments were to remain deferred: cloning DNA from highly pathogenic organisms, genes for potent toxins, and large-scale work with agents whose products might be harmful.

## The containment framework

Two kinds of containment were specified, to be combined.

**Physical containment** graded laboratories from minimal to high — the P1 to P4 scale, later reworked into the biosafety levels still in use. Higher grades required negative-pressure rooms, airlocks, filtered exhaust, and restricted access.

**Biological containment** was the more original idea: engineer the host and the vector so that an escaped organism cannot survive outside the laboratory. Enfeebled strains of *E. coli* K-12 were developed that depended on nutrients unavailable in the human gut or the environment, and vectors were built that could replicate only in those strains. This principle, that safety can be built into the organism rather than only around it, reappears in modern work on [[recoded-organisms]] and synthetic auxotrophy, in the construction of [[synthetic-genomes]], and in the biocontainment discussions surrounding [[gene-drive|gene drives]] and [[mirror-life]].

> [!key] The idea that outlasted the meeting
> Asilomar's specific risk assessments were mostly wrong — the hazards of recombinant *E. coli* proved far smaller than feared. What survived is the structural move: matching a graded control regime to a graded risk estimate, and building containment into the biology. That template now governs work far removed from 1975's concerns.

## Aftermath

The National Institutes of Health issued formal guidelines in 1976 that codified the Asilomar framework for federally funded work, enforced through institutional biosafety committees and administered with advice from the Recombinant DNA Advisory Committee. Over the following six years the guidelines were progressively relaxed as the feared hazards failed to materialise.

Public reaction was not uniformly deferential. The city council of Cambridge, Massachusetts imposed a local moratorium in 1976 and convened a review board of ordinary citizens — a machinist, a nurse, a nun among them — which after months of hearings recommended permitting the research under the NIH guidelines with additional local conditions. The episode is cited by both sides: as evidence that non-specialists can adjudicate technical controversy responsibly, and as evidence that they will generally arrive where the scientists were already standing.

Commercial biotechnology arrived immediately afterwards. Genentech was founded in 1976, and within a few years recombinant insulin was in production. Asilomar took place in the last moment when the field had no industry, no share prices, and no patents at stake, which is one reason its participants could agree to stop. The advisory committee it inspired went on to review individual [[somatic-gene-therapy]] protocols for three decades. The fourteen-day limit on culturing human embryos, agreed by advisory bodies in the years that followed and now under pressure from [[synthetic-embryos|stem-cell-based embryo models]], is the other durable example of a research boundary that scientists drew around themselves.

## Criticism

Historians of the controversy have argued that the meeting's narrow framing was its central political act rather than an incidental feature. By defining the question as biohazard containment, the participants placed it inside their own expertise and outside anyone else's; questions about who should benefit, who bears risk, and whether some applications should exist at all were ruled off-topic and never reinstated.[^krimsky1982][^wright1994] On this reading Asilomar succeeded in its unstated purpose, which was to forestall statutory regulation by demonstrating that the field could police itself. Several bills to regulate recombinant DNA were introduced in the US Congress in the late 1970s; none passed.

A second criticism concerns who was in the room. Every substantive participant was a researcher. No patients, no workers who would handle the organisms, no members of affected publics took part. The later observation that Asilomar established scientists as the arbiters of which questions count has been developed by scholars examining its use as a precedent for genome editing.[^hurlbut2015]

## Why the analogy is contested

When [[crispr-cas9]] made human embryo editing feasible, a 2015 meeting in Napa convened by [[jennifer-doudna]] and colleagues, several of whom had been at Asilomar, explicitly invoked the precedent and called for a moratorium on clinical germline use pending broader discussion.[^baltimore2015] The comparison is imperfect in at least four ways.

The question is different in kind. Whether an enfeebled bacterium can survive in a gut is an empirical matter on which molecular biologists have genuine authority. Whether it is acceptable to alter the genome of a person who does not yet exist is a question of value on which they have none, and Asilomar's precedent of scientists deciding the scope of the debate is, in that setting, part of the problem rather than the solution.

The community is different. In 1975 recombinant DNA work required rare reagents, rare skills, and a handful of laboratories whose leaders all knew one another. Editing an embryo in 2026 requires equipment available in any IVF clinic and reagents that can be ordered online, distributed across thousands of institutions in dozens of legal jurisdictions — to say nothing of the amateur culture described under [[biohacking]]. A norm agreed by a professional community binds only its members, as the [[he-jiankui-affair]] demonstrated and as privately funded ventures continue to demonstrate.

The moratorium was different. Asilomar's applied to experiments, was expected to be temporary, and lasted less than a year. Proposals for germline moratoria concern applications and are open-ended, which makes them harder to sustain and easier to defect from.

Finally, the stakes are structurally different. A containment failure in 1975 would have been an accident with victims and a remedy. A heritable edit is transmitted to descendants who cannot consent and cannot be restored, which is why the frameworks discussed in [[governance-of-genome-editing]] reach for law rather than professional agreement, and why arguments from the [[precautionary-principle]] and from [[differential-technological-development]] have more traction here than they did in 1975. Whether any equivalent of the 1974 letter could command the same compliance today has not been tested, and there is little reason for confidence that it would.

## See also

- [[germline-editing]]
- [[governance-of-genome-editing]]
- [[he-jiankui-affair]]
- [[crispr-cas9]]
- [[dual-use-research]]
- [[precautionary-principle]]
- [[gene-drive]]
- [[mirror-life]]

## References

[^berg1974]: `paper` Berg, P. et al. "Potential Biohazards of Recombinant DNA Molecules." *Science*, 1974. {The short letter that proposed the voluntary moratorium and called the meeting; it asks for deferral pending an assessment rather than presenting one.}
[^berg1975]: `paper` Berg, P., Baltimore, D., Brenner, S., Roblin, R.O., Singer, M.F. "Summary Statement of the Asilomar Conference on Recombinant DNA Molecules." *Proceedings of the National Academy of Sciences*, 1975.
[^krimsky1982]: `book` Krimsky, S. *Genetic Alchemy: The Social History of the Recombinant DNA Controversy*. MIT Press, 1982.
[^wright1994]: `book` Wright, S. *Molecular Politics: Developing American and British Regulatory Policy for Genetic Engineering, 1972–1982*. University of Chicago Press, 1994.
[^hurlbut2015]: `paper` Hurlbut, J.B. "Limits of Responsibility: Genome Editing, Asilomar, and the Politics of Deliberation." *Hastings Center Report*, 2015.
[^baltimore2015]: `paper` Baltimore, D. et al. "A prudent path forward for genomic engineering and germline gene modification." *Science*, 2015. {A position piece by a group including Baltimore and Berg of the 1975 organising committee; it recommends a pause and carries no legal force.}

==============================================================================
ARTICLE: aubrey-de-grey
TITLE:   Aubrey de Grey
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/aubrey-de-grey
SOURCE:  https://futurehumanwiki.com/raw/aubrey-de-grey
==============================================================================

---
title: "Aubrey de Grey"
slug: "aubrey-de-grey"
type: "person"
status: "contested"
horizon: "present"
categories: ["people", "longevity"]
tags: ["aging", "sens", "biogerontology", "life extension", "rejuvenation", "advocacy"]
summary: "British biogerontologist who proposed the SENS damage-repair framework for rejuvenation and popularized the idea of longevity escape velocity."
updated: "2026-07-27"
issues: ["The 2021 misconduct section carries no citation for the investigation or its outcome."]
---

```infobox
{
  "caption": "Biogerontologist and advocate",
  "rows": [
    { "label": "Born", "value": "20 April 1963, London" },
    { "label": "Nationality", "value": "British" },
    { "label": "Education", "value": "University of Cambridge" },
    { "label": "Known for", "value": "SENS; longevity escape velocity", "link": "/wiki/longevity-escape-velocity" },
    { "label": "Field", "value": "Biogerontology" },
    { "label": "Affiliation", "value": "LEV Foundation" },
    { "label": "Former role", "value": "CSO, SENS Research Foundation", "link": "/wiki/sens-research-foundation" }
  ]
}
```

**Aubrey de Grey** is a British biogerontologist who argues that aging should be treated as a
finite set of accumulated molecular and cellular lesions that periodic repair can remove, rather
than as a metabolic process that medicine must learn to slow. He formalized that position as SENS
— Strategies for Engineered Negligible Senescence — and popularized the term
[[longevity-escape-velocity]]. He is the most visible advocate of radical life extension and among
the most heavily criticized figures in the field.

## Overview

De Grey's importance is organizational and rhetorical more than experimental. He has run almost no
wet-lab research of his own, and his primary contribution is a framework: a claim that the damage
accumulated during aging falls into a small number of categories, that each category has a
conceivable repair strategy, and that the list is complete enough to serve as an engineering
specification. Mainstream biogerontology has largely accepted his framing question — whether aging
is tractable at all — while rejecting his timelines and much of his confidence.

## Career

De Grey read computer science at Trinity Hall, Cambridge, and spent his early career as a software
engineer, including work in artificial-intelligence tooling. His entry into biology came through
his marriage to the Cambridge geneticist Adelaide Carpenter and through self-directed reading in
the aging literature. In 1999 he published *The Mitochondrial Free Radical Theory of Aging*, which
argued that mutant mitochondrial DNA spreads clonally within cells; Cambridge awarded him a PhD by
published work on the strength of it.[^dg1999]

Through the early 2000s he organized the SENS conference series in Cambridge, edited the journal
*Rejuvenation Research*, and built a donor network. In 2009 he co-founded the
[[sens-research-foundation]] in California, and later committed a substantial personal inheritance
to it and to allied organizations. The foundation funded work on lysosomal enzymes for
[[proteostasis|aggregate clearance]], allotopic expression of mitochondrial genes, and immune
approaches to [[cellular-senescence|senescent cells]] — several years before [[senolytics]] became
a mainstream target.

## SENS and the damage-repair thesis

The name is borrowed from comparative biology: a species shows
[[negligible-senescence]] when its mortality rate does not rise measurably with age. SENS
proposes to engineer that condition rather than find it. It partitions age-related damage into
seven categories: cell loss and atrophy, cancerous
mutations, [[mitochondrial-dysfunction|mitochondrial DNA mutations]], death-resistant cells,
extracellular crosslinks, extracellular aggregates, and intracellular aggregates. For each, de Grey
proposes a repair rather than a prevention: replace lost cells with
[[induced-pluripotent-stem-cells|stem-cell derivatives]], remove senescent cells, back up
mitochondrial genes in the nucleus, break crosslinks pharmacologically, clear amyloid immunologically,
and supply cells with enzymes capable of digesting what lysosomes cannot.[^ea2007]

> [!key] Why the framing matters
> Damage repair is agnostic about the causes of aging. If the categories are complete, a therapy
> works even if nobody knows why the damage forms — the same logic by which a mechanic restores a
> car without a theory of corrosion. The whole argument rests on completeness, which is exactly what
> critics dispute.

The scheme predates and partly overlaps the [[hallmarks-of-aging]] taxonomy published in 2013, and
the two are often compared. The hallmarks are a consensus classification derived from experimental
literature; SENS is a repair programme derived from an engineering premise.

## Longevity escape velocity

De Grey argued in 2004 that rejuvenation therapies need not be perfect to matter, only to improve
faster than the residual damage they fail to remove.[^dg2004] If each generation of therapy buys a
treated person enough additional years for the next generation to arrive, remaining life expectancy
stops falling. He has repeatedly given roughly even odds of reaching that threshold within a
decade or two, conditional on funding, and has said that the first person to live past 1,000 may
already be born — a claim that sits far outside anything the record of
[[maximum-human-lifespan|documented human longevity]] supports. No demographic or biological
evidence establishes a date, and de Grey presents these as conditional forecasts rather than
projections from data. The framing has the useful property of shifting attention from
[[healthspan|how long people stay well]] to the rate at which therapies improve.

## Reception

The strongest institutional test of SENS came in 2005, when *Technology Review* offered a prize to
any molecular biologist who could demonstrate that SENS was too wrong to merit debate. A panel
including Craig Venter, Nathan Myhrvold and Rodney Brooks concluded in 2006 that no submission met
that bar, while also finding SENS highly speculative and largely unsupported by reproducible
results; a critique led by Preston Estep received a partial award.[^tr2006] The verdict has been
cited by both sides ever since — as vindication that SENS is not pseudoscience, and as confirmation
that it is unproven.

A 2005 *EMBO Reports* commentary signed by a large group of biogerontologists, including Steven
Austad and Huber Warner, argued that none of the SENS proposals had been shown to work in a mammal
and that presenting them as a near-term engineering plan misrepresented the state of the
science.[^warner2005] That remains the mainstream position: the categories are a useful checklist,
the timelines are not credible, and the gap between a mouse result and a human therapy is
systematically understated.

> [!debate] The disagreement in one line
> De Grey holds that the remaining obstacles are engineering problems with known shapes; most
> biogerontologists hold that at least three of the seven categories lack any demonstrated repair
> in a living mammal.

## Departure from SENS Research Foundation

In August 2021, two women in the longevity field publicly described inappropriate conduct by de
Grey. The SENS Research Foundation placed him on leave, commissioned an independent legal
investigation, and terminated his employment. De Grey denied wrongdoing and disputed the process.
The episode split the organization and its donor base, and prompted broader discussion of conduct
norms in a field that depends heavily on a small number of private funders.

## Legacy and current work

Since 2022 de Grey has led the Longevity Escape Velocity Foundation, whose main programme is the
Robust Mouse Rejuvenation study: rather than testing single compounds, it combines interventions
already reported individually to extend mouse lifespan — [[rapamycin]] among them — and starts them
in middle age, asking whether the effects add. That design addresses a real gap, since the field has
run very few combination studies, and its results, whatever they show, are more checkable than any
of his forecasts.

His durable influence is that he moved the question. Arguments that aging research was intrinsically
disreputable were common when he began; the [[geroscience-hypothesis]] is now respectable enough to
be discussed with regulators, and senolytics, [[epigenetic-reprogramming]] and
[[aging-biomarkers]] are funded programmes at major institutions. Whether that shift is
attributable to him or merely coincident with him is contested. What is not contested is that the
therapies he described in 2005 as achievable within decades remain, twenty years on, unavailable in
any clinic.

## See also

- [[longevity-escape-velocity]]
- [[sens-research-foundation]]
- [[hallmarks-of-aging]]
- [[senolytics]]
- [[geroscience-hypothesis]]
- [[methuselah-foundation]]
- [[david-sinclair]]
- [[transhumanism]]

## References

[^dg1999]: `book` de Grey, A. *The Mitochondrial Free Radical Theory of Aging*. Landes Bioscience, 1999.
[^dg2004]: `paper` de Grey, A. "Escape Velocity: Why the Prospect of Extreme Human Life Extension Matters Now." *PLoS Biology*, 2004. {An essay setting out the escape-velocity argument; it reasons from an assumed rate of therapeutic improvement rather than an observed one.}
[^ea2007]: `book` de Grey, A. and Rae, M. *Ending Aging: The Rejuvenation Breakthroughs That Could Reverse Human Aging in Our Lifetime*. St. Martin's Press, 2007. {A trade book by the subject setting out his own programme; it argues that the repairs are feasible rather than reporting results.}
[^tr2006]: `news` Pontin, J. "Is Defeating Aging Only a Dream?" *Technology Review* SENS Challenge report, 2006.
[^warner2005]: `paper` Warner, H. et al. "Science fact and the SENS agenda: What can we reasonably expect from ageing research?" *EMBO Reports*, 2005. {A signed commentary by a group of biogerontologists rather than a research paper; it reports their judgment of the SENS proposals.}

==============================================================================
ARTICLE: autophagy
TITLE:   Autophagy
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/autophagy
SOURCE:  https://futurehumanwiki.com/raw/autophagy
==============================================================================

---
title: "Autophagy"
slug: "autophagy"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity"]
tags: ["aging", "autophagy", "lysosome", "mtor", "cell biology", "mechanisms"]
summary: "The lysosomal pathway by which cells digest their own damaged components, and the process through which most known lifespan-extending interventions appear to act."
updated: "2026-07-27"
humanEvidence: "The pathway itself is established in human cells, but autophagic flux cannot be measured in a living person, and every lifespan result comes from yeast, worms, flies and mice."
issues: ["The spermidine claims in the human-data paragraph carry no citations."]
---

```infobox
{
  "caption": "Cellular degradation pathway",
  "rows": [
    { "label": "Named", "value": "1963, by Christian de Duve" },
    { "label": "Genetics established", "value": "1992–1993, yeast screens" },
    { "label": "Nobel Prize", "value": "2016, Yoshinori Ohsumi" },
    { "label": "Core genes", "value": "ATG family, ~40 in yeast" },
    { "label": "Main negative regulator", "value": "mTORC1" },
    { "label": "Hallmark status", "value": "Added to the list in 2023" }
  ]
}
```

**Autophagy** is the process by which a cell delivers its own cytoplasmic contents — misfolded proteins, damaged organelles, invading bacteria, surplus lipid — to the lysosome for degradation and recycling. It is the bulk arm of [[proteostasis|protein quality control]] and the only route by which a cell can dispose of a structure too large for the proteasome. Disabled macroautophagy was added as a separate entry to the [[hallmarks-of-aging]] in 2023, and interfering with autophagy genes abolishes the lifespan extension produced by most known geroprotective interventions in invertebrates.

```keyfacts
[
  { "value": "3", "label": "Forms of autophagy", "note": "macroautophagy, microautophagy, chaperone-mediated" },
  { "value": "2016", "label": "Nobel Prize", "note": "Yoshinori Ohsumi, for the genetics of autophagy" },
  { "value": "0", "label": "Validated human flux assays", "note": "autophagic flux cannot yet be measured in a living person" }
]
```

## Three forms

Macroautophagy, the form usually meant by the unqualified term, engulfs cargo in a double-membraned vesicle that then fuses with a lysosome. Microautophagy invaginates the lysosomal membrane directly to take up small volumes of cytosol. Chaperone-mediated autophagy is selective for individual proteins carrying a KFERQ-like motif, which HSC70 recognizes and delivers to the lysosomal receptor LAMP2A for unfolding and direct translocation across the membrane. Only mammals and some other vertebrates have the third form; it declines steeply with age in rodent liver.

## Mechanics

Macroautophagy begins with the ULK1 kinase complex nucleating an isolation membrane, or phagophore, at a site supplied with lipid by ATG9-containing vesicles and by the class III phosphatidylinositol 3-kinase complex built around VPS34 and Beclin-1. Two ubiquitin-like conjugation cascades then act on the growing membrane. The first builds an ATG12–ATG5–ATG16L1 complex; the second, using that complex as its ligase, attaches LC3 to the lipid phosphatidylethanolamine in the phagophore membrane. Lipidated LC3 recruits cargo receptors, drives membrane expansion, and remains the standard experimental marker of autophagosome formation.

The closed autophagosome traffics along microtubules and fuses with a lysosome. Acid hydrolases degrade the contents, and permeases return amino acids, fatty acids and nucleosides to the cytosol. The whole cycle takes minutes.

## Discovery

```timeline
[
  { "year": "1955", "title": "Lysosome identified", "text": "Christian de Duve characterizes the lysosome by cell fractionation, establishing that cells contain a dedicated degradative compartment." },
  { "year": "1963", "title": "Autophagy named", "text": "De Duve coins the term for the observation that lysosomes sometimes contain recognizable pieces of the cell's own cytoplasm." },
  { "year": "1992–1993", "title": "Genes found", "text": "Yoshinori Ohsumi observes autophagic bodies accumulating in protease-deficient yeast vacuoles and identifies a set of genes required for the process." },
  { "year": "2003–2008", "title": "Link to longevity", "text": "Work in C. elegans shows autophagy genes are required for the lifespan extension produced by reduced insulin/IGF-1 signalling, dietary restriction and TOR inhibition." },
  { "year": "2016", "title": "Nobel Prize", "text": "Ohsumi receives the Nobel Prize in Physiology or Medicine for elucidating the mechanisms of autophagy." }
]
```

De Duve saw autophagy in electron micrographs and named it, but the process resisted study for three decades because it lacked genetic handles. Ohsumi's contribution was to make it tractable: by blocking vacuolar proteases in yeast so that autophagic bodies accumulated visibly, he created a screenable phenotype and identified the core machinery, which turned out to be conserved from yeast to humans.[^tsukada1993]

## Regulation

Autophagy is controlled by the same nutrient-sensing network that governs growth. mTORC1, active when amino acids and growth factors are abundant, phosphorylates ULK1 and ATG13 and holds initiation off; it also phosphorylates TFEB, the transcription factor that drives lysosomal and autophagy gene expression, keeping it out of the nucleus. AMPK, active when the AMP:ATP ratio rises, does the opposite on both counts. Starvation, exercise and [[rapamycin]] therefore converge on the same switch from different directions, which is why [[caloric-restriction|dietary restriction]] and mTOR inhibition produce overlapping transcriptional signatures.

## Selective autophagy

Bulk autophagy is non-specific, but most autophagy under normal conditions is selective. Cargo is ubiquitinated and recognized by receptors — p62/SQSTM1, NBR1, optineurin, NDP52 — which bind both the tag and LC3 on the forming membrane. Named subtypes include mitophagy for [[mitochondrial-dysfunction|damaged mitochondria]], aggrephagy for protein aggregates, lipophagy for lipid droplets, ferritinophagy for iron stores, and xenophagy for intracellular pathogens. Mitophagy is the subtype most directly tied to aging, and the PINK1–Parkin route that governs it is mutated in familial early-onset Parkinson's disease.

Selectivity matters for a specific reason in aging biology. Adult stem cells depend on autophagy to clear the damaged mitochondria and protein aggregates that would otherwise be partitioned asymmetrically at division, and loss of autophagy in haematopoietic and muscle stem cells produces a prematurely aged, myeloid-biased or poorly regenerating compartment in mice — a direct mechanistic bridge to [[stem-cell-exhaustion|stem cell exhaustion]]. In macrophages, failure to degrade cytosolic DNA and damaged organelles activates inflammasome signalling, one of the proposed sources of [[inflammaging|chronic sterile inflammation]].

## Why longevity interventions route through it

The strongest evidence that autophagy is not incidental to aging comes from epistasis. In *C. elegans*, autophagy genes were shown to be required for the lifespan extension of the long-lived insulin-signalling mutants [[cynthia-kenyon|Cynthia Kenyon's]] laboratory characterized, and later for the extension produced by dietary restriction and by TOR inhibition.[^melendez2003][^hansen2008] The pattern holds in *Drosophila*. Gain-of-function experiments in mice point the same way: animals engineered to express extra ATG5, and animals carrying a Beclin-1 point mutation that frees it from inhibition by BCL2, both show elevated basal autophagy, better metabolic health, and extended lifespan.[^pyo2013][^fernandez2018]

The argument is epistatic rather than sufficient. Showing that autophagy is required for an intervention's effect does not show that raising autophagy alone would reproduce it, and the mouse gain-of-function experiments alter the pathway from conception rather than in an already-aged animal. Drugs promoted on autophagy grounds — [[metformin]] through AMPK, [[nad-precursors|NAD⁺ precursors]] through sirtuin-dependent regulation of the machinery — have not been shown to raise autophagic activity in human tissue at the doses used.

> [!key] Why this matters
> Autophagy is the mechanism that most candidate longevity interventions have in common. Rapamycin, dietary restriction, exercise, spermidine and several NAD⁺-related manipulations all raise autophagic activity in model systems, which is either evidence that autophagy is the shared final pathway or evidence that autophagy is easy to raise and easy to measure.

Human data are much weaker. [[exercise-and-aging|Exercise]] induces autophagy in mouse muscle, and the metabolic benefits of exercise are lost in mice unable to mount that response. Spermidine, a polyamine that induces autophagy across species and is abundant in some fermented foods, extends lifespan in yeast, flies, worms and mice, and is associated with lower mortality in observational human cohorts; randomized trials of [[dietary-supplements|supplementation]] in older adults have not shown clear cognitive benefit. Autophagy is an explicit programme focus for at least one well-funded longevity company, [[retro-biosciences|Retro Biosciences]].

## Limits and counterexamples

More autophagy is not uniformly better. Established tumours depend on autophagy to survive nutrient stress, which is why autophagy inhibitors such as hydroxychloroquine have been trialled as cancer adjuvants — the opposite of the geroprotective logic. Monoallelic loss of *BECN1* is found in some breast and ovarian cancers, so the pathway is tumour-suppressive in one phase and tumour-supporting in another, a pattern it shares with [[cellular-senescence|cellular senescence]]. Variants in *ATG16L1* predispose to Crohn's disease through impaired handling of intestinal bacteria rather than through any failure of recycling.

There is also a measurement problem that limits every human claim in this article. Autophagy is a flux, not a level. A high LC3-II signal can mean rapid autophagosome formation or blocked degradation, and distinguishing the two requires adding a lysosomal inhibitor, which cannot be done in a living person. No assay of autophagic flux has been qualified as an [[aging-biomarkers|aging biomarker]], and the field's human evidence therefore rests on downstream functional endpoints rather than on the pathway itself. The consequence is that the central claim — that raising autophagy in an aged human would slow the processes described by the [[geroscience-hypothesis|geroscience hypothesis]] — is currently untestable rather than untested.

## See also

- [[hallmarks-of-aging]]
- [[proteostasis]]
- [[rapamycin]]
- [[mitochondrial-dysfunction]]
- [[caloric-restriction]]
- [[retro-biosciences]]

## References

[^tsukada1993]: `paper` Tsukada, M., Ohsumi, Y. "Isolation and characterization of autophagy-defective mutants of Saccharomyces cerevisiae." *FEBS Letters*, 1993.
[^melendez2003]: `paper` Meléndez, A. et al. "Autophagy genes are essential for dauer development and life-span extension in C. elegans." *Science*, 2003. {Worms, not mammals; the experiment shows autophagy genes are required for the lifespan extension, not that raising autophagy alone extends life.}
[^hansen2008]: `paper` Hansen, M. et al. "A role for autophagy in the extension of lifespan by dietary restriction in C. elegans." *PLoS Genetics*, 2008.
[^pyo2013]: `paper` Pyo, J.-O. et al. "Overexpression of Atg5 in mice activates autophagy and extends lifespan." *Nature Communications*, 2013.
[^fernandez2018]: `paper` Fernández, Á. F. et al. "Disruption of the beclin 1–BCL2 autophagy regulatory complex promotes longevity in mice." *Nature*, 2018. {The mice carry a knock-in mutation present from conception, so the result does not test raising autophagy in an already-aged animal.}

==============================================================================
ARTICLE: base-editing
TITLE:   Base editing
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/base-editing
SOURCE:  https://futurehumanwiki.com/raw/base-editing
==============================================================================

---
title: "Base editing"
slug: "base-editing"
type: "technology"
status: "emerging"
horizon: "late 2020s"
trl: 7
categories: ["genetics"]
tags: ["crispr", "genome editing", "gene therapy", "point mutations", "biotechnology"]
summary: "A genome editing method that chemically converts one DNA base into another at a targeted site without cutting both strands of the double helix."
updated: "2026-07-27"
humanEvidence: "All human use is early-phase or one-off: base-edited donor T cells were given to a teenager with relapsed leukaemia in 2022, in vivo liver editing has lowered cholesterol in early trials, and a bespoke editor was dosed into one infant."
access: "No base-editing therapy is approved anywhere as of 2026; the only routes are early-phase clinical trials and one-off bespoke treatments authorised for a single patient."
reversibility: "irreversible"
issues: ["The Verve and alpha-1 antitrypsin trial results in Current state carry no citations."]
---

```infobox
{
  "caption": "Genome editing platform",
  "rows": [
    { "label": "Type", "value": "Deaminase–nickase fusion" },
    { "label": "First described", "value": "2016 (cytosine), 2017 (adenine)" },
    { "label": "Developed by", "value": "David Liu laboratory" },
    { "label": "Edits possible", "value": "4 of 12 substitutions" },
    { "label": "Double-strand break", "value": "No" },
    { "label": "First human use", "value": "2022, leukaemia CAR-T cells" },
    { "label": "Readiness", "value": "TRL 7" }
  ]
}
```

**Base editing** converts one DNA letter directly into another at a chosen position, using an enzyme that chemically modifies a base rather than a nuclease that severs the chromosome. It was built to solve a specific weakness of [[crispr-cas9]]: most disease-causing mutations are single-letter substitutions, and correcting a single letter with a double-strand break is inefficient and messy. Base editors handle four of the twelve possible substitutions, which covers a large fraction of known pathogenic point mutations but by no means all of them.

## How it works

A base editor is a fusion protein. The targeting half is a catalytically impaired Cas9 that binds DNA under guide-RNA control but cuts at most one strand. The working half is a deaminase, an enzyme that strips an amine group from a base and changes what the cell's polymerases read.

Cytosine base editors, described in 2016, use a cytidine deaminase to convert cytosine to uracil.[^komor2016] Uracil is read as thymine during replication, so a C•G pair becomes T•A. Because cells actively excise uracil from DNA, the editor also carries a uracil glycosylase inhibitor to block that repair, and nicks the opposite strand so the cell rebuilds it using the edited strand as template.

Adenine base editors, described in 2017, had no natural starting enzyme: no known deaminase acts on adenine in DNA. The solution was directed evolution of a bacterial tRNA adenine deaminase until it accepted single-stranded DNA.[^gaudelli2017] The product, inosine, is read as guanine, converting A•T to G•C.

Both editors act on the short stretch of DNA displaced when Cas9 unwinds the helix, giving an editing window of roughly five nucleotides within the protospacer. The window is the source of the method's characteristic error: if more than one editable base falls inside it, all of them may be converted. These **bystander edits** are usually silent or tolerated, but they are not always, and guide design often becomes a search for a position that puts only the intended base in range.

```compare
{
  "columns": ["Nuclease editing", "Base editing", "Prime editing"],
  "rows": [
    { "label": "Double-strand break", "values": ["Yes", "No (single-strand nick)", "No (single-strand nick)"] },
    { "label": "Changes possible", "values": ["Disruption; templated replacement", "4 of 12 substitutions", "All 12 substitutions plus small indels"] },
    { "label": "Needs a repair template", "values": ["Yes, for replacement", "No", "Template carried on the guide"] },
    { "label": "Works in non-dividing cells", "values": ["Poorly for replacement", "Yes", "Yes"] },
    { "label": "Typical unintended product", "values": ["Indels, large deletions", "Bystander edits", "Unwanted insertions at the nick"] },
    { "label": "Cargo size", "values": ["Smallest", "Larger", "Largest"] }
  ]
}
```

## Development history

Both classes came out of David Liu's laboratory, and both were engineered rather than discovered: the adenine editor in particular required many rounds of laboratory evolution before it worked on DNA at all. Successive generations improved on-target activity, narrowed the editing window, and reduced the deaminase's promiscuity. Faster variants shortened the exposure time needed, which matters because most unintended editing accumulates with duration of expression.

The first use in a person came in 2022, when a team at Great Ormond Street Hospital in London treated a teenager with relapsed T-cell acute lymphoblastic leukaemia using donor T cells in which three genes had been disabled by base editing, converting them into a CAR-T product that would not attack itself.[^chiesa2023] The editing was done outside the body, which sidesteps the delivery problem entirely.

## Current state

As of 2026, base editing is in clinical trials rather than approved practice, and the trials divide into two families.

Ex vivo programmes edit cells outside the body and return them. A sickle cell disease candidate uses an adenine editor to recreate a naturally occurring promoter variant that keeps fetal haemoglobin switched on in adulthood — a different route to the same physiological end as [[casgevy]], which disrupts an erythroid enhancer of the repressor gene *BCL11A* instead. Both approaches share the burden that limits autologous stem cell therapy generally: patients need conditioning chemotherapy to clear marrow space, and that conditioning carries real mortality and infertility risk independent of the edit.

In vivo programmes deliver the editor systemically, almost always to the liver, using [[lipid-nanoparticles]]. The first in-human in vivo base editing trial, run by Verve Therapeutics, disabled *PCSK9* in hepatocytes to lower LDL cholesterol permanently; a successor candidate from the same programme reported substantial cholesterol reductions in 2025. A separate liver programme has reported correction — as opposed to disruption — of a disease-causing point mutation in patients with alpha-1 antitrypsin deficiency, which if it holds up would be the first such correction in a person.

> [!key] The bespoke-therapy case
> In 2025 clinicians at the Children's Hospital of Philadelphia treated an infant with a severe urea-cycle disorder using a base editor designed for his specific mutation and delivered by lipid nanoparticle, from diagnosis to dosing in roughly six months.[^musunuru2025] It is the clearest existing demonstration that editing can be personalised, and the clearest illustration that regulatory and manufacturing pathways for one-patient medicines do not yet exist.

## Limitations

The four accessible substitutions are the transitions — C to T, G to A, A to G, T to C. The eight transversions require different chemistry. Glycosylase-based editors that produce C-to-G changes exist but are less efficient and less clean, and for most transversions [[prime-editing]] remains the practical option.

Targeting is still constrained by the PAM requirement inherited from Cas9, which limits how the editing window can be positioned over a given base. Cargo size is a problem for viral delivery: a base editor plus guide RNA exceeds the packaging capacity of a single [[aav-vectors|adeno-associated virus]], forcing split-vector designs. And systemic delivery beyond liver and haematopoietic cells remains largely unsolved, which is the same constraint that limits [[targeted-drug-delivery]] generally.

## Risks

Base editors avoid double-strand breaks but introduce a distinct failure mode: deaminase activity that does not depend on the guide RNA at all. Cytosine editors have been shown to cause genome-wide off-target single-nucleotide changes in mouse embryos and rice through Cas9-independent deamination of transiently exposed single-stranded DNA.[^zuo2019] Both editor classes can also edit cellular RNA, since the deaminases do not distinguish substrates as sharply as intended.[^grunewald2019] Engineered variants reduce both effects but do not abolish them, and detecting scattered single-base changes is harder than detecting the indels left by [[crispr-off-target-effects|nuclease off-target activity]], because there is no cut site to look near.

Bystander edits raise a subtler issue for [[somatic-gene-therapy]]: a silent bystander change in one cell type may not be silent in another, and the tissue is a mosaic of edit patterns rather than a uniform genotype.

For heritable use, the apparent tidiness of base editing has been used as an argument that [[germline-editing]] is now safe enough to attempt. Most of the field rejects that inference. Embryo editing still produces mosaicism, the Cas9-independent off-target mechanism operates during the exact developmental window in question, and no whole-genome assay can certify the absence of scattered point mutations in a single embryo without destroying it. The regulatory posture that followed the [[he-jiankui-affair]] has not shifted, and the frameworks surveyed in [[governance-of-genome-editing]] treat editor chemistry as largely beside the point.

## Outlook

The near-term question is whether permanent single-dose edits to common risk genes — a lifetime cholesterol reduction from one infusion, for instance — can clear a safety bar set for chronic disease rather than for fatal rare disease. That bar is high, because the comparator is a cheap daily pill and the edit is irreversible. The technical frontier is delivery to tissues other than liver, editors small enough for single-vector packaging, and reliable assays for scattered off-target deamination. The economic frontier, shared with [[gene-therapy-for-aging]] and every other one-time genetic medicine, is whether a treatment designed for one child can ever be paid for, a problem that runs directly into [[access-and-inequality]] and that no current reimbursement system is built to solve.

## See also

- [[prime-editing]]
- [[crispr-cas9]]
- [[epigenome-editing]]
- [[crispr-off-target-effects]]
- [[lipid-nanoparticles]]
- [[casgevy]]
- [[somatic-gene-therapy]]
- [[germline-editing]]

## References

[^komor2016]: `paper` Komor, A. C., Kim, Y. B., Packer, M. S., Zuris, J. A., Liu, D. R. "Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage." *Nature*, 2016.
[^gaudelli2017]: `paper` Gaudelli, N. M. et al. "Programmable base editing of A•T to G•C in genomic DNA without DNA cleavage." *Nature*, 2017.
[^chiesa2023]: `paper` Chiesa, R. et al. "Base-edited CAR7 T cells for relapsed T-cell acute lymphoblastic leukemia." *New England Journal of Medicine*, 2023.
[^zuo2019]: `paper` Zuo, E. et al. "Cytosine base editor generates substantial off-target single-nucleotide variants in mouse embryos." *Science*, 2019. {The method compared edited and unedited cells descended from the same embryo, which is why it detected changes that sequencing predicted sites would miss.}
[^grunewald2019]: `paper` Grünewald, J. et al. "Transcriptome-wide off-target RNA editing induced by CRISPR-guided DNA base editors." *Nature*, 2019.
[^musunuru2025]: `paper` Musunuru, K. et al. "Patient-specific in vivo gene editing to treat a rare genetic disease." *New England Journal of Medicine*, 2025. {A report of one infant treated under an individualised protocol, not a trial; it establishes feasibility rather than efficacy.}

==============================================================================
ARTICLE: bioconservatism
TITLE:   Bioconservatism
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/bioconservatism
SOURCE:  https://futurehumanwiki.com/raw/bioconservatism
==============================================================================

---
title: "Bioconservatism"
slug: "bioconservatism"
type: "concept"
status: "established"
horizon: "present"
categories: ["society"]
tags: ["bioethics", "philosophy", "precaution", "human dignity", "politics"]
summary: "The position that biotechnological modification of human beings should face a strong presumption against it, grounded in dignity, precaution, or the value of the given."
updated: "2026-07-27"
issues: ["The stem-cell-clinic harms and the 2019 plasma warning are stated without sources."]
---

```infobox
{
  "caption": "Position in bioethics and politics",
  "rows": [
    { "label": "Core claim", "value": "Presumption against modification" },
    { "label": "Term originated", "value": "Early 2000s" },
    { "label": "Principal authors", "value": "Kass, Fukuyama, Habermas, McKibben" },
    { "label": "Institutional peak", "value": "US President's Council, 2001–2009" },
    { "label": "Political alignment", "value": "Right and left strands" },
    { "label": "Chief opponent", "value": "Transhumanism", "link": "/wiki/transhumanism" }
  ]
}
```

**Bioconservatism** is the view that deliberate biotechnological alteration of human nature should meet a strong presumption against it, and that some such alterations should be refused outright rather than regulated. The label was coined in transhumanist circles in the early 2000s and is used mostly by opponents of the position; few of the writers it describes adopt it. What unites them is not a shared politics but a shared structure of argument: that the burden of justification rests on those who would change the human constitution, not on those who would leave it alone.

## The core arguments

Four distinct arguments run under the heading, and they are frequently confused.

**Human dignity.** Francis Fukuyama's version holds that equal moral and political rights rest on a complex of species characteristics — he calls it Factor X — which is not reducible to any single trait and which would be fragmented by selective modification.[^fukuyama2002] He named transhumanism the world's most dangerous idea in a 2004 magazine symposium, on the ground that a society containing beings of engineered and unengineered kinds would have no stable basis for treating its members as equals.

**The given and the made.** Jürgen Habermas argues that a person who learns their dispositions were chosen by another to a design stands in an asymmetric relation to that designer which no naturally born person occupies, and that the symmetry of mutual recognition among free and equal persons depends on nobody having authored anybody.[^habermas2003] Michael Sandel's related case against perfection is developed in [[bioethics-of-enhancement]].

**Precaution about complex systems.** Bill McKibben's *Enough* makes an environmentalist argument: interventions in systems that are not understood have a poor track record, human biology is such a system, and germline changes are irreversible in a way that ordinary technological mistakes are not.[^mckibben2003] This strand is the closest to the [[precautionary-principle]] as it operates in environmental law.

**Repugnance as evidence.** Leon Kass argued that visceral revulsion at practices like [[human-cloning]] may register a moral truth that argument has not yet articulated. The claim is the most widely criticised in the whole literature — historical repugnance has tracked prejudice at least as often as insight — and Kass's more substantial contributions, particularly the argument that mortality structures the meaning of a human life, do not depend on it. [[leon-kass]] treats his position in detail.

## Strands

Bioconservatism is not a single political tendency, and treating it as the religious right's position on biotechnology misdescribes it.

The **religious** strand is the oldest and the most institutionally organised. Catholic teaching, set out in the 1987 instruction *Donum Vitae* and the 2008 instruction *Dignitas Personae*, opposes embryo destruction, most assisted reproduction, and heritable modification on grounds of the embryo's moral status. It is a specific and internally consistent position that generalises poorly to enhancement questions not involving embryos.

The **secular-philosophical** strand — Kass, Fukuyama, Habermas, Sandel — argues from dignity, meaning, and social structure rather than from religious premises, though its critics dispute how much the arguments retain when the premises are removed.

The **left and feminist** strand opposes the same technologies for opposite reasons. Feminist critics of reproductive technology in the 1980s argued that it medicalised women's bodies and served the interests of clinics rather than patients. The Center for Genetics and Society, founded in 2001, opposes heritable modification and [[polygenic-embryo-screening]] from an explicitly progressive position centred on equality and the history of eugenics rather than on the sanctity of nature. Disability-rights scholarship supplies a further line, developed in [[disability-rights-and-enhancement]], which objects to selection against traits its authors regard as forms of human variation rather than defects.

> [!note] Not the same as opposing biotechnology
> Almost every writer in this tradition supports therapeutic medicine, including [[somatic-gene-therapy]]. The dispute concerns modification of the germline, enhancement beyond species-typical function, and technologies that alter reproduction itself. Framing it as anti-science makes the actual disagreement invisible.

## Institutional influence

The position's high-water mark was the US President's Council on Bioethics, chaired by Kass from 2001 and continuing to 2009, whose 2003 report *Beyond Therapy* remains the fullest official statement of the case against enhancement.[^ptc2003] The Council was unusual among bioethics bodies in commissioning literary and philosophical material alongside technical review, and unusual in the criticism it drew: two members were not reappointed in 2004 after publicly dissenting from its stem-cell positions, prompting accusations that it had been assembled to reach conclusions. It was dissolved in 2009 and replaced by a body with a narrower, more procedural remit.

In law, bioconservative reasoning is visible in the German Embryo Protection Act, in the Council of Europe's Oviedo Convention prohibition on heritable modification, and in the US restrictions described in [[governance-of-genome-editing]]. In practice these instruments were mostly drafted in response to cloning and embryo research rather than to enhancement, and they constrain what may be done to embryos far more tightly than what may be done to adults.

## Criticism

The standard objection is status-quo bias. Nick Bostrom and Toby Ord proposed a diagnostic: if a proposal to increase some parameter is judged bad, ask whether decreasing it would also be judged bad. If both directions are bad, either the current value happens to be optimal — which requires an argument — or the judgment is tracking the fact of change rather than its content.[^bo2006] Applied to memory, disease resistance, or lifespan, the test is uncomfortable for anyone who would refuse an increase while also refusing a decrease. [[nick-bostrom]] and the broader movement described in [[transhumanism]] treat it as decisive; bioconservatives reply that it assumes traits are separable parameters rather than components of an integrated form of life, which is precisely the point at issue.

A second criticism is that appeals to nature treat evolution as if it were a designer with good judgment. Evolution optimises for reproductive fitness in ancestral conditions, not for wellbeing, and the resulting body is full of arrangements no engineer would defend. A third is empirical: the technologies most often invoked as imminent have repeatedly failed to arrive, so that decades of argument have been conducted about capabilities that remain out of reach.

## Where the position has been vindicated

Bioconservatism has a better predictive record than its opponents usually concede, and the vindications are specific rather than general.

The [[he-jiankui-affair]] confirmed that professional self-regulation of the kind inherited from the [[asilomar-conference]] does not bind a determined actor, and that the first heritable edits would be performed without medical justification, without adequate consent, and without the technical safeguards the field had described as prerequisites. Unregulated stem-cell clinics have caused blindness and death while operating in regulatory gaps. Clinics selling young-plasma infusions on the strength of mouse experiments prompted an explicit US regulatory warning in 2019; [[parabiosis-and-young-blood]] describes what the underlying science does and does not show. The tracheal-transplant scandal recounted in [[decellularized-scaffolds]] showed how far ahead of evidence a celebrated surgeon could operate before institutions responded.

None of these vindicates the dignity argument or the repugnance argument. What they vindicate is the narrower claim that enthusiasm systematically outruns evidence in this field, and that the actors most confident about timelines have generally been wrong. Whether that supports a presumption against modification or merely a demand for better regulation is the question the two camps have never resolved, and it is not obvious that any amount of further argument would settle it — the disagreement appears to rest on prior judgments about how much risk a society should accept in exchange for capability it cannot yet evaluate.

## See also

- [[bioethics-of-enhancement]]
- [[leon-kass]]
- [[transhumanism]]
- [[precautionary-principle]]
- [[germline-editing]]
- [[human-enhancement]]
- [[disability-rights-and-enhancement]]
- [[posthuman]]

## References

[^fukuyama2002]: `book` Fukuyama, F. *Our Posthuman Future: Consequences of the Biotechnology Revolution*. Farrar, Straus and Giroux, 2002.
[^habermas2003]: `book` Habermas, J. *The Future of Human Nature*. Polity Press, 2003.
[^mckibben2003]: `book` McKibben, B. *Enough: Staying Human in an Engineered Age*. Times Books, 2003.
[^ptc2003]: `report` President's Council on Bioethics. *Beyond Therapy: Biotechnology and the Pursuit of Happiness*. Washington, DC, 2003. {An advisory report to a US president, written by appointees of the administration that convened the council, not by a scientific consensus body.}
[^bo2006]: `paper` Bostrom, N. and Ord, T. "The Reversal Test: Eliminating Status Quo Bias in Applied Ethics." *Ethics*, 2006. {The test diagnoses status-quo bias in a judgment; it does not by itself show that any particular enhancement would be beneficial.}

==============================================================================
ARTICLE: bioethics-of-enhancement
TITLE:   Bioethics of enhancement
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/bioethics-of-enhancement
SOURCE:  https://futurehumanwiki.com/raw/bioethics-of-enhancement
==============================================================================

---
title: "Bioethics of enhancement"
slug: "bioethics-of-enhancement"
type: "concept"
status: "established"
horizon: "present"
categories: ["society", "enhancement"]
tags: ["bioethics", "enhancement", "philosophy", "autonomy", "justice", "therapy"]
summary: "The philosophical and policy debate over whether biomedical means may legitimately be used to push human capacities beyond species-typical function."
updated: "2026-07-27"
issues: ["Needs a source for the 2003 US approval of growth hormone for idiopathic short stature."]
---

```infobox
{
  "caption": "Field of applied ethics",
  "rows": [
    { "label": "Central question", "value": "Therapy versus enhancement" },
    { "label": "Formative period", "value": "1998–2011" },
    { "label": "Key critics", "value": "Kass, Sandel, Habermas", "link": "/wiki/bioconservatism" },
    { "label": "Key proponents", "value": "Harris, Savulescu, Bostrom" },
    { "label": "Canonical report", "value": "Beyond Therapy, 2003" },
    { "label": "Status", "value": "Unresolved; largely pre-empirical" }
  ]
}
```

**Bioethics of enhancement** is the branch of applied ethics that asks whether, and under what conditions, biomedical interventions may be used to raise human capacities above what is typical for the species rather than to restore what disease has taken away. The debate acquired its modern shape around 2000, when [[crispr-cas9|genome editing]], psychopharmacology, and embryo screening made the question look imminent rather than literary. Two decades later the philosophical positions are well developed and the technologies mostly are not, which gives the literature an unusual character: it is a mature argument about capabilities that largely do not yet exist.

## The line and why it moves

The organising distinction is between therapy, which restores or maintains normal function, and enhancement, which improves on it. The line does real institutional work. Insurers reimburse therapy; medical licensing is justified by therapeutic purpose; most genome-editing statutes permit intervention only for "serious disease" without defining the term.

It also fails under examination. Vaccination does not restore function but adds immunity a person never had. Contraception disables a working system. Caffeine, corrective lenses, and literacy all improve on the unaided baseline. Growth hormone was approved in the United States in 2003 for idiopathic short stature, a condition defined by height alone rather than by any identified pathology — the same molecule, the same effect on the same trait, classified as therapy or enhancement depending on whether a cause was found. The same ambiguity now operates at scale with [[glp-1-receptor-agonists|GLP-1 weight-loss drugs]], approved for obesity and used cosmetically by people at or near typical weight.

Norman Daniels offered the most influential defence of the line: what justice requires medicine to provide is defined by species-typical normal functioning, because departures from it restrict the range of life plans a person can pursue.[^daniels2000] The distinction, on this account, is not about what is natural but about what a society owes. Critics respond that the boundary of the normal is itself statistical and shifting, and that an argument about the limits of entitlement has been repurposed as an argument about the limits of permission. [[human-enhancement]] surveys the modalities; this article concerns the arguments about them.

> [!note] Terminology
> "Enhancement" is used in at least three incompatible senses in the literature: any improvement on an individual's prior state, any improvement beyond the species-typical range, and any improvement not justified by a medical indication. Arguments frequently trade on the ambiguity.

## The case against

Leon Kass, chairing the US President's Council on Bioethics from 2001, framed the objection in terms of ends rather than risks. The Council's 2003 report *Beyond Therapy* accepted that enhancement technologies might work as advertised and argued that this was the problem: better memory, better mood, and longer life pursued as consumer goods would corrode the practices that give those capacities meaning.[^ptc2003] Kass had earlier defended the evidential weight of moral disgust — the "wisdom of repugnance" — as a signal that something deep is at stake even when argument cannot articulate it. That move is widely regarded as the weakest part of the case; repugnance has a poor historical record.

Michael Sandel's version is more durable. In *The Case Against Perfection* he argues that the drive to enhance expresses a refusal of "giftedness" — the recognition that a person's talents are not wholly their own doing — and that three social goods depend on that recognition: humility, limits on responsibility, and solidarity.[^sandel2007] If height, temperament, and intelligence become choices, they become someone's fault. Insurance pools rest on shared exposure to luck; remove the luck and the case for pooling weakens. This is an argument about social structure, not about nature, and it does not depend on any claim that the given is good.

Jürgen Habermas adds a claim about moral symmetry: a person who knows their dispositions were selected by their parents to a design stands in an asymmetric relation to those parents that no naturally born person occupies, and this may undermine their standing as an equal author of their own life.[^habermas2003] Francis Fukuyama argues that human rights rest on a complex of shared species characteristics he calls Factor X, and that fragmenting it fragments the basis of equal treatment.[^fukuyama2002] These positions are collected under [[bioconservatism]].

## The case for

John Harris makes the simplest argument: there is no morally relevant difference between preventing harm and conferring benefit. If it is good to cure a child's poor eyesight, it is good to prevent it, and there is no principled point at which improving vision becomes wrong.[^harris2007] Julian Savulescu extends this to reproduction as [[procreative-beneficence]], the claim that prospective parents have reason to select the child expected to have the best life.

Nick Bostrom and Anders Sandberg supply the most useful methodological contribution on the pro-enhancement side, and it cuts against naive optimism. Their evolutionary optimality challenge asks: if this modification is so beneficial, why has evolution not made it? A proposed enhancement is credible only if the proponent can answer — because the environment has changed, because evolution optimises for reproductive fitness rather than wellbeing, or because a developmental constraint blocked the path.[^bs2009] Applied honestly, the heuristic rules out most speculative enhancements and supports a few, which is what a good heuristic should do.

Allen Buchanan's *Beyond Humanity?* attacks both camps: the bioconservative appeal to nature treats an unplanned evolutionary process as if it were a designer with good judgment, while enthusiasts underrate how much of human capacity is already the product of institutions rather than biology.[^buchanan2011] Much of what is claimed for cognitive enhancement was delivered historically by literacy, nutrition, and public health — the position developed under [[intelligence-amplification]].

## Authenticity and the self

A distinct strand asks whether a chemically or genetically improved trait is really the person's own. Carl Elliott catalogued the unease of patients who did well on antidepressants and felt they had cheated.[^elliott2003] Erik Parens reframed the disagreement as one between two ethical vocabularies — gratitude for what is given, and creativity in remaking it — each internally coherent and neither obviously prior.

The counterargument is that authenticity is not the same as origin. A person who chooses a trait after reflection may own it more fully than one who inherited it. The dispute recurs in every domain the wiki covers: in [[deep-brain-stimulation]], where a minority of patients report changes in personality or in the sense of agency; in [[moral-enhancement]], where the thing modified is the faculty that would evaluate the modification; and in [[personal-identity-and-continuity]], where the question becomes whether the enhanced person is the same person at all.

## Fairness, coercion, and access

The distributive objection is that enhancement bought at market prices entrenches advantage across generations, converting income inequality into biological inequality. It is the strongest argument in practice and the one most often answered with a promise that costs will fall. Prices for licensed cell and gene therapies have not fallen; several have been withdrawn from markets that refused to pay. [[access-and-inequality]] examines the record.

The coercion objection is structurally different and does not require anyone to be excluded. Where an enhancement confers positional advantage, its availability changes the choice set of people who do not want it: surgeons whose colleagues take modafinil, students whose peers do, athletes competing against the pharmacologically prepared. Nobody is forced and everybody's options worsen. This is the subject of [[enhancement-arms-race]], and it is the mechanism the anti-doping regime described in [[enhancement-in-sport]] exists to interrupt.

A third objection comes from disability scholarship: enhancement raises the baseline against which impairment is measured, so that traits currently unremarkable become deficits. [[disability-rights-and-enhancement]] treats this and the related expressivist objection to selection.

> [!debate] The disagreement in one line
> Critics argue that enhancement changes what it means to be human and should therefore face a presumption against it; proponents reply that the presumption smuggles in the assumption that the current human is well designed, which nobody defends when asked directly.

## The empirical gap

The literature's most striking feature is how little of it is constrained by evidence. Cognitive pharmacology has produced no agent with a reliable effect on healthy adults beyond wakefulness; the evidence base is reviewed in [[nootropics]]. Transcranial direct-current stimulation, for a decade the leading non-invasive enhancement candidate, has a poor replication record, as [[non-invasive-neuromodulation]] describes. [[polygenic-embryo-screening]] delivers expected gains that are real but small and poorly characterised outside the ancestry groups its predictors were trained on. [[genetic-enhancement-of-intelligence]] is limited by the fact that the trait is spread across thousands of variants of tiny effect, each with unknown pleiotropic consequences.

Two of the field's founding worries have meanwhile been tested. The [[he-jiankui-affair]] showed what happens when a researcher acts outside professional constraint — not a slippery slope but a single episode, prosecuted, with no imitators for years afterward. And the enhancement that did arrive at scale arrived through [[embryo-selection|selection]] rather than editing, which almost no pre-2010 framework anticipated and which most regulatory instruments still do not reach.

## Open problems

Three questions remain genuinely unsettled rather than merely contested.

**Who decides what counts as better.** Enhancement presupposes a direction of improvement. For visual acuity or infection resistance the direction is uncontroversial. For temperament, risk tolerance, or empathy it is not, and the people choosing are usually parents choosing for someone who does not exist yet and cannot object — the structure examined under [[designer-babies]].

**Whether the therapy line can be replaced.** If it cannot be defended philosophically but is doing necessary institutional work, the alternative is a substantive account of which capacities societies should underwrite and which they should merely permit. No such account commands agreement, and regulators continue to use a distinction that most ethicists on both sides regard as incoherent.

**Whether the debate has been about the right technologies.** Twenty-five years of argument concentrated on germline modification and cognitive drugs. The interventions that have actually changed human capability in that period are digital rather than biological, and the closest thing to a live enhancement question in 2026 concerns access to and dependence on external cognitive systems — the territory of [[human-ai-merger]], which the enhancement literature has barely begun to metabolise.

## See also

- [[human-enhancement]]
- [[bioconservatism]]
- [[procreative-beneficence]]
- [[disability-rights-and-enhancement]]
- [[enhancement-arms-race]]
- [[moral-enhancement]]
- [[leon-kass]]
- [[julian-savulescu]]

## References

[^daniels2000]: `paper` Daniels, N. "Normal Functioning and the Treatment-Enhancement Distinction." *Cambridge Quarterly of Healthcare Ethics*, 2000.
[^ptc2003]: `report` President's Council on Bioethics. *Beyond Therapy: Biotechnology and the Pursuit of Happiness*. Washington, DC, 2003. {An advisory report to a US president, written by appointees of the administration that convened the council, not by a scientific consensus body.}
[^sandel2007]: `book` Sandel, M. *The Case Against Perfection: Ethics in the Age of Genetic Engineering*. Harvard University Press, 2007.
[^habermas2003]: `book` Habermas, J. *The Future of Human Nature*. Polity Press, 2003.
[^fukuyama2002]: `book` Fukuyama, F. *Our Posthuman Future: Consequences of the Biotechnology Revolution*. Farrar, Straus and Giroux, 2002.
[^harris2007]: `book` Harris, J. *Enhancing Evolution: The Ethical Case for Making Better People*. Princeton University Press, 2007.
[^bs2009]: `book` Bostrom, N. and Sandberg, A. "The Wisdom of Nature: An Evolutionary Heuristic for Human Enhancement." In *Human Enhancement*, Oxford University Press, 2009. {A book chapter arguing a heuristic from evolutionary reasoning rather than reporting experiments; it concerns what is plausible, not what has been shown to work.}
[^buchanan2011]: `book` Buchanan, A. *Beyond Humanity? The Ethics of Biomedical Enhancement*. Oxford University Press, 2011.
[^elliott2003]: `book` Elliott, C. *Better Than Well: American Medicine Meets the American Dream*. W. W. Norton, 2003. {Reporting and cultural criticism drawn from interviews and clinical anecdote, not a study of outcomes.}

==============================================================================
ARTICLE: biohacking
TITLE:   Biohacking and grinders
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/biohacking
SOURCE:  https://futurehumanwiki.com/raw/biohacking
==============================================================================

---
title: "Biohacking and grinders"
slug: "biohacking"
type: "concept"
status: "contested"
horizon: "present"
categories: ["enhancement", "people"]
tags: ["diy", "implants", "self-experimentation", "subculture", "regulation", "citizen science"]
summary: "The practice of modifying one's own biology outside institutional medicine, ranging from subdermal implants to do-it-yourself gene transfer and open-source medical devices."
updated: "2026-07-28"
---

```infobox
{
  "caption": "Subculture and practice",
  "rows": [
    { "label": "Emerged", "value": "c. 2005–2010" },
    { "label": "Main strands", "value": "Grinders, DIYbio, quantified self" },
    { "label": "Typical interventions", "value": "Magnets, RFID, supplements" },
    { "label": "Oversight", "value": "None; no adverse-event reporting" },
    { "label": "Notable figures", "value": "Amal Graafstra, Josiah Zayner" },
    { "label": "Legal status", "value": "Mostly lawful; sale often is not" }
  ]
}
```

**Biohacking and grinders** are, respectively, a loose label for do-it-yourself modification of one's own biology and the name the implant subculture within it uses for itself. The term covers activities with almost nothing in common: inserting a magnet in a fingertip, running an open-source insulin pump, taking a stack of supplements, and injecting an unapproved gene-transfer construct. What unites them is the deliberate bypassing of the institutions — clinical trials, prescribing physicians, regulators — that normally stand between a person and an intervention on their own body.

## Strands

**Grinders** implant hardware subdermally. The canonical devices are neodymium magnets in the fingertip, which transduce alternating magnetic fields into a felt vibration and function as a crude added sense of the kind discussed in [[sensory-augmentation]], and glass-encapsulated RFID or NFC transponders in the web of the hand, used to open doors and pair with phones. Amal Graafstra implanted an RFID tag in 2005 and later founded a company supplying implant hardware; several thousand people, concentrated in Sweden and the United States, are thought to carry NFC implants, though no register exists. The word "grinder" was borrowed from a comic and adopted by the community around 2010.

**DIY biology** works with molecules rather than hardware. Community laboratories such as Genspace in New York and BioCurious in California, both founded around the turn of the 2010s, offer bench space and training to non-professionals, and the DIYbio network that formed in 2008 gave the movement a shared code of conduct.[^ledford2010] Most of the work is unremarkable molecular biology. A small fraction is self-experimentation.

**Open-source medicine** builds tools that industry has not. The most substantial example is do-it-yourself automated insulin delivery, in which people with type 1 diabetes reverse-engineered pump communication protocols and wrote closed-loop control algorithms years before commercial systems were approved, under the slogan "we are not waiting". An open-source system of this kind was subsequently tested in a randomised controlled trial and found safe and effective.[^burnside2022] This is the single clearest case of a biohacking project producing validated clinical benefit.

**Lifestyle biohacking** is the largest strand by participation and the loosest by definition: supplement regimens, [[continuous-glucose-monitoring|continuous glucose monitors]] worn by people without diabetes, [[heat-and-cold-exposure|cold exposure and sauna use]], sleep tracking, and expensive self-quantification programmes. It shades into the [[nootropics]] market on one side and into the [[quantified-self|quantified-self movement]], named by two *Wired* editors in 2007, on the other.[^wolf2010] Consumer tests reporting a [[biological-age]] have become the strand's characteristic instrument, despite measuring something whose relation to health outcomes is unsettled.

```timeline
[
  { "year": "2005", "title": "First hobbyist RFID implant", "text": "Amal Graafstra implants a radio-frequency tag in his hand to open doors, establishing the template for hardware implants outside medicine." },
  { "year": "2007–2008", "title": "Quantified self and DIYbio", "text": "The quantified-self label is coined and the DIYbio network forms, giving amateur biology a shared identity and a code of conduct." },
  { "year": "2013", "title": "Circadia", "text": "Grindhouse Wetware implants a Bluetooth-enabled temperature sensor in a member's forearm, installed by a body-modification artist rather than a physician." },
  { "year": "2015", "title": "Open-source insulin delivery spreads", "text": "Do-it-yourself closed-loop systems built from reverse-engineered pumps reach a substantial user base well ahead of approved equivalents." },
  { "year": "2017", "title": "Self-administered gene transfer", "text": "Josiah Zayner injects himself with a CRISPR construct targeting myostatin on stage at a synthetic-biology conference; no effect on muscle is reported and he later describes the act as a mistake." },
  { "year": "2018", "title": "A death and a reckoning", "text": "Aaron Traywick, who had self-administered an experimental herpes treatment in public, is found dead in a flotation tank; the episode ends the brief vogue for stunt self-dosing." },
  { "year": "2019–2020", "title": "First targeted law", "text": "California requires gene-therapy kits sold to consumers to carry a notice that they are not for self-administration, following a federal warning that such sales are unlawful." }
]
```

## The gene-editing episodes

The self-experiments that drew the most attention involved gene transfer. In 2017 Josiah Zayner, a former NASA researcher who founded a company selling molecular-biology kits, injected himself with a [[crispr-cas9]] construct intended to disrupt the myostatin gene and increase muscle mass. No effect was observed. The relevant biology, set out in [[myostatin-inhibition]], makes that unsurprising: an intramuscular injection of naked plasmid reaches a negligible fraction of fibres, and even successful [[somatic-gene-therapy|somatic gene transfer]] against this target has produced mass without proportional strength gains in trials. Zayner has since said the demonstration was irresponsible and that its main effect was to encourage imitation. Anti-doping authorities have treated the same constructs as a prospective cheating method rather than a hobby; see [[gene-doping]].

The United States Food and Drug Administration stated in 2017 that the sale of gene-therapy products intended for self-administration is unlawful.[^fda2017] California followed in 2019 with a statute requiring consumer gene-therapy kits to carry a warning against self-use, the first law written specifically for this practice.[^sb180] Separately, unregulated clinics and companies operating in permissive jurisdictions have supplied follistatin and telomerase constructs to paying individuals, a commercial activity distinct from the hobbyist scene but frequently confused with it; see [[gene-therapy-for-aging]].

> [!caution] What self-experiments can and cannot show
> An uncontrolled intervention on one person, with no blinding and an outcome the person is invested in, cannot establish efficacy. The self-experiment tradition in medicine — self-inoculation, self-catheterisation — produced findings because the effects were unmistakable and immediate. Nothing in the enhancement space has that property, so the epistemic value of these demonstrations is close to zero even when the participant is honest.

## Safety

Implant risks are those of any foreign body inserted without sterile technique: local infection, migration, extrusion, and allergic response to coating materials. Magnets whose parylene or silicone coating fractures cause local necrosis and must be removed. Implanted electronics complicate magnetic resonance imaging and are difficult to explant intact. Installation is usually performed by body-modification practitioners, who in most jurisdictions may not administer local anaesthetic, so procedures are done without it.

Molecular self-experiments carry the risks catalogued in clinical gene therapy with none of its safeguards: immune reaction to an [[aav-vectors|adeno-associated viral vector]], uncontrolled expression, contamination of preparations made outside a controlled facility, and the absence of anyone monitoring the outcome. The community's own harm-reduction norms — sterile technique, avoiding viral vectors, keeping a witness present — are unenforceable.

Because nothing is reported to any authority, the actual complication rate across the whole practice is unknown. That is itself the most reliable statement available about its safety.

## Law, governance, and the biosecurity question

Modifying one's own body is generally lawful. Selling something for that purpose usually is not, and practising medicine without a licence certainly is not; enforcement has concentrated on the sellers rather than the users. Consent law imposes an outer limit that surprises many participants: in England and Wales, the Court of Appeal has held that consent is not a defence to charges arising from serious body modification performed by a non-medical practitioner, a ruling discussed further in [[morphological-freedom]].

The biosecurity concern attached to community laboratories has been examined repeatedly and has consistently been judged low. Amateur laboratories lack the tacit skill and the equipment for anything dangerous, and the community has cooperated with law enforcement since the late 2000s. Analysts of citizen science have argued that the more realistic policy problem is not amateur bioterror but an absence of any route by which useful amateur work — the insulin systems, for instance — can be evaluated and adopted.[^guerrini2018] The genuinely serious version of the biosecurity question concerns commercial DNA synthesis and design tools rather than garages, and is treated in [[dual-use-research]].

## Criticism and standing

Mainstream bioethics treats the movement's central claim — that individuals should be able to accept risk on their own behalf — as reasonable in principle and badly executed in practice. It is the same claim that organises the wider debate over [[human-enhancement]], stripped of the institutional framing that debate usually assumes, and it is assessed in [[bioethics-of-enhancement]]. The objection is not paternalism but that self-experimenters generate no usable knowledge, expose themselves to harms they cannot assess, and, when they perform for cameras, recruit others with less understanding. The disability-rights critique in [[disability-rights-and-enhancement]] adds that the subculture's framing of the body as a platform to be upgraded sits awkwardly with the experience of people whose bodies are already objects of medical intervention they did not choose.

Within the [[transhumanism|transhumanist]] tradition the movement occupies an ambiguous position. It is the only strand that acts rather than argues, and the actions have produced one validated medical technology and a series of failed stunts. The interesting question is why the ratio is that way: the insulin work succeeded because it targeted a well-characterised control problem with an immediate, measurable endpoint, and every failed enhancement self-experiment has targeted a diffuse outcome measurable only over years. Whether any enhancement goal has the first shape is unclear, and if none does, the practice has a ceiling that no amount of nerve will raise.

## See also

- [[human-enhancement]]
- [[morphological-freedom]]
- [[sensory-augmentation]]
- [[gene-therapy-for-aging]]
- [[nootropics]]
- [[somatic-gene-therapy]]
- [[dual-use-research]]
- [[transhumanism]]

## References

[^ledford2010]: `news` Ledford, H. "Garage biotech: Life hackers." *Nature* news feature, 2010.
[^wolf2010]: `news` Wolf, G. "The Data-Driven Life." *The New York Times Magazine*, 2010. {Wolf is one of the two editors who named the quantified-self movement, so the piece is a participant's account of it.}
[^burnside2022]: `paper` Burnside, M. J. et al. "Open-Source Automated Insulin Delivery in Type 1 Diabetes." *New England Journal of Medicine*, 2022. {A randomised trial in New Zealand in children and adults with type 1 diabetes, testing a community-built system under clinical supervision.}
[^fda2017]: `regulator` U.S. Food and Drug Administration. "Information About Self-Administration of Gene Therapy." Public statement, 2017.
[^sb180]: `law` California Senate Bill 180 (2019), requiring consumer notice on gene-therapy kits sold in California. {The statute requires a warning notice at the point of sale; it does not prohibit selling the kits or using them on oneself.}
[^guerrini2018]: `paper` Guerrini, C. J., Majumder, M. A., Lewellyn, M. J. and McGuire, A. L. "Citizen science, public policy." *Science*, 2018.

==============================================================================
ARTICLE: biological-age
TITLE:   Biological age
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/biological-age
SOURCE:  https://futurehumanwiki.com/raw/biological-age
==============================================================================

---
title: "Biological age"
slug: "biological-age"
type: "concept"
status: "contested"
horizon: "present"
categories: ["longevity"]
tags: ["aging", "biomarkers", "measurement", "epigenetics", "frailty"]
summary: "An estimate of how far an individual has progressed along the trajectory of age-related decline, expressed in years and compared against their chronological age."
updated: "2026-07-28"
humanEvidence: "The measures are built and tested in human cohorts, where several predict mortality; none is accepted by a regulator as a surrogate endpoint, and different measures in the same person agree only weakly."
issues: ["The reversal-claims section cites none of the studies whose pattern it describes."]
---

```infobox
{
  "caption": "Concept in gerontology",
  "rows": [
    { "label": "Field", "value": "Biogerontology, epidemiology" },
    { "label": "Contrasted with", "value": "Chronological age" },
    { "label": "Main measure families", "value": "Methylation, clinical, functional" },
    { "label": "Gold standard", "value": "None established" },
    { "label": "Consumer availability", "value": "Widespread" },
    { "label": "Status", "value": "Definitionally contested" }
  ]
}
```

**Biological age** is an estimate of where a person sits on the trajectory of age-related functional decline, expressed on the scale of years so that it can be compared with the date on their birth certificate. Two seventy-year-olds can differ enormously in organ function, disease burden and remaining life expectancy, and the concept exists to name that difference. It is used constantly in aging research and has no agreed definition, no gold-standard measurement, and no accepted regulatory status.

```keyfacts
[
  { "value": "1969", "label": "Comfort's call for an aging test battery", "note": "still unmet by any single accepted measure" },
  { "value": "Weak", "label": "Agreement between measures", "note": "different biological-age metrics in the same person correlate only loosely" },
  { "value": "None", "label": "Measures accepted as surrogate endpoints", "note": "no regulator recognises one for aging" }
]
```

## What the concept is trying to capture

Chronological age is the best single predictor of death and of most chronic disease, and it is measured perfectly. The reason to want something else is that it is uninformative about individuals: it cannot distinguish a robust eighty-year-old from a frail one, and it cannot respond to an intervention. A measure that could do both would let researchers test geroprotectors in years rather than decades, which is the practical motivation described under [[aging-biomarkers]].

The underlying assumption is that there exists a single latent quantity, the extent of accumulated aging, that many measurable things reflect imperfectly. This assumption is doing a great deal of work and may not be true. If the processes catalogued in the [[hallmarks-of-aging]] proceed at partly independent rates in different tissues, then a person does not have one biological age but many.

## Origins

Gerontologists have sought a functional measure of aging since at least the 1960s. Alex Comfort argued in 1969 for a battery of tests that could establish an individual's rate of aging, and set out roughly what such a battery would need to do.[^comfort1969] Efforts through the following decades combined physiological measures such as vital capacity, blood pressure, reaction time and visual accommodation into composite indices, generally by regressing them against chronological age.

The statistical framework most often used today comes from Klemera and Doubal, who showed that regressing biomarkers on age and then inverting the relationship produces a better estimate than the obvious approach of regressing age on biomarkers.[^klemera2006] A pattern runs through the whole history. Each generation of measures was assembled from whatever assay had just become cheap and high-throughput: physiological test batteries, then routine blood chemistry, then methylation arrays, then mass-spectrometry proteomics. The selection of inputs has been driven by instrument availability rather than by any theory of what aging is, which is one reason the resulting indices agree with each other so poorly.

## Measurement families

**Clinical composites** combine routine laboratory values, blood pressure and spirometry. PhenoAge was derived this way from a national health survey before being ported onto methylation data.

**Methylation clocks** dominate current research, following the multi-tissue predictor [[steve-horvath]] published in 2013. See [[epigenetic-clock]] for the generations, their training targets, and their reliability problems.

**Deficit accumulation** takes a different route entirely. A frailty index counts the proportion of a long list of health deficits that a person has, without weighting them, and predicts mortality and institutionalisation robustly. It requires no laboratory at all.

**Proteomic measures** read plasma protein levels. Work from Tony Wyss-Coray's laboratory used organ-enriched plasma proteins to estimate the age of individual organs, finding that a substantial fraction of people show one organ aging markedly faster than the rest, with disease risk tracking that organ.[^oh2023] **Immune clocks** built from cytometry and cytokine profiles capture a related axis, linked to [[inflammaging]].

**Haematological measures** exploit the fact that blood is the most accessible aging tissue: clonal expansions in the marrow, a marker of [[stem-cell-exhaustion]], become common after middle age and predict cardiovascular as well as haematological outcomes.

**Imaging measures** estimate brain age from MRI and retinal age from fundus photographs. **Functional measures** such as gait speed, grip strength, chair-rise time and VO2max are the oldest approach and among the best validated for predicting outcomes; they are also the endpoints that [[exercise-and-aging|exercise]] moves most reliably.

> [!caution] The measures disagree
> When multiple biological-age metrics are computed on the same cohort, they correlate with each other only weakly, often more weakly than each correlates with chronological age. Belsky and colleagues found this across telomere length, several methylation clocks and clinical composites.[^belsky2018] Either they are measuring different things, or most of them are measuring noise.

## Why disagreement between measures matters

The weak inter-measure agreement admits two readings. On the optimistic reading, aging is multidimensional and each measure captures a genuine, partly independent component, in which case the right response is a panel rather than a number. On the pessimistic reading, most of the variance in these measures is technical and cohort-specific, and the shared signal is small.

Organ-specific results support the multidimensional reading. So does the observation that different measures predict different outcomes: methylation clocks predict mortality, frailty indices predict disability and institutional care, and cardiorespiratory fitness predicts cardiovascular events. Aggregating them into one number discards that structure.

Settings that accelerate aging-like change provide a partial test. Long-duration spaceflight produces bone loss, immune dysregulation and cardiovascular change on a compressed timescale, discussed under [[space-medicine]], and cancer survivors treated with cytotoxic chemotherapy show elevated readings on several measures. If competing metrics agreed about who is aging fast, these are the populations where the agreement should be easiest to see. It is not consistently observed.

## Reversal claims

Interventions are routinely reported to have "reversed biological age by X years". Almost always the claim rests on a single measure, usually a methylation clock, in a small and often uncontrolled study. The pattern recurs across [[caloric-restriction]] trials, supplement studies, [[metformin]] cohorts and early [[senolytics]] work. Three cautions apply. First, the measure has test-retest variation on the order of the reported effect. Second, a measure trained to predict an outcome is not thereby a cause of that outcome, so moving it need not move the outcome. Third, several of the interventions that shift clock readings, notably [[epigenetic-reprogramming]], act directly on the molecular substrate the clock reads, which makes a clock a particularly poor referee for them.

Regression to the mean compounds the problem in [[consumer-blood-testing|consumer testing]]. A person who tests high, changes their habits and retests will frequently see a lower number regardless of what they did.

## Open problems

The field lacks an accepted ground truth. Measures are validated against mortality, which is the outcome they are supposed to predict, so validation is partly circular and cannot distinguish a measure of aging from a measure of current illness. Nothing yet distinguishes measures that track a causal process from measures that track its consequences, which is the same difficulty that dogs [[cellular-senescence]] markers and [[telomeres-and-telomerase|telomere length]].

Standardisation is beginning through consortium work, and prize competitions such as [[xprize-healthspan]] have sidestepped the problem by specifying functional restoration in muscle, cognition and immunity as the endpoint rather than any biomarker. That choice is a judgement about the state of the field: the organisers concluded that no biological-age measure is yet trustworthy enough to award a nine-figure prize on. Whether the [[geroscience-hypothesis|geroscience]] programme can obtain a validated surrogate before it needs one for [[healthspan]]-extension trials is the practical form of the question.

## See also

- [[epigenetic-clock]]
- [[aging-biomarkers]]
- [[hallmarks-of-aging]]
- [[healthspan]]
- [[geroscience-hypothesis]]
- [[compression-of-morbidity]]
- [[maximum-human-lifespan]]

## References

[^comfort1969]: `paper` Comfort, A. "Test-battery to measure ageing-rate in man." *The Lancet*, 1969.
[^klemera2006]: `paper` Klemera, P. & Doubal, S. "A new approach to the concept and computation of biological age." *Mechanisms of Ageing and Development*, 2006.
[^oh2023]: `paper` Oh, H.S.-H. et al. "Organ aging signatures in the plasma proteome track health and disease." *Nature*, 2023. {Organ ages are estimated from plasma proteins enriched in each organ, not from measuring the organs themselves.}
[^belsky2018]: `paper` Belsky, D.W. et al. "Eleven telomere, epigenetic clock, and biomarker-composite quantifications of biological aging: do they measure the same thing?" *American Journal of Epidemiology*, 2018. {Run within a single birth cohort at one age, so it tests agreement between the measures rather than their ability to rank people of different ages.}

==============================================================================
ARTICLE: blue-zones
TITLE:   Blue Zones
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/blue-zones
SOURCE:  https://futurehumanwiki.com/raw/blue-zones
==============================================================================

---
title: "Blue Zones"
slug: "blue-zones"
type: "concept"
status: "contested"
horizon: "present"
categories: ["longevity", "society"]
tags: ["aging", "demography", "diet", "public health", "epidemiology", "centenarians"]
summary: "Five regions claimed to have unusually high concentrations of centenarians, and the lifestyle explanations built on that claim."
updated: "2026-07-27"
humanEvidence: "The regional claims rest on retrospective age records in places with incomplete birth registration; the best-documented case, the Adventist cohort at Loma Linda, enrolled living people with verified ages."
issues: ["The Japanese and Greek pension-audit figures are stated without sources."]
---

```infobox
{
  "caption": "Demographic claim and popular health concept",
  "rows": [
    { "label": "Coined by", "value": "Michel Poulain and Gianni Pes" },
    { "label": "First region", "value": "Sardinia, 2004" },
    { "label": "Popularized by", "value": "Dan Buettner, 2005–2008" },
    { "label": "Regions claimed", "value": "Five" },
    { "label": "Central critique", "value": "Age-record error and pension fraud" },
    { "label": "Status", "value": "Demography disputed; advice conventional" }
  ]
}
```

**Blue Zones** are five regions identified as having exceptional concentrations of people living past 100, together with the set of lifestyle explanations proposed to account for them. The term originated in demographic fieldwork in Sardinia and was popularized in books and television by the writer Dan Buettner. The demographic claim underlying it has come under sustained attack since the late 2010s, chiefly from work showing that regions with poor birth records produce inflated numbers of very old people.

```keyfacts
[
  { "value": "5", "label": "Regions in the canonical list", "note": "Sardinia, Okinawa, Nicoya, Ikaria, Loma Linda" },
  { "value": "2004", "label": "First region identified", "note": "Sardinian AKEA study" },
  { "value": "2024", "label": "Ig Nobel awarded to the critique", "note": "Saul Newman, demography prize" }
]
```

## Origins

The term comes from the work of the Belgian demographer Michel Poulain and the Italian physician Gianni Pes, who mapped centenarian density in Sardinia and found an unusual cluster in the mountainous interior of Nuoro province. They marked the villages in blue ink, and the phrase stuck.[^poulain2004] The Sardinian finding was notable less for the absolute number of centenarians than for the sex ratio: the local excess appeared in men, whereas centenarians almost everywhere are overwhelmingly women.

Dan Buettner extended the concept in a 2005 *National Geographic* cover story and a 2008 book, adding Okinawa in Japan, the Nicoya Peninsula in Costa Rica, Ikaria in Greece, and the Seventh-day Adventist community of Loma Linda, California.[^buettner2008] A 2023 Netflix series brought the framework to a much wider audience. The list has not expanded since; Buettner's organization has periodically discussed candidate regions without formally adding them.

## The claimed common factors

Buettner's synthesis reduces the regions' commonality to nine habits: natural daily movement rather than deliberate exercise, a sense of purpose, routines that reduce stress, stopping eating at partial fullness, a mostly plant-based diet with beans as a staple, moderate alcohol in most of the regions, religious or community belonging, prioritizing family, and social circles that reinforce the rest.

None of these is controversial as health advice. Each is supported by evidence from ordinary epidemiology gathered outside the Blue Zones entirely, and several — particularly [[exercise-and-aging|habitual physical activity]] and moderate energy intake, the human echo of [[caloric-restriction]] — have mechanistic support running through nutrient-sensing pathways, [[autophagy]] and the same mTOR signaling that [[rapamycin]] inhibits pharmacologically. The dispute is not about whether these behaviors are good but about whether they explain a regional longevity excess that may not exist.

It is also worth separating the outcome being claimed. Most of the nine habits have their best-documented effect on [[healthspan]] and on the [[compression-of-morbidity|timing of disability]] rather than on maximum age, and the strongest evidence for any of them concerns cardiovascular and metabolic disease rather than the [[hallmarks-of-aging|underlying processes of aging]].

> [!note] Two different claims
> "This region has many centenarians" and "this region has high life expectancy" are separate statements that can diverge. Centenarian counts depend on the size of the birth cohort a century ago, on out-migration, and on record quality. Several Blue Zones have life expectancy at or below their national average.

## The demographic critique

The most damaging line of attack comes from the demographer Saul Newman, whose analysis of age records across many countries won an Ig Nobel Prize in 2024.[^newman2024] His argument is that regions producing remarkable-age records share a distinctive profile: relatively high poverty, low average life expectancy, incomplete or late birth registration, and in some cases pension systems that reward age misreporting.

Several strands support it.

**Record-keeping is the strongest predictor.** In the United States, the introduction of statewide birth registration was followed by a very large fall — on the order of three-quarters — in the number of supercentenarian records subsequently generated by that state. Documented ages fell when documents improved.

**Audits have repeatedly found phantom old people.** Japan's 2010 review of its centenarian registry found on the order of two hundred thousand people recorded as living at 100 or above who were dead or untraceable. Greek pension audits around 2011 removed large numbers of centenarian beneficiaries found not to exist. Both countries contain Blue Zones.

**Okinawa's health has deteriorated in ways the narrative did not predict.** Okinawan male life expectancy fell sharply in Japan's prefectural rankings around 2000, and the prefecture has since carried some of the country's highest obesity rates. If a durable regional practice were responsible for the earlier longevity, its collapse within one generation requires explanation. The more parsimonious account is that the surviving Okinawan cohort had been shaped by wartime privation and postwar records destroyed in the 1945 fighting.

Defenders respond that the Sardinian data in particular were validated against parish registers, civil records, marriage documents and military rolls, cross-checked independently, and that blanket claims of error do not apply to cohorts verified this way. Newman's central papers circulated for years primarily as preprints, which critics have noted, and the specific accusation of fraud has not been substantiated for every region. The disagreement is unresolved and mostly conducted outside the peer-reviewed literature.

> [!caution] Contested
> The claim that specific regions harbor genuine longevity outliers is disputed by mainstream demography. The claim that beans, walking and social connection are good for health is not disputed by anyone. Coverage of the topic routinely treats support for the second as support for the first.

## What survives

The Loma Linda case is the most defensible, because it is not a geographic claim at all but a cohort study with good records. The Adventist Health Studies followed tens of thousands of Seventh-day Adventists in California with birth documentation and prospective data collection, and reported that Adventist men lived several years longer than other Californian men, with vegetarian diet, non-smoking, exercise and body weight accounting for much of the gap.[^fraser2001] There is no age-verification problem because the subjects were enrolled while alive in a country with reliable registration.

That case, however, undercuts the exotic reading of the concept. What it shows is that ordinary modifiable risk factors produce a substantial life-expectancy difference within a rich country — the same result cardiovascular epidemiology has produced repeatedly. It does not require a special region, a special diet, or an unexplained cluster.

The residual scientific interest lies in whether any Blue Zone population carries genetic or environmental features worth studying, in the way that comparative work on [[negligible-senescence]] examines species with unusual mortality curves. Sardinian genetic isolation is real and has been studied for other traits. Whether it contributes to longevity is unestablished. Centenarian genetics has been pursued more systematically in cohorts assembled outside these regions, among them the Ashkenazi families studied by [[nir-barzilai]]. One of the few findings from that work to replicate across populations, an association between variants near FOXO3 and exceptional survival, points to the insulin and IGF-1 signaling pathway that [[cynthia-kenyon|Cynthia Kenyon's]] worm genetics opened rather than to anything regional. Studies looking for a distinctive [[inflammaging|inflammatory profile]] or a favorable [[cellular-senescence|senescent cell burden]] in these populations have been small and have not converged.

Biological measures could in principle discipline the field. An [[epigenetic-clock]] cannot verify a claimed age of 115 — the error bars on any current [[aging-biomarkers|aging biomarker]] are several years wide — but it can distinguish a claimed 115-year-old who is biologically 90 from one who is not, and applying [[biological-age]] estimates to disputed cohorts would be a cheap test that nobody has systematically run.

## Commercialization

The concept became a business. Blue Zones LLC licenses the brand for community health programmes, works with municipalities on food and built-environment policy, and was acquired by a hospital system in 2020. Certified Blue Zones products and restaurant menus followed.

The programmes themselves are conventional public health: sidewalks, tobacco restrictions, school lunch policy, walking groups. Evaluations have reported improvements in measured health behaviors in participating communities, though the designs are before-and-after comparisons without randomization. The oddity is that a marketing frame built on contested demography has been used to deliver interventions that were already justified without it.

The framing has a defensible side. Population-level environmental change is close to the only longevity intervention that does not raise the [[access-and-inequality]] problem, since sidewalks and food policy reach everyone in a jurisdiction rather than those who can pay. If the economic case set out in the [[longevity-dividend]] literature holds anywhere, it holds most clearly for measures of this kind.

## Outlook

The useful residue of the Blue Zones idea is methodological. It drew attention to the fact that claims about very old people are only as good as the paperwork behind them, and it prompted the audits and validation work that now constrain the study of [[maximum-human-lifespan]]. That is a real contribution, if not the one intended, and it applies equally to the extrapolations behind [[longevity-escape-velocity]] and to any [[geroscience-hypothesis|geroscience]] trial that proposes to use historical longevity data as a comparator.

What remains unanswered is whether any population anywhere has a longevity advantage that is not explained by record error, cohort selection, or the standard risk factors. Settling it requires prospective cohorts with verified birth records rather than retrospective centenarian counts, and those cohorts take a century to mature. Until then, region-level longevity claims should be read as hypotheses about data quality first and biology second.

## See also

- [[maximum-human-lifespan]]
- [[healthspan]]
- [[compression-of-morbidity]]
- [[exercise-and-aging]]
- [[caloric-restriction]]
- [[heat-and-cold-exposure]]
- [[negligible-senescence]]
- [[longevity-dividend]]

## References

[^poulain2004]: `paper` Poulain, M., Pes, G.M., Grasland, C. et al. "Identification of a geographic area characterized by extreme longevity in the Sardinia island: the AKEA study." *Experimental Gerontology*, 2004.
[^buettner2008]: `book` Buettner, D. *The Blue Zones: Lessons for Living Longer from the People Who've Lived the Longest*. National Geographic Books, 2008. {A popular book by a journalist; the nine common factors are the author's synthesis across regions, not a published statistical result.}
[^newman2024]: `preprint` Newman, S.J. "Supercentenarian and remarkable age records exhibit patterns indicative of clerical errors and pension fraud." *bioRxiv* preprint, 2024. {A statistical analysis of patterns across many regions of age records rather than a document audit of any single Blue Zone cohort.}
[^fraser2001]: `paper` Fraser, G.E. and Shavlik, D.J. "Ten Years of Life: Is It a Matter of Choice?" *Archives of Internal Medicine*, 2001.

==============================================================================
ARTICLE: brain-preservation
TITLE:   Brain preservation
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/brain-preservation
SOURCE:  https://futurehumanwiki.com/raw/brain-preservation
==============================================================================

---
title: "Brain preservation"
slug: "brain-preservation"
type: "technology"
status: "experimental"
horizon: "indefinite"
trl: 4
categories: ["minds", "longevity"]
tags: ["cryopreservation", "connectome", "uploading", "neuroscience", "vitrification", "death"]
summary: "Preserving the structural information in a brain well enough that a future technology could reconstruct its memories, rather than revive the tissue itself."
updated: "2026-07-27"
humanEvidence: "Verification is from rabbit and pig brains only; no preserved human brain has been assessed the same way, and no preserved brain of any species has been scanned and shown to yield a working model."
access: "Offered after death by a small number of organizations, at lower cost than liquid-nitrogen cryonics; no provider is known to perform the premortem perfusion the published protocol assumes."
reversibility: "irreversible"
issues: ["The startups pursuing reversible brain cryopreservation are described without being named."]
---

```infobox
{
  "caption": "Preservation technology and research programme",
  "rows": [
    { "label": "Leading method", "value": "Aldehyde-stabilized cryopreservation" },
    { "label": "Demonstrated in", "value": "Rabbit and pig brains" },
    { "label": "Verification", "value": "Whole-brain electron microscopy" },
    { "label": "Prize awarded", "value": "2016 and 2018" },
    { "label": "Reversible", "value": "No, by design" },
    { "label": "Target", "value": "Connectome, not viability" }
  ]
}
```

**Brain preservation** is the storage of a brain in a state that retains the structural information thought to encode memory and personality, on the assumption that this information — not the living tissue — is what would have to survive for a person to be recoverable. It differs from [[cryonics]] in an important respect: the leading method is chemically irreversible and makes no attempt to keep the brain biologically viable. It is a bet on scanning, not on resuscitation, and it is taken up mainly by people who do not expect [[longevity-escape-velocity]] to arrive in time for them.

## What is being preserved

The premise is that a person's memories and character are physically instantiated in the arrangement of their neurons and synapses, together with molecular details such as receptor composition and synaptic strength. If that arrangement can be frozen in place — literally or chemically — then in principle it can later be read out, whatever the state of the tissue.

This shifts the success criterion. Cryonics asks whether damage can be repaired; brain preservation asks whether information has been destroyed. The two lead to different procedures. A method that crosslinks every protein in the brain into an immobile mesh would be catastrophic for revival and excellent for structural preservation, and that is roughly what the field's leading technique does.

> [!key] The trade being made
> Aldehyde fixation is the standard preparation for electron microscopy precisely because it locks ultrastructure in place. It also makes the tissue permanently non-living. Brain preservation accepts that trade; classical cryonics does not.

## Aldehyde-stabilized cryopreservation

Aldehyde-stabilized cryopreservation (ASC) was published in 2015 by Robert McIntyre and Greg Fahy, whose earlier work produced the low-toxicity vitrification solutions that [[cryonics]] organizations use.[^mcintyre2015] The procedure perfuses the brain's vasculature with glutaraldehyde, which crosslinks proteins within seconds and halts decay, then introduces a high concentration of a cryoprotectant — ethylene glycol in the published protocol — and cools the brain below its glass transition, around −135 °C, where it can be stored without ice formation or further chemical change.

Fixation solves the problem that limits conventional vitrification. Because the tissue is already stabilized, cryoprotectant can be introduced slowly and at concentrations that would be fatally toxic to living cells, and osmotic damage no longer matters. Fixation also stops the clock immediately, whereas a purely cryogenic protocol must race cooling against ischemia.

The published demonstrations used a rabbit brain and later a whole pig brain, with electron micrographs sampled across the specimen showing intact cell membranes, synapses and synaptic vesicles. Verification was the point: earlier preservation claims had rested on gross appearance rather than on ultrastructure at synaptic resolution.

## The Brain Preservation Prize

```timeline
[
  { "year": "2010", "title": "Foundation established", "text": "The neuroscientist Kenneth Hayworth founds the Brain Preservation Foundation and announces a prize for demonstrably complete preservation of a mammalian connectome." },
  { "year": "2015", "title": "ASC published", "text": "McIntyre and Fahy describe aldehyde-stabilized cryopreservation in Cryobiology, combining chemical fixation with vitrification." },
  { "year": "2016", "title": "Small Mammal Prize", "text": "The prize is awarded for an ASC-preserved rabbit brain whose ultrastructure is verified across the whole organ by electron microscopy." },
  { "year": "2018", "title": "Large Mammal Prize", "text": "A pig brain, preserved by the same method and evaluated the same way, wins the second and final tier of the prize." },
  { "year": "2018", "title": "Nectome controversy", "text": "MIT ends a research subcontract with a startup commercializing the method after coverage highlights that the procedure requires a living patient and is invariably fatal." }
]
```

The prize's design mattered more than its money. By fixing the evaluation criterion in advance — uniform preservation of synaptic ultrastructure across an entire brain, judged by independent electron microscopists — it forced a field that had traded in assertion into producing checkable images. It also, by its own terms, said nothing about whether a preserved brain could ever be read.

## The connectome criterion

The argument that structural preservation suffices runs through [[connectomics]]. If memory is stored in the pattern and strength of synaptic connections, then a preserved connectome plus enough molecular annotation is the person in storage, and the remaining work is scanning and simulation — the pipeline described in the [[whole-brain-emulation]] roadmap by [[anders-sandberg]] and [[nick-bostrom]].[^sandberg2008] Nothing in the method requires the [[medical-nanorobots|cell-by-cell repair machinery]] that biological revival scenarios depend on, which is the main reason its advocates consider it the more tractable bet.

The objections are substantial and not resolved.

**A wiring diagram may not be enough.** The complete connectome of *C. elegans* has been available since 1986, and its 302 neurons still do not yield a predictive model of the animal's behavior.[^white1986] Connectivity omits synaptic weights, neuromodulatory tone, receptor subtypes, gene expression state, and the extrasynaptic signaling that shapes circuit function.

**Fixation may destroy what matters.** Glutaraldehyde preserves gross ultrastructure but alters protein conformation and does not preserve small molecules, ion gradients, or short-lived phosphorylation states. If any of those carry information relevant to identity, ASC loses it and electron microscopy would not reveal the loss.

**Perfusion is uneven in real cases.** The demonstrations used healthy animals perfused under controlled conditions. A human brain after cardiac arrest has collapsed capillaries and swollen tissue, and fixative may not reach everywhere.

**The success test is unspecified.** Even a perfect reconstruction would raise the question of how anyone would know it worked. Behavioral resemblance underdetermines the answer, as the debates over [[machine-consciousness]] and the [[neural-correlates-of-consciousness]] make clear, and the preserved person is not available to check against.

**Nothing has been read back.** No preserved brain of any species has been scanned and shown to yield a functioning model. Until something is, the criterion "connectome preserved" is a proxy for success whose relationship to actual success is unmeasured. This is the same gap that separates a [[memory-prosthesis]] that can write a hippocampal code from one that can read an existing memory.

## Ethics and law

Because good perfusion requires intact circulation, the ideal application of ASC is to a patient who is anesthetized and alive at the moment fixation begins. That makes the procedure a cause of death rather than a response to one, and places it squarely inside the [[right-to-die-and-right-to-live|assisted-dying]] debate rather than beside it. Advocates ground the claim in [[morphological-freedom]] and the right to dispose of one's own body; opponents note that no other elective procedure is both irreversible and immediately lethal, and that consent to it is being sought on the strength of a technology nobody can demonstrate.

The startup Nectome ran into this directly. Founded by McIntyre to commercialize ASC, it attracted attention in 2018 for a service described in press coverage as "100 percent fatal", accepted refundable deposits from a waiting list, and lost its research subcontract with MIT shortly afterward.[^regalado2018] The scientific objection raised at the time was not that the preservation failed but that selling it implied a claim about future revival that nobody could support.

A smaller practice has continued. Some organizations offer chemical fixation with storage at higher temperatures, which is far cheaper than liquid-nitrogen maintenance and may be adequate if the goal is structure rather than viability. A few well-funded startups, including one founded by the longevity investor Laura Deming, have taken the opposite tack and are pursuing genuinely reversible brain cryopreservation, which if achieved would make the fixation trade unnecessary.

> [!caution] What the prize did and did not show
> The Brain Preservation Prize established that a large mammalian brain can be preserved with synapses visibly intact throughout. It did not establish that memories survive, that the preserved state is sufficient for reconstruction, or that any scanning technology capable of reading a whole human brain is achievable.

## Outlook

Brain preservation occupies an unusual position: its central technical claim is verifiable and largely verified, while its central practical claim is not testable at all with present knowledge. The bottleneck has moved to the read-out side, where progress depends on the throughput of volume electron microscopy and on whether models built from [[neural-decoding|decoded]] structure can reproduce function.

The field's honest position is that it preserves an option rather than delivering a result, and that the option's value depends on premises — [[substrate-independence]], the sufficiency of structure, the eventual existence of the scanning pipeline — that are argued rather than demonstrated. Whether preserving that option is worth doing, given that it must be exercised at the cost of a life that would otherwise have ended slightly later, is a question the technical results cannot settle. It is the same question [[mind-uploading]] raises, arriving decades earlier because the preservation half is already available.

## See also

- [[cryonics]]
- [[connectomics]]
- [[whole-brain-emulation]]
- [[mind-uploading]]
- [[personal-identity-and-continuity]]
- [[digital-immortality]]
- [[alcor]]
- [[head-transplant]]

## References

[^mcintyre2015]: `paper` McIntyre, R.L. and Fahy, G.M. "Aldehyde-stabilized cryopreservation." *Cryobiology*, 2015. {Demonstrated on rabbit brains; the verification is electron microscopy of preserved structure, which cannot show whether stored information survived.}
[^white1986]: `paper` White, J.G., Southgate, E., Thomson, J.N. and Brenner, S. "The structure of the nervous system of the nematode *Caenorhabditis elegans*." *Philosophical Transactions of the Royal Society B*, 1986.
[^sandberg2008]: `report` Sandberg, A. and Bostrom, N. *Whole Brain Emulation: A Roadmap*. Future of Humanity Institute, University of Oxford, 2008.
[^regalado2018]: `news` Regalado, A. "A startup is pitching a mind-uploading service that is '100 percent fatal'." *MIT Technology Review*, 2018. {Reporting on a startup's proposed service and the reaction to it; it describes a business plan, not a scientific result.}

==============================================================================
ARTICLE: brain-to-brain-interface
TITLE:   Brain-to-brain interfaces
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/brain-to-brain-interface
SOURCE:  https://futurehumanwiki.com/raw/brain-to-brain-interface
==============================================================================

---
title: "Brain-to-brain interfaces"
slug: "brain-to-brain-interface"
type: "technology"
status: "experimental"
horizon: "2050s+"
trl: 3
categories: ["cybernetics", "minds"]
tags: ["bci", "telepathy", "tms", "neural decoding", "neurostimulation", "bandwidth"]
summary: "Systems that link one nervous system to another by decoding activity from a sender and delivering stimulation to a receiver, so far at rates of a few bits per minute."
updated: "2026-07-27"
humanEvidence: "All human demonstrations are non-invasive: volunteers have exchanged roughly one bit per trial through scalp EEG and magnetic stimulation, and no healthy person has been implanted for this purpose."
access: "Nothing to obtain: the systems exist only as laboratory setups assembled from research EEG and stimulation equipment, with no product or service on offer."
reversibility: "reversible"
issues: ["The ethics section is thinner than the technical sections."]
---

```infobox
{
  "caption": "Experimental neurotechnology",
  "rows": [
    { "label": "First animal demonstration", "value": "2013 (rat to rat)" },
    { "label": "First human demonstration", "value": "2013–2014" },
    { "label": "Read method", "value": "EEG or implanted electrodes" },
    { "label": "Write method", "value": "TMS, ultrasound, microstimulation" },
    { "label": "Demonstrated rate", "value": "A few bits per minute" },
    { "label": "Largest network", "value": "Three humans" },
    { "label": "Readiness", "value": "TRL 3" }
  ]
}
```

**Brain-to-brain interfaces** are systems that connect two nervous systems by decoding a signal
from one brain and delivering stimulation to another, without the sender speaking or the receiver
listening. Every demonstration so far chains together two established technologies: a
[[brain-computer-interface]] that reads out a single variable, and a neurostimulation device that
writes a crude signal in. Nothing about the arrangement is telepathic in the sense the popular
coverage implies.

The achieved information rates are the essential fact. Across the published human experiments,
transmission has amounted to roughly one bit per trial, with trials taking tens of seconds — a few
bits per minute, against roughly forty bits per second for ordinary speech. What is transmitted is
not a thought but a pre-agreed binary code that the experimenters defined in advance.

## How it works

The sender's side is a decoder. In the non-invasive human experiments the sender looks at one of two
flickering targets, producing a steady-state visual evoked potential at the corresponding frequency
in occipital cortex, which scalp electroencephalography can classify reliably. Alternatively the
sender imagines moving a hand, producing a detectable change in sensorimotor rhythms. Either way,
the output is a single binary decision — the simplest possible case of [[neural-decoding]].

The receiver's side is a stimulator, drawn from the standard toolkit of
[[non-invasive-neuromodulation]]. Transcranial magnetic stimulation over primary motor cortex can
evoke an involuntary hand movement; over occipital cortex it can evoke a phosphene, a brief spot of
perceived light. The bit is encoded as stimulate or do not stimulate. In animal work the
write channel has been intracortical microstimulation through an implanted array, or focused
ultrasound delivered through the skull.

What passes between the brains is therefore a bit chosen by the experimental protocol. The
receiver does not experience the sender's percept; the receiver sees a flash, or feels a hand move,
and applies a rule learned beforehand about what a flash means.

## Development history

```timeline
[
  { "year": "2013", "title": "Rat to rat", "text": "A Duke group links an encoder rat's motor cortex to a decoder rat's via microstimulation; the decoder rat performs a lever task above chance using information it could not otherwise obtain." },
  { "year": "2013", "title": "Human to rat", "text": "A Harvard-affiliated team uses human EEG to trigger focused ultrasound over a rat's motor cortex, producing a tail movement." },
  { "year": "2013–2014", "title": "Human to human", "text": "A University of Washington pair transmit a motor-imagery signal from one person's EEG to magnetic stimulation over another's motor cortex, moving a finger to fire in a video game." },
  { "year": "2014", "title": "Words across continents", "text": "A team encodes 'hola' and 'ciao' in binary, sending them from an EEG subject in India to subjects in France who perceive the bits as phosphenes." },
  { "year": "2015", "title": "Multi-animal networks", "text": "Nicolelis's group links several rat and several monkey brains into 'brainets' that jointly perform computations or control a virtual arm." },
  { "year": "2019", "title": "BrainNet", "text": "Three people collaborate on a Tetris-like task, two sending rotate-or-not decisions by EEG and the third receiving them as phosphenes, with accuracy around 80 percent." }
]
```

The 2013 rat experiment remains the most substantive result, because it involved implanted
electrodes on both ends and the receiving animal genuinely improved at a task using transmitted
information.[^paisvieira2013] The human work has all been non-invasive, and therefore restricted to
the coarsest possible read and write.[^yoo2013][^rao2014][^grau2014] BrainNet, which linked three
people in a shared task, is the largest network demonstrated and still transmitted a single bit per
sender per turn.[^jiang2019] Multi-animal "brainets" have shown that several brains can be pooled
to control one output, which is a different and somewhat less surprising result: it is population
decoding across skulls.[^ramakrishnan2015]

> [!caution] Why "telepathy" is the wrong word
> None of these systems transmits meaning. The bit is meaningful only because both participants were
> told in advance what it would signify. A system with the same architecture and a light bulb
> instead of magnetic stimulation would work identically, faster, and with less equipment. The
> neuroscientific content of the demonstrations is that a decoded signal can be delivered as a
> percept, not that brains can share content.

## Why the bandwidth is so low

Three obstacles compound, and only one of them is an engineering problem.

**Reading.** Non-invasive read-out through the skull resolves centimetre-scale cortical activity at
best, which caps the sender's channel at a few bits per trial. Implanted electrodes would raise this
substantially, as the results from [[utah-array]] and [[ecog-interfaces]] work show, but nobody has
implanted a healthy volunteer to serve as a sender.

**Writing.** This is the harder half. Transcranial magnetic stimulation activates on the order of a
cubic centimetre of cortex indiscriminately; focused ultrasound is better localized but still
addresses large populations; intracortical microstimulation is finer but still activates hundreds
of cells in a pattern nothing like natural activity. The read-out fidelity of current
neurotechnology far exceeds its write-in fidelity, which is also why sensory feedback in
[[neuroprosthetics]] lags motor decoding. Precise write-in through optical control is possible in
animals via [[optogenetics]], but requires genetically modifying the target neurons.

**No common code.** Even with perfect read and write, the deeper problem is that two brains do not
share a representational format. The population activity encoding a particular concept in one
person's cortex has no fixed correspondence to activity in another's; representations are shaped by
individual developmental and learning history. Delivering A's neural pattern to B would not
reproduce A's experience in B, any more than installing one computer's memory contents in another
with a different architecture would run the program.

The plausible workaround is learning rather than translation. A receiver could, in principle, learn
to interpret an arbitrary but consistent stimulation code, in the way that users of
[[sensory-augmentation]] devices come to experience a substituted signal as a perception rather
than a puzzle. That has not been attempted at any scale for brain-to-brain transmission, and it
would make the channel a new sense to be learned rather than a shortcut past learning.

## Ethics

The experiments raise a problem unusual in neurotechnology: the receiver's body moves without the
receiver intending it. Participants in the human studies consented to precisely that, but the
arrangement makes the question of agency concrete in a way that a cursor decoder does not. Related
concerns about who holds the transmitted signal, and what could be inferred from it, fall under
[[mental-privacy]] and the emerging [[neurorights]] proposals.

There is also a persistent framing problem. Commercial neurotechnology has borrowed the vocabulary
— [[neuralink]] markets its cursor-control product under the name Telepathy — in ways that inflate
public expectations of what neural interfaces do. The distance between typing with a decoder and
sharing an experience is not a matter of degree.

## Outlook

No path from the current demonstrations to meaningful brain-to-brain communication has been
articulated in technical detail. The requirements are a high-resolution write channel that does not
exist, a solution to the shared-code problem that no one has proposed beyond "the receiver learns",
and an ethical case for implanting healthy people that nothing currently justifies.

The idea nonetheless persists because it sits at the centre of several other speculative
programmes: proposals for [[human-ai-merger]] often assume a rich bidirectional neural channel,
scenarios built on [[whole-brain-emulation]] assume that a mind's content can be extracted and
moved, and arguments about merged or networked minds in the [[mind-uploading]] literature take
shared access between substrates for granted. Those arguments generally inherit the assumption that transmitting
neural activity transmits its content. The brain-to-brain experiments are the closest thing to a
direct test of that assumption, and what they show is a single bit crossing a gap that everything
else about the two brains has to be arranged in advance to interpret.

## See also

- [[brain-computer-interface]]
- [[neural-decoding]]
- [[non-invasive-neuromodulation]]
- [[sensory-augmentation]]
- [[mental-privacy]]
- [[human-ai-merger]]
- [[neuroprosthetics]]
- [[mind-uploading]]

## References

[^paisvieira2013]: `paper` Pais-Vieira, M., Lebedev, M., Kunicki, C., Wang, J. and Nicolelis, M. A. L. "A brain-to-brain interface for real-time sharing of sensorimotor information." *Scientific Reports*, 2013.

[^yoo2013]: `paper` Yoo, S.-S. et al. "Non-invasive brain-to-brain interface (BBI): establishing functional links between two brains." *PLOS ONE*, 2013.

[^rao2014]: `paper` Rao, R. P. N. et al. "A direct brain-to-brain interface in humans." *PLOS ONE*, 2014.

[^grau2014]: `paper` Grau, C. et al. "Conscious brain-to-brain communication in humans using non-invasive technologies." *PLOS ONE*, 2014.

[^jiang2019]: `paper` Jiang, L., Stocco, A., Losey, D. M., Abernethy, J. A., Prat, C. S. and Rao, R. P. N. "BrainNet: a multi-person brain-to-brain interface for direct collaboration between brains." *Scientific Reports*, 2019.

[^ramakrishnan2015]: `paper` Ramakrishnan, A. et al. "Computing arm movements with a monkey brainet." *Scientific Reports*, 2015. {The monkeys were not connected to each other; a computer combined their recorded activity to drive one output.}

==============================================================================
ARTICLE: brain-computer-interface
TITLE:   Brain–computer interface
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/brain-computer-interface
SOURCE:  https://futurehumanwiki.com/raw/brain-computer-interface
==============================================================================

---
title: "Brain–computer interface"
slug: "brain-computer-interface"
type: "technology"
status: "emerging"
horizon: "late 2020s"
trl: 6
categories: ["cybernetics", "foundations"]
tags: ["bci", "neurotechnology", "paralysis", "neural decoding", "implants", "assistive technology"]
summary: "A system that measures nervous-system activity, decodes an intention or state from it, and converts that decoded signal into a command for an external device."
updated: "2026-07-27"
humanEvidence: "Implanted arrays have given research participants with paralysis cursor control, robotic-arm use and conversational-rate speech decoding; no implanted BCI has been shown to improve any capability in a healthy person."
access: "No implanted BCI is approved by any regulator; the only route is enrolment in an early feasibility trial. Consumer EEG headsets are sold freely as wellness devices."
reversibility: "difficult"
issues: ["The speech-decoding section states vocabulary and error-rate figures without citations."]
---

```infobox
{
  "caption": "Class of neurotechnology",
  "rows": [
    { "label": "Term coined", "value": "1973, Jacques Vidal" },
    { "label": "Signal sources", "value": "Spikes, LFP, ECoG, EEG, fNIRS" },
    { "label": "First human implant trial", "value": "2004 (BrainGate)" },
    { "label": "Primary clinical use", "value": "Severe paralysis, anarthria" },
    { "label": "Regulatory status", "value": "Investigational; no approved implanted BCI" },
    { "label": "Readiness", "value": "TRL 6" }
  ]
}
```

**Brain–computer interface** (BCI) is the general term for a system that measures activity in the
nervous system, extracts a variable of interest from that activity, and turns it into a command for
a computer, a prosthesis, or the person's own muscles. The defining feature, in the standard
formulation, is that the command bypasses the ordinary output path of peripheral nerve and
muscle.[^wolpaw2002] A BCI is therefore useful mainly to people
whose ordinary output path is broken, and as of 2026 every serious clinical result comes from that
population: participants with tetraplegia, brainstem stroke, or amyotrophic lateral sclerosis.

The field has produced genuinely striking demonstrations — a person with no usable movement
controlling a robotic arm, another producing fluent synthesized speech from motor-cortex activity
alone — in a very small number of research participants, using hardware that is tethered, requires
frequent recalibration, and has not been approved by any regulator for permanent use. The gap
between those demonstrations and a product is largely a gap in reliability, longevity, and
manufacturing, not in principle.

```figure
{"key": "braingate-array", "caption": "The implanted array against a human hand. The entire recording surface is a few millimetres square."}
```

## The signal chain

Every BCI is the same four stages: acquisition, feature extraction, decoding, and output, usually
with a feedback loop closing back to the user.

*Acquisition* determines everything downstream. Signals differ by how close the sensor sits to the
neurons producing them. Penetrating microelectrodes recording action potentials from individual
cells carry the most information per channel; scalp electroencephalography, separated from cortex
by dura, skull, and scalp, carries the least, because the skull acts as a spatial low-pass filter
that smears contributions from many square centimetres of cortex into one measurement.

*Feature extraction* reduces raw voltage into something a decoder can use: firing rates binned over
tens of milliseconds, power in a frequency band, the amplitude of an evoked potential. In cortical
recordings the workhorse features are threshold crossings and high-gamma power (roughly 70–200 Hz),
the latter being a reasonable proxy for local population spiking that can be measured without
penetrating the cortex.

*Decoding* maps features to intent. Early systems used linear filters and the population-vector
approach inherited from primate motor physiology; current systems use Kalman filters, recurrent
neural networks, and increasingly transformer-style sequence models paired with language models.
[[neural-decoding]] is where most of the recent performance gains have come from — the electrodes
in the best 2025 speech systems are not fundamentally better than those used in 2012.

*Output* is a cursor, a robotic arm, a synthesized voice, a wheelchair, or stimulation delivered to
the user's own spinal cord or to a powered [[exoskeleton]]. Bidirectional systems also write
information back in, which is a much harder problem than reading it out and remains the main
limitation for restoring touch.

### Non-invasive systems

Scalp EEG supports three well-characterized paradigms. The P300 speller flashes rows and columns of
a character grid and detects the evoked response that follows an attended flash; steady-state
visually evoked potential systems tag each option with a distinct flicker frequency and read off
which frequency dominates visual cortex; sensorimotor-rhythm systems ask the user to imagine
movement of a limb and detect the resulting suppression of mu and beta rhythms. All three work.
None is fast: information transfer rates are conventionally reported in the range of tens of bits
per minute, on the order of a few characters, against several hundred bits per minute for competent
typing.

Functional near-infrared spectroscopy and functional MRI measure haemodynamics rather than
electrical activity, and are therefore limited by the seconds-long delay of the blood-oxygen
response. fMRI remains valuable for decoding research precisely because it covers the whole brain,
but a scanner is not a wearable device. Consumer EEG headsets sit at the weakest end
of the spectrum, typically with a handful of dry electrodes and heavy artefact contamination from
eye movement and jaw muscle.

```compare
{
  "columns": ["Penetrating microelectrodes", "Subdural ECoG", "Endovascular", "Scalp EEG"],
  "rows": [
    { "label": "Surgery", "values": ["Craniotomy, cortical penetration", "Craniotomy or slot", "Catheter via jugular vein", "None"] },
    { "label": "Resolution", "values": ["Single neurons", "Millimetre populations", "Coarse populations", "Centimetre populations"] },
    { "label": "Typical channels", "values": ["96–1,024", "64–1,024", "16", "8–256"] },
    { "label": "Long-term stability", "values": ["Degrades over months to years", "Good over years", "Good over years", "Good, but user-dependent"] },
    { "label": "Demonstrated ceiling", "values": ["Multi-degree-of-freedom limb control, fluent speech", "Fluent speech, walking via spinal stimulation", "Discrete clicks", "Spellers, simple switches"] }
  ]
}
```

## Development history

The name is older than the capability. Jacques Vidal proposed in 1973 that a computer could
interpret electroencephalographic signals in real time as deliberate control commands, coining the
phrase and sketching the architecture two decades before hardware could support it.[^vidal1973]
Animal work in the late 1990s established that ensembles of cortical neurons carry enough movement
information to drive a machine; the human trials followed within five years.

```timeline
[
  { "year": "1924", "title": "Human EEG recorded", "text": "Hans Berger records electrical rhythms from the human scalp, establishing that cortical activity can be measured non-invasively." },
  { "year": "1973", "title": "The term is coined", "text": "Jacques Vidal at UCLA publishes 'Toward direct brain-computer communication', proposing that computers could interpret EEG in real time as control signals." },
  { "year": "1999–2002", "title": "Motor decoding in animals", "text": "Groups led by John Chapin, Miguel Nicolelis, and Andrew Schwartz show that rats and monkeys can control levers and robotic arms using decoded cortical ensemble activity." },
  { "year": "2004–2006", "title": "First long-term human implant", "text": "Matthew Nagle, paralysed by a spinal cord injury, receives a Utah array in the BrainGate pilot and controls a cursor, a television, and a prosthetic hand." },
  { "year": "2012–2013", "title": "Robotic arm control", "text": "BrainGate and University of Pittsburgh teams report multi-degree-of-freedom robotic arm control by participants with tetraplegia, including self-feeding." },
  { "year": "2016", "title": "Fully implanted communication BCI", "text": "A Utrecht group implants a wireless electrocorticography system in a person with late-stage ALS, who uses it at home for years to select letters." },
  { "year": "2021–2024", "title": "Speech decoding becomes fluent", "text": "UCSF, Stanford, and UC Davis teams decode attempted speech at conversational rates from motor cortex, with large vocabularies and single-digit word error rates in the best case." },
  { "year": "2023", "title": "Brain–spine interface", "text": "A Lausanne team restores volitional walking in a man with spinal cord injury by linking a cortical implant to an epidural spinal stimulator." },
  { "year": "2024–2026", "title": "Commercial implants enter trials", "text": "Neuralink, Synchron, Precision Neuroscience, Paradromics, and several Chinese groups run early feasibility studies; none has regulatory approval for routine use." }
]
```

## Current state

Three capability areas have credible human results.

**Cursor and device control.** People with tetraplegia can move a computer pointer, click, type on a
virtual keyboard, and drive a robotic arm through several degrees of freedom.[^hochberg2006] The
2012 and 2013 robotic-arm reports, in which participants reached, grasped, and drank
unassisted, remain the reference demonstrations.[^hochberg2012][^collinger2013] Performance in the
best participants approaches a slow but usable pointing device. The [[utah-array]] remains the most
common sensor for this work, and the multi-site BrainGate consortium has produced most of the
long-run data. A fully implanted surface-electrode system used at home by a person with late-stage
ALS demonstrated that the approach can survive outside a laboratory.[^vansteensel2016]

**Speech.** Between 2021 and 2025 decoding of attempted speech moved from a 50-word vocabulary at
around fifteen words per minute to conversational-rate output over vocabularies of more than a
hundred thousand words, with the best reported word error rates in the single digits for one
participant. Some systems now synthesize a voice reconstructed from pre-illness recordings, or
animate an avatar. See [[speech-neuroprosthesis]].

**Movement restoration.** Rather than driving an external robot, a cortical implant can drive
stimulation of the user's own spinal cord or muscles, re-establishing a "digital bridge" around a
lesion. This has been demonstrated for grasping and, in a 2023 report, for volitional walking in a
man with an incomplete spinal cord injury, who retained some improvement even with the system
switched off.[^lorach2023]

Device approaches now in human trials span the invasiveness spectrum: penetrating arrays
([[neuralink]], Paradromics, Blackrock Neurotech), thin-film surface arrays ([[ecog-interfaces]],
Precision Neuroscience), and endovascular electrodes threaded through a vein ([[stentrode]],
developed by [[synchron]]). Each trades bandwidth against surgical risk. Several Chinese academic
and state-backed groups have also reported implants in participants since 2024, though public
detail is limited and independent verification scarce.

What has not been demonstrated is equally worth stating. No BCI has restored a rich sense of touch;
the best feedback so far is a small number of discriminable pressure percepts delivered by
intracortical or peripheral stimulation, well short of what a hand needs. No system has restored
memory in a clinically meaningful way, despite the encoding-model work described under
[[memory-prosthesis]]. And no implanted BCI has been shown to improve any capability in a healthy
person, which is the premise on which most enhancement-oriented interest rests.

> [!caution] Sample sizes are tiny
> The headline results in this field typically come from one to four participants, often the same
> individuals across multiple papers, with hardware maintained by a research team. Nothing here has
> been tested at the scale that would establish reliability, and single-participant results have
> repeatedly failed to generalize.

## The bandwidth problem

Popular framing treats BCI progress as an electrode-count race, on the assumption that information
transfer scales with channels. It does not scale cleanly. Motor cortical activity during reaching
is well described by a low-dimensional dynamical structure — a neural manifold — in which a few
dozen latent dimensions capture most of the variance.[^gallego2017] Adding electrodes samples that same structure
more densely rather than revealing new independent signals, so returns diminish. This is why a
96-channel array from the 1990s can support performance comparable to systems with ten times the
channels, and why decoder architecture has mattered more than sensor count.

Where more channels do help is coverage: sampling several cortical areas, or sampling a functional
map finely enough to separate articulators in speech cortex. The genuinely hard constraints are
elsewhere — the number of independent control dimensions a person can learn to modulate, the
non-stationarity of the recorded population from day to day, and the near-total absence of a
high-resolution write channel. Reading intention out is a solved-enough problem for many
applications; writing rich sensory information in is not, which is why touch feedback in
[[neuroprosthetics]] remains crude compared with motor decoding.

## Limitations

Implanted arrays lose channels. The brain mounts a foreign-body response — microglial activation,
astrocytic encapsulation — that displaces neurons from the recording tip, while insulation and
metal traces degrade in a warm saline environment. Signal yield typically declines over months to
years, though some implants have supported usable control beyond a thousand days and a few for
several years; the failure statistics are set out under [[utah-array]].

Percutaneous connectors, still standard in academic systems, carry infection risk and tether the
user to a lab. Fully implanted wireless systems solve this but constrain power and data rate; the
most aggressive proposal for removing the tether, untethered ultrasonic motes, is covered under
[[neural-dust]] and has no human results.
Decoders drift, requiring recalibration that ranges from minutes daily to occasional, depending on
the system. Almost all reported performance comes from controlled settings with a technician
present.

There is also a structural risk specific to commercial neurotechnology: a company can fail while
its devices remain in people's heads. The discontinuation of the Argus II retinal system left users
with unsupported implants, a precedent that shadows every BCI trial. See [[retinal-implant]].

## Ethics and governance

Implanted BCIs raise questions that older neurotechnology such as [[deep-brain-stimulation]] and
the [[cochlear-implant]] introduced in milder form: who owns the neural data, what happens when a
decoder misreads intent, and whether a device that shapes behaviour compromises agency. The
specific novelty is that recorded activity can support inferences the user did not intend to
disclose, an issue taken up under [[mental-privacy]]. Legislatures have begun responding —
constitutional and state-level neural-data provisions exist in several jurisdictions — under the
banner of [[neurorights]], though enforcement mechanisms remain thin.

Consumer neurotechnology is a separate regulatory category and a much weaker technical one.
Wearable EEG headsets and [[non-invasive-neuromodulation]] devices are sold for attention training,
sleep, and gaming with evidence that ranges from thin to absent, and the "brain-reading" claims
attached to them substantially outrun what scalp electrodes can resolve. Regulators have generally
treated these as wellness products rather than medical devices, the same line that governs
[[wearable-health-sensors]], which is why the emerging neural-data statutes target data handling
rather than efficacy claims.

> [!debate] How invasive is worth it
> Synchron's position is that a device delivered by catheter, with sixteen electrodes, will reach
> far more patients than one requiring a craniotomy, and that discrete reliable control is enough
> for most assistive tasks. The counter-position, taken by the penetrating-array developers, is
> that low-bandwidth control caps the addressable applications at switch-like interaction and rules
> out speech and dexterous limb control. Both claims are testable and neither has been settled.

## Outlook

The near-term agenda is legible: more participants, wireless and fully implanted systems, decoders
intended to survive a week without recalibration, and eventually a pivotal trial in a well-defined
indication such as loss of speech in ALS. If that sequence holds, it could produce an approved
assistive device for a small clinical population, though no timeline for it is well supported and
none of the current studies is large enough to predict one.

The longer-term claims — that BCIs will serve as a general-purpose channel for
[[human-ai-merger]], or provide the read-out fidelity that [[whole-brain-emulation]] would require
— rest on assumptions the clinical work does not test. A speech decoder recovers a motor plan, not
a thought; a 1,024-channel array samples a millionth of a cortical hemisphere. Whether elective
implantation in healthy people ever clears the risk-benefit bar is a question about surgical
complication rates and device lifetimes, and neither is currently known well enough to answer.

## See also

- [[neural-decoding]]
- [[utah-array]]
- [[stentrode]]
- [[speech-neuroprosthesis]]
- [[neuroprosthetics]]
- [[neuralink]]
- [[brain-to-brain-interface]]
- [[neurorights]]

## References

[^vidal1973]: `paper` Vidal, J. J. "Toward direct brain-computer communication." *Annual Review of Biophysics and Bioengineering*, 1973.

[^hochberg2006]: `paper` Hochberg, L. R. et al. "Neuronal ensemble control of prosthetic devices by a human with tetraplegia." *Nature*, 2006.

[^hochberg2012]: `paper` Hochberg, L. R. et al. "Reach and grasp by people with tetraplegia using a neurally controlled robotic arm." *Nature*, 2012.

[^collinger2013]: `paper` Collinger, J. L. et al. "High-performance neuroprosthetic control by an individual with tetraplegia." *The Lancet*, 2013. {A single participant, whose performance was reached over months of training with the research team present.}

[^vansteensel2016]: `paper` Vansteensel, M. J. et al. "Fully implanted brain–computer interface in a locked-in patient with ALS." *New England Journal of Medicine*, 2016.

[^gallego2017]: `paper` Gallego, J. A. et al. "Neural manifolds for the control of movement." *Neuron*, 2017. {A perspective drawn largely from animal motor cortex; the inference that added electrodes give diminishing returns is an extrapolation from it.}

[^lorach2023]: `paper` Lorach, H. et al. "Walking naturally after spinal cord injury using a brain–spine interface." *Nature*, 2023.

[^wolpaw2002]: `paper` Wolpaw, J. R. et al. "Brain–computer interfaces for communication and control." *Clinical Neurophysiology*, 2002.

==============================================================================
ARTICLE: buck-institute
TITLE:   Buck Institute for Research on Aging
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/buck-institute
SOURCE:  https://futurehumanwiki.com/raw/buck-institute
==============================================================================

---
title: "Buck Institute for Research on Aging"
slug: "buck-institute"
type: "organization"
status: "established"
horizon: "present"
categories: ["organizations", "longevity"]
tags: ["aging", "biogerontology", "research institutes", "senescence", "training", "nonprofit"]
summary: "An independent research institute in Novato, California, opened in 1999 as the first in the world dedicated solely to the biology of aging."
updated: "2026-07-27"
issues: ["Scale figures for laboratory count and annual expenditure carry no source."]
---

```infobox
{
  "caption": "Independent research institute",
  "rows": [
    { "label": "Opened", "value": "1999" },
    { "label": "Location", "value": "Novato, California" },
    { "label": "Origin", "value": "Buck Trust bequest, Marin County" },
    { "label": "Architect", "value": "I. M. Pei" },
    { "label": "President", "value": "Eric Verdin" },
    { "label": "Focus", "value": "Biology of aging; geroscience" },
    { "label": "Distinction", "value": "First institute devoted solely to aging" }
  ]
}
```

**Buck Institute for Research on Aging** is an independent, nonprofit biomedical research institute in Novato, California, opened in 1999 and the first research organization in the world dedicated exclusively to the biology of aging. It sits at the centre of the modern [[geroscience-hypothesis|geroscience]] programme both intellectually and institutionally: much of the foundational work on [[cellular-senescence]] and the senescence-associated secretory phenotype was done there, and a substantial share of the field's principal investigators and company founders trained in its laboratories.

## Overview

The institute's organizing claim is the geroscience premise: that the major chronic diseases of later life share upstream biological drivers, and that intervening on those drivers would do more good than treating each disease separately. Its research is organized around aging mechanisms rather than around organ systems or diseases, which is unusual for a biomedical institute and was more unusual still when it opened.

It runs roughly two dozen laboratories, a graduate programme, and a translational arm intended to move findings toward companies. Its scale is modest — annual expenditure in the tens of millions of dollars — relative to [[calico]] or [[altos-labs]], and its output per dollar has been correspondingly high.

## History

The institute exists because of a lawsuit. Leonard and Beryl Buck left a charitable trust for use in Marin County, California; the trust's assets grew far beyond expectation after the value of its holdings rose, and the resulting litigation in the 1980s over whether the money had to be spent within one wealthy county produced a settlement that funded several new institutions, among them a research centre on aging.

Construction on a campus designed by I. M. Pei began in the 1990s and the institute opened in 1999. Dale Bredesen was its founding president. Brian Kennedy, a yeast geneticist known for work on the genetics of lifespan, led it during the first half of the 2010s, and Eric Verdin, an immunologist and metabolism researcher, has been president since 2016.

```timeline
[
  { "year": "1980s", "title": "The Buck Trust litigation", "text": "A dispute over whether a charitable bequest must be spent in Marin County ends in a settlement that funds a new institute on aging." },
  { "year": "1999", "title": "Doors open", "text": "The Buck Institute begins operations on an I. M. Pei-designed campus in Novato, the first research organization devoted solely to the biology of aging." },
  { "year": "2008", "title": "The SASP described", "text": "Judith Campisi's group characterizes the senescence-associated secretory phenotype, showing that senescent cells actively signal to their neighbours rather than simply sitting inert." },
  { "year": "2011", "title": "Senescent-cell clearance", "text": "Work at the Mayo Clinic, building on the senescence biology developed at the Buck and elsewhere, shows that removing senescent cells improves healthspan measures in mice." },
  { "year": "2011", "title": "Unity Biotechnology", "text": "A company is founded to develop senolytic drugs, with Campisi among its scientific founders." },
  { "year": "2016", "title": "Verdin era", "text": "Eric Verdin becomes president and expands the institute's work on metabolism, immune aging and translation." }
]
```

## Research programme

### Cellular senescence

The institute's most consequential contribution is the biology of senescent cells. Judith Campisi, who worked at the Buck from its early years until her death in 2024, established that senescent cells are not merely arrested but metabolically active, secreting a complex mixture of cytokines, proteases and growth factors now called the senescence-associated secretory phenotype.[^coppe2008] That finding reframed senescence from a cell-culture curiosity into a mechanism by which a small number of cells could drive tissue-wide dysfunction, and it is the direct intellectual ancestor of [[senolytics]] as a drug class.

Campisi also articulated the antagonistic-pleiotropy account of senescence: the programme suppresses cancer in young animals by arresting damaged cells, and harms old ones through the same secretory activity — an evolutionary tradeoff rather than a design flaw.[^campisi2013] The proof that this matters causally came from elsewhere, when a Mayo Clinic group showed that genetically clearing p16-positive cells delayed several age-associated pathologies in mice,[^baker2011] but the mechanistic account that made the experiment interpretable was largely built at the Buck. It also connects senescence directly to [[inflammaging]], since the secretory phenotype is one of the principal sources of the chronic sterile inflammation that characterizes aged tissue.

### Autophagy and proteostasis

Several laboratories work on [[autophagy]] and the mechanisms of [[proteostasis]] failure, including their role in neurodegeneration. This connects directly to the observation that most interventions extending lifespan in model organisms require an intact autophagy pathway.

### Metabolism and nutrient sensing

Work on ketone bodies, NAD metabolism and dietary interventions links the institute to the [[caloric-restriction]] literature and to the pharmacology of [[rapamycin]] and [[nad-precursors]]. Verdin's own research concerns sirtuins and the metabolic control of gene expression.

### Model organisms and biomarkers

The institute maintains substantial invertebrate work — *C. elegans* and *Drosophila* screens for compounds and genes affecting lifespan — alongside mouse studies, and contributes to the [[aging-biomarkers]] effort that the field needs before any intervention can be tested efficiently in humans. Invertebrate screens are cheap and fast; they are also the reason the field has a long list of compounds that extend lifespan in worms and almost none that have done anything measurable in a person, a filtering problem that the [[epigenetic-clock]] literature has not solved either.

> [!note] Why an independent institute
> Aging research is poorly served by disease-oriented funding, because no funding agency has a study section for a process that is not a disease. An institute organized around mechanisms rather than organs can support work that would fall between the cracks of a university's departmental structure, which is the argument the Buck's founders made and which the [[sens-research-foundation]] later made in a different register.

## Funding

The institute is funded by a combination of the original endowment, competitive federal grants — principally from the National Institute on Aging — philanthropy, and industry collaboration. It is not a university and grants no undergraduate degrees, but it operates a doctoral programme in the biology of aging in partnership with the University of Southern California, one of very few graduate programmes anywhere organized around the subject.

That programme is the institute's second major contribution. Aging biology had no training pipeline of its own until relatively recently, and researchers entered it laterally from cancer biology, genetics or metabolism. A dedicated programme produces people who frame their questions in the field's own terms from the start.

## Reception and influence

The Buck occupies an unusual position: it is respected by mainstream biology in a way that most institutions associated with life extension are not, and it is treated with suspicion by parts of the longevity advocacy community for the same reason. Its researchers publish in conventional venues, avoid claims about lifespan extension in humans, and have been publicly critical of the field's commercial overreach.

Its spin-outs and alumni are widely distributed. Unity Biotechnology, founded on senescence biology with Campisi among its scientific founders, ran some of the earliest clinical trials of senolytic drugs — and produced results that were largely negative, a fact the institute's researchers have discussed openly rather than minimized. Other alumni run laboratories and companies across the sector, and the institute's people appear on the advisory boards of most large longevity ventures, including as consultants to the well-funded reprogramming companies.

The main criticism directed at the institute is not specific to it. A quarter-century of well-regarded mechanistic work on aging has produced no approved intervention against aging in humans, and the institute's own scientists are among those who point out that the gap between mouse and human results has not narrowed.

## Outlook

The institute's near-term significance depends on the same thing the whole field depends on: whether a validated [[biological-age|measure of biological aging]] can be qualified as an endpoint, which would make aging-directed trials affordable. The Buck contributes to that effort and would be a principal beneficiary of success.

A second question concerns the field's centre of gravity. When the institute opened, it was the only place doing this work at scale; as of 2026 several private companies have larger budgets and can pay more. Whether an independent academic institute retains a distinctive role in that environment — as a trainer of researchers, as a source of results that are published rather than held, and as a check on claims made by the companies that recruit from it — is the question its next decade will settle.

## See also

- [[cellular-senescence]]
- [[senolytics]]
- [[geroscience-hypothesis]]
- [[hallmarks-of-aging]]
- [[calico]]
- [[sens-research-foundation]]
- [[aging-biomarkers]]
- [[inflammaging]]

## References

[^coppe2008]: `paper` Coppé, J.-P., Patil, C.K., Rodier, F. et al. "Senescence-Associated Secretory Phenotypes Reveal Cell-Nonautonomous Functions of Oncogenic RAS and the p53 Tumor Suppressor." *PLoS Biology*, 2008. {The secretory phenotype was characterised in cultured cells; its role in aged tissue was inferred from later work.}
[^baker2011]: `paper` Baker, D.J. et al. "Clearance of p16Ink4a-positive senescent cells delays ageing-associated disorders." *Nature*, 2011. {The mice were a progeroid strain carrying a transgene that let p16-positive cells be killed on demand, not normally aged animals.}
[^campisi2013]: `paper` Campisi, J. "Aging, Cellular Senescence, and Cancer." *Annual Review of Physiology*, 2013.

==============================================================================
ARTICLE: calico
TITLE:   Calico Life Sciences
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/calico
SOURCE:  https://futurehumanwiki.com/raw/calico
==============================================================================

---
title: "Calico Life Sciences"
slug: "calico"
type: "organization"
status: "established"
horizon: "2030s"
categories: ["organizations", "longevity"]
tags: ["aging", "biotech", "drug discovery", "comparative biology", "google", "research institutes"]
summary: "Alphabet's independent biotechnology company, founded in 2013 to study the biology of aging and develop interventions against age-related disease."
updated: "2026-07-27"
issues: ["The AbbVie funding totals and the reported narrowing of the partnership carry no source."]
---

```infobox
{
  "caption": "Private biotechnology company",
  "rows": [
    { "label": "Founded", "value": "September 2013" },
    { "label": "Parent", "value": "Alphabet Inc." },
    { "label": "Chief executive", "value": "Arthur Levinson" },
    { "label": "Headquarters", "value": "South San Francisco, California" },
    { "label": "Main partner", "value": "AbbVie" },
    { "label": "Focus", "value": "Biology of aging; age-related disease" },
    { "label": "Lead clinical asset", "value": "eIF2B activator for ALS" }
  ]
}
```

**Calico Life Sciences**, originally the California Life Company, is a biotechnology company founded by Google in September 2013 to study the biology of aging and translate it into treatments for age-related disease. It was the first attempt by a large technology company to fund aging research at pharmaceutical scale, it operates with unusual secrecy, and after more than a decade its public output remains modest relative to the capital committed. It is the reference case, cited both by supporters and critics, for what happens when the [[geroscience-hypothesis]] is given a large budget and a long horizon.

## Overview

Calico's founding proposition was that aging is a tractable biological problem that pharmaceutical companies neglect because it is not a disease with a regulatory pathway, and that a company insulated from quarterly pressure could do the basic biology first and find the drug later. Arthur Levinson, formerly chief executive of Genentech and chairman of Apple, was named chief executive at launch and has held the role since.

The company divides its work between discovery research — comparative biology, cellular aging mechanisms, computational genomics — and a drug development organization run in partnership with AbbVie. The two halves are connected loosely enough that Calico's most advanced clinical assets do not obviously derive from its aging research.

## History

Google announced Calico in September 2013 with a *Time* magazine cover asking whether the company could solve death, a framing Levinson has consistently avoided. David Botstein, a geneticist known for foundational work on genetic linkage mapping, joined as chief scientific officer and built the discovery organization. Hal Barron joined as president of research and development in 2014 and left in 2018 for GSK; he later became chief executive of [[altos-labs]].

In September 2014, Calico and AbbVie announced a collaboration in which each committed several hundred million dollars, with AbbVie providing pharmaceutical development capability and commercial rights and Calico providing discovery. The partnership was expanded twice, and total committed funds across the extensions have been reported in the billions. Reports from the mid-2020s describe the collaboration as having narrowed, with fewer programmes carried forward.

```timeline
[
  { "year": "2013", "title": "Founded by Google", "text": "Calico is announced with Arthur Levinson as chief executive and a mandate to study aging and age-related disease." },
  { "year": "2014", "title": "AbbVie partnership", "text": "A collaboration is announced with each side committing several hundred million dollars, later expanded, giving Calico a route to clinical development." },
  { "year": "2018", "title": "Naked mole-rat mortality", "text": "Calico researchers publish evidence that naked mole-rat mortality risk does not rise measurably with age, an unusual departure from the Gompertz pattern." },
  { "year": "2018", "title": "Longevity heritability revised down", "text": "Working with Ancestry.com pedigree data, Calico scientists argue that published estimates of the heritability of human lifespan are inflated by assortative mating." },
  { "year": "2020s", "title": "Clinical entry", "text": "An eIF2B activator developed with AbbVie enters trials in amyotrophic lateral sclerosis and a rare leukodystrophy, becoming the partnership's most advanced asset." },
  { "year": "2023–2025", "title": "Partnership narrowed", "text": "Reports describe the collaboration being scaled back in scope as programmes are prioritized, with Calico's focus adjusting accordingly." }
]
```

## Research programme

### Comparative biology

Calico recruited Rochelle Buffenstein and her naked mole-rat colony, one of the largest anywhere. The colony produced the field's most-cited quantitative claim about [[negligible-senescence]]: an analysis of thousands of animal-years finding that the hazard of death in naked mole-rats does not increase detectably with age after sexual maturity, in contrast to the Gompertzian doubling seen in nearly every other mammal.[^ruby2018] The result does not mean the animals are immortal — they die of infection, injury and, rarely, cancer — but it establishes that mammalian mortality acceleration is not obligatory.

The company has also worked on the short-lived African turquoise killifish as a vertebrate aging model, and on yeast under Botstein. Comparative work of this kind is the main route by which the field learns which of the [[cellular-senescence|damage mechanisms]] observed in laboratory mice are general features of mammalian aging and which are artefacts of one heavily inbred species with a three-year lifespan.

### Human genetics

A Calico collaboration with Ancestry.com analysed pedigree data from hundreds of millions of individuals and concluded that published heritability estimates for human lifespan — commonly given as roughly 25 per cent — are substantially inflated because people choose partners with similar life expectancy, which pedigree methods misattribute to genetics.[^ruby2018b] The corrected estimate was well under 10 per cent. This is one of the field's more consequential negative results, and it constrains what genomic approaches to [[maximum-human-lifespan]] can be expected to deliver. The estimate is a population average; the centenarian-genetics programmes run by [[nir-barzilai]] and others instead look for variants in exceptionally long-lived families.

### Drug programmes

Calico's most advanced clinical asset with AbbVie is a small-molecule activator of eIF2B, the translation-initiation factor whose inhibition drives the integrated stress response. It has been studied in amyotrophic lateral sclerosis and in vanishing white matter disease, a rare leukodystrophy caused by eIF2B mutations. Other disclosed programmes have included work on [[proteostasis]] and on cancer immunology. None is an aging indication, because no such indication exists at any regulator — the problem the [[metformin|TAME trial]] was designed to address.

> [!key] Why the pipeline is not about aging
> A company cannot run a registrational trial for "aging" because no regulator recognizes it as a treatable condition and no [[aging-biomarkers|surrogate endpoint]] has been qualified. Every well-capitalized longevity company therefore ends up developing drugs for specific diseases and arguing that the underlying biology generalizes. One route around the constraint has been veterinary: [[loyal]] is pursuing a lifespan indication in dogs, which the animal-drug regulator will entertain and its human counterpart will not. Calico's pipeline is the clearest illustration of the constraint.

## Funding and structure

Calico is a subsidiary of Alphabet and does not publish financial statements. Alphabet's filings have disclosed its holdings in the "Other Bets" segment without breaking out Calico specifically. Public estimates of total commitment from Alphabet and AbbVie combined run to several billion dollars over a decade, which places it in the same order of magnitude as Altos Labs.

Its structure is unusual in a way that cuts both directions. Independence from Alphabet's core business gave it a long horizon; dependence on AbbVie for development gave it a partner whose priorities are pharmaceutical rather than gerontological, and when those priorities shifted the collaboration narrowed accordingly.

## Reception

Criticism of Calico centres on output. For a decade of work at pharmaceutical scale the public record consists of a respectable but not extraordinary body of papers, no approved drug, and no aging-specific programme. Researchers in the field have noted that the secrecy makes it impossible to distinguish between slow progress and undisclosed progress, and that a company drawing heavily on academic talent while publishing selectively imposes a cost on the field it does not repay.

A more sympathetic reading is that Calico did exactly what it said it would. It funded basic biology for years without a product, produced at least two results — the naked mole-rat mortality analysis and the heritability correction — that changed what specialists believe, and reached the clinic with a mechanism-driven drug. Measured against the founding rhetoric about solving death it has failed; measured against a realistic prior for a preclinical research organization, it is unremarkable in both directions. Nothing it has produced has been shown to match the benefit of [[exercise-and-aging|exercise]], which remains the comparator any candidate geroprotector has to beat, and it has not attempted the kind of repurposing trial that [[rapamycin]] and [[senolytics]] advocates have pursued with far less money.

The company also serves as a cautionary reference for the [[longevity-dividend]] argument. Money and time have been supplied at a level the field had never previously seen, and the constraint that bound was not funding but the absence of a validated target and a measurable endpoint.

## Outlook

The signals worth watching are whether the eIF2B programme produces a positive phase-three result in ALS, which would validate the discovery organization independently of aging; whether Calico discloses a programme aimed at one of the [[hallmarks-of-aging]] as revised in 2023 rather than at a named disease;[^lopez2023] and whether Alphabet's commitment survives the narrowing of the AbbVie relationship.

The larger question Calico poses is not about Calico. If more than a decade of well-funded, competently run work on the biology of aging has not produced a candidate intervention against aging itself, the constraint is unlikely to be money, and the field's assumption that scale was the missing ingredient — an assumption [[retro-biosciences]], [[newlimit]] and Altos are all now testing again with fresh capital — needs a better justification than it currently has.

## See also

- [[altos-labs]]
- [[geroscience-hypothesis]]
- [[negligible-senescence]]
- [[hallmarks-of-aging]]
- [[retro-biosciences]]
- [[buck-institute]]
- [[aging-biomarkers]]
- [[longevity-dividend]]

## References

[^ruby2018]: `paper` Ruby, J.G., Smith, M. and Buffenstein, R. "Naked Mole-Rat mortality rates defy Gompertzian laws by not increasing with age." *eLife*, 2018. {The records come from captive animals in laboratory colonies, not from a wild population.}
[^ruby2018b]: `paper` Ruby, J.G. et al. "Estimates of the Heritability of Human Longevity Are Substantially Inflated due to Assortative Mating." *Genetics*, 2018. {Built from assembled public family trees rather than genotyped cohorts, so it revises a statistical estimate rather than implicating any gene.}
[^lopez2023]: `paper` López-Otín, C., Blasco, M.A., Partridge, L., Serrano, M. and Kroemer, G. "Hallmarks of aging: An expanding universe." *Cell*, 2023.

==============================================================================
ARTICLE: caloric-restriction
TITLE:   Caloric restriction
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/caloric-restriction
SOURCE:  https://futurehumanwiki.com/raw/caloric-restriction
==============================================================================

---
title: "Caloric restriction"
slug: "caloric-restriction"
type: "intervention"
status: "experimental"
horizon: "present"
trl: 6
categories: ["longevity"]
tags: ["dietary restriction", "aging", "fasting", "calerie", "nutrient sensing", "clinical trials"]
summary: "Sustained reduction of energy intake without malnutrition, the oldest and most reproducible lifespan-extending intervention in animals and the least proven in humans."
updated: "2026-07-27"
humanEvidence: "CALERIE, the only randomized trial in healthy non-obese adults, ran two years at roughly half the intended restriction and moved cardiometabolic and immune markers; no human lifespan data exist, and the two rhesus studies disagree."
access: "Requires no prescription and nothing to buy, since it is a change in eating; commercial fasting-mimicking formulations are sold direct to consumers."
reversibility: "reversible"
issues: ["Fasting-mimicking diet claims are unsourced; the paragraph needs a named trial."]
---

```infobox
{
  "caption": "Dietary intervention",
  "rows": [
    { "label": "First reported", "value": "1935 (McCay et al., rats)" },
    { "label": "Typical protocol", "value": "20–40% below ad libitum intake" },
    { "label": "Proven in", "value": "Yeast, worms, flies, rodents" },
    { "label": "Primate evidence", "value": "Conflicting" },
    { "label": "Human trial", "value": "CALERIE, 2 years, biomarker endpoints" },
    { "label": "Human lifespan effect", "value": "Unknown" },
    { "label": "Readiness", "value": "TRL 6" }
  ]
}
```

**Caloric restriction** is a sustained reduction in energy intake, usually 20 to 40 per cent below
what an animal would eat freely, while maintaining adequate vitamins, minerals, and protein. In
laboratory rodents it remains the most reproducible intervention in biogerontology, extending median
and maximum lifespan and delaying most age-related pathologies. In primates the picture is
genuinely conflicted, and in humans the evidence stops at intermediate biomarkers measured over two
years.

## Origins

Underfeeding and longevity were linked before anyone had a mechanism. Clive McCay's 1935 experiments
at Cornell showed that rats fed a restricted but nutritionally complete diet from weaning grew more
slowly, matured later, and lived substantially longer than freely fed controls.[^mccay1935] For
decades the effect was attributed to slowed growth, then to reduced metabolic rate, then to reduced
oxidative damage. None of those explanations survived scrutiny; restricted animals do not, in
general, have lower mass-specific metabolic rates, and the free-radical account has weakened alongside
the wider retreat of that theory in [[mitochondrial-dysfunction|mitochondrial biology]].

The current framing is nutrient sensing, which appears in most catalogues of the
[[hallmarks-of-aging]] as a distinct axis. That framing owes much to invertebrate genetics:
[[cynthia-kenyon|Cynthia Kenyon's]] laboratory showed that partial loss of a single
insulin/IGF-1-family receptor doubles nematode lifespan, which gave nutrient sensing a demonstrated
control point rather than a correlation. Restriction reduces signalling through insulin and IGF-1,
lowers mTORC1 output, activates AMPK, and raises [[autophagy]], engaging the same targets addressed
pharmacologically by [[rapamycin]] and [[metformin|biguanides]]. It also lowers markers of
[[inflammaging]] and, in rodents, reduces the burden of [[cellular-senescence|senescent cells]] that
[[senolytics|senolytic drugs]] are designed to remove. Work in flies and mice complicates even this: restricting protein, or the
single amino acid methionine, reproduces much of the effect without reducing calories at all, which
suggests the operative variable is nutrient composition rather than energy as such.[^solonbiet2014]

## The primate studies

Two long-running rhesus macaque experiments produced opposite headline results, and reconciling them
is the most instructive episode in the field.

```compare
{
  "columns": ["Wisconsin (WNPRC)", "National Institute on Aging"],
  "rows": [
    { "label": "Control feeding", "values": ["Fed ad libitum", "Portion-controlled, not ad libitum"] },
    { "label": "Diet base", "values": ["Purified, higher sucrose", "Natural-ingredient, lower sucrose"] },
    { "label": "Age at onset", "values": ["Adult animals", "Cohorts starting in youth and in old age"] },
    { "label": "Headline survival result", "values": ["Fewer age-related deaths", "No significant survival benefit"] },
    { "label": "Health measures", "values": ["Improved", "Improved in several measures"] }
  ]
}
```

A joint reanalysis by both teams concluded that the divergence came largely from the
controls and the diets rather than from the restriction.[^mattison2017] The NIA control animals were
already portion-fed and lean, so the study compared moderate restriction with mild restriction, not
with overfeeding. The Wisconsin control diet was higher in sucrose. The lesson is that "caloric
restriction extends lifespan" may in practice mean "avoiding chronic overfeeding on a poor diet
extends lifespan", which is a considerably weaker and more ordinary claim.

## Human evidence

CALERIE remains the only randomized controlled trial of sustained restriction in healthy,
non-obese adults. Participants were asked for 25 per cent restriction over two years and achieved
roughly half that on average, which is itself a finding about adherence. The intervention improved
cardiometabolic risk markers, reduced inflammatory markers, and produced changes in thymic tissue
consistent with reduced immune aging.[^spadaro2022] It also reduced bone mineral density and lean
mass.

A secondary analysis applying [[epigenetic-clock|DNA-methylation clocks]] found a small slowing of
one pace-of-aging measure and no significant change in the other clocks tested.[^waziry2023] The result is
frequently reported as evidence that restriction slows human aging. It is better described as
evidence that a two-year dietary change moves one composite predictor by a small amount, with the
relationship between that predictor and any clinical outcome still unestablished — the general
problem addressed under [[aging-biomarkers]] and [[biological-age]].

> [!caution] What has not been shown
> No human study has tested whether caloric restriction extends lifespan, and none realistically can:
> the trial would run for decades with an intervention almost nobody sustains. Human evidence is
> restricted to surrogate measures over a few years.

## Variants

Because sustained restriction is difficult, most current interest is in schedules that reproduce parts
of the signal intermittently.

*Time-restricted eating* confines intake to a window of roughly eight to ten hours. Randomized trials
in humans have generally found weight loss no greater than matched calorie reduction, and at least one
found a concerning share of the loss came from lean mass.[^lowe2020]

*Alternate-day and periodic fasting* produce larger metabolic swings. In the largest controlled
lifespan study in genetically diverse mice, both fasting schedules and graded restriction extended
lifespan, with the most restricted group living longest; notably, the degree of metabolic improvement
did not predict which animals lived longer, while traits related to resilience and genetic background
did.[^difrancesco2024] That dissociation undercuts the common assumption that a favourable metabolic
panel is the mechanism of benefit.

*Fasting-mimicking diets* use several days per month of a low-calorie, low-protein formulation
intended to trigger fasting responses while permitting some food. Trials report improvements in
cardiometabolic risk factors; claims about reduced biological age rest on the same surrogate measures
whose validity is in question.

## Tradeoffs and risks

Restricted animals are cold-intolerant, slower to heal wounds, and more vulnerable to some infections
and to physical stress. In humans the documented costs of sustained restriction include loss of bone
density and lean mass, reduced strength, cold sensitivity, persistent hunger, reduced libido, and
menstrual disruption — with implications for [[reproductive-longevity|reproductive function]] that
are rarely discussed alongside the longevity claims. Severe restriction interacts badly with the
sarcopenia of later life, so the population most likely to want a geroprotector is the population
least suited to this one.

The effect is also not universal. In panels of genetically diverse mice, restriction shortens life in
a minority of strains. There is no reason to assume humans are uniform where mice are not, and no
biomarker exists to identify who responds.

> [!debate] Does the benefit shrink with lifespan?
> Restriction adds a large proportional gain in short-lived species and a smaller, contested one in
> primates. One reading is that it works by triggering a famine-response programme whose payoff scales
> with how quickly an organism's reproductive schedule can be deferred, in which case a species that
> already lives eighty years has little left to gain. This bears directly on how much weight
> [[blue-zones|dietary explanations of exceptional longevity]] deserve.

## Outlook

The practical trajectory has moved away from restriction itself and toward compounds that engage its
downstream signalling without requiring hunger — mTOR inhibitors, AMPK activators, and the wider set
of nutrient-sensing drugs discussed under the [[geroscience-hypothesis]]. Separately,
[[glp-1-receptor-agonists]] have made sustained reductions in intake pharmacologically achievable,
though they act mainly on excess adiposity rather than holding a lean animal below its normal
intake. Comparative work on
[[negligible-senescence|exceptionally long-lived species]] and on the genetics of restriction
response may eventually explain why some individuals benefit and others do not.

What remains unresolved is the counterfactual that matters most for people: whether restriction adds
anything to [[healthspan]], in a well-nourished adult of normal weight, beyond what
[[exercise-and-aging|regular exercise]] and the avoidance of overfeeding already deliver. CALERIE was
not designed to answer that, and no trial currently planned will.

## See also

- [[autophagy]]
- [[rapamycin]]
- [[hallmarks-of-aging]]
- [[exercise-and-aging]]
- [[aging-biomarkers]]
- [[healthspan]]
- [[blue-zones]]
- [[geroscience-hypothesis]]

## References

[^mccay1935]: `paper` McCay, C.M., Crowell, M.F., Maynard, L.A. "The effect of retarded growth upon the length of life span and upon the ultimate body size." *Journal of Nutrition*, 1935. {The rats were restricted from weaning, so retarded growth and reduced intake are confounded in the original design.}
[^solonbiet2014]: `paper` Solon-Biet, S.M. et al. "The ratio of macronutrients, not caloric intake, dictates cardiometabolic health, aging, and longevity in ad libitum-fed mice." *Cell Metabolism*, 2014. {The mice were fed freely throughout, which is what isolates macronutrient ratio from any reduction in energy intake.}
[^mattison2017]: `paper` Mattison, J.A. et al. "Caloric restriction improves health and survival of rhesus monkeys." *Nature Communications*, 2017.
[^spadaro2022]: `paper` Spadaro, O. et al. "Caloric restriction in humans reveals immunometabolic regulators of health span." *Science*, 2022.
[^waziry2023]: `paper` Waziry, R. et al. "Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial." *Nature Aging*, 2023.
[^lowe2020]: `paper` Lowe, D.A. et al. "Effects of time-restricted eating on weight loss and other metabolic parameters in women and men with overweight and obesity: the TREAT randomized clinical trial." *JAMA Internal Medicine*, 2020. {A short trial in adults with overweight or obesity measuring weight and body composition; no aging endpoint was assessed.}
[^difrancesco2024]: `paper` Di Francesco, A. et al. "Dietary restriction impacts health and lifespan of genetically diverse mice." *Nature*, 2024.

==============================================================================
ARTICLE: cancer
TITLE:   Cancer
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/cancer
SOURCE:  https://futurehumanwiki.com/raw/cancer
==============================================================================

---
title: "Cancer"
slug: "cancer"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity", "foundations"]
tags: ["cancer", "aging", "oncology", "somatic mutation", "mortality", "mechanisms", "risk"]
summary: "The class of diseases in which a somatic cell lineage escapes the controls on its own division, and the age-related cause of death that constrains most life-extension arguments."
updated: "2026-07-28"
humanEvidence: "Incidence rises steeply with age in every human population measured and US mortality rates have fallen by about a third since 1991; the claim that slowing aging would lower cancer risk has not been tested in people."
issues: ["No account of how carcinogenicity is tested in geroscience trials, which the longevity section assumes."]
---

```infobox
{
  "caption": "Disease class and constraint on lifespan",
  "rows": [
    { "label": "Kind of thing", "value": "Somatic clonal disease" },
    { "label": "Principal risk factor", "value": "Age" },
    { "label": "New cases, 2022", "value": "~20 million worldwide" },
    { "label": "Deaths, 2022", "value": "~9.7 million worldwide" },
    { "label": "Organizing framework", "value": "Hallmarks of cancer" },
    { "label": "First stated", "value": "2000, revised 2011 and 2022" },
    { "label": "If eliminated", "value": "~3 years of life expectancy" }
  ]
}
```

**Cancer** is the class of diseases in which a somatic cell lineage escapes the controls on its own division and invades tissue that is not its own. It is among the leading causes of death worldwide, and it is the disease whose incidence climbs most sharply with age. That combination puts it at the centre of nearly every argument this wiki makes about extending human life: it is the ceiling that life-extension arithmetic runs into, the risk that cell and gene therapies have to price in, and the reason two of the [[hallmarks-of-aging]] are usually explained as tumour suppression in the first place.

```keyfacts
[
  { "value": "~20M", "label": "New cases worldwide in 2022", "note": "GLOBOCAN estimate, including non-melanoma skin cancer" },
  { "value": "34%", "label": "Fall in the US cancer death rate", "note": "from the 1991 peak through 2023" },
  { "value": "~3 yrs", "label": "Life expectancy gained if all cancer deaths ended", "note": "a 1990 cause-elimination estimate for the United States" }
]
```

## An age-related disease

Cancer incidence rises as a steep power of age. Armitage and Doll fitted that relationship in 1954 and read the log-log slope as evidence that carcinogenesis requires six or seven rate-limiting steps in a single cell lineage, which is why risk accumulates rather than arriving at a threshold.[^armitage1954] The multi-stage picture has been refined repeatedly since, but the demographic fact it was built on has not moved: most cancers are diagnosed in people over 65, and the age curve is the single strongest predictor in oncology.

The pattern is not uniform. The American Cancer Society's 2026 statistical report projects roughly 2.1 million new US cases and 626,000 deaths, with 12% of diagnoses in people under 50 and colorectal cancer incidence in that group rising year on year.[^siegel2026] Part of the most recent acceleration is attributed to screening beginning at 45 rather than 50, which finds disease earlier without changing how much of it there is; the longer-running rise that preceded the change in screening age is not accounted for that way. Early-onset disease is a reminder that "cancer is a disease of aging" describes a rate, not a rule.

> [!key] Why this matters here
> Eliminating every cancer death would add roughly three years to US life expectancy at birth, because the people saved would die of something else soon afterwards.[^olshansky1990] That figure is the origin of the [[geroscience-hypothesis|geroscience]] argument: attacking one disease at a time buys little, because the underlying risk is shared. It also sets the constraint in the other direction. Any intervention that slows aging must not raise cancer incidence, because cancer is the failure mode with the shortest fuse.

## The mechanism

Cancer is somatic evolution: mutation supplies variation, and clonal selection does the rest. The surprise of the last decade is how much of the first half is already present in healthy tissue. Ultra-deep sequencing of normal, sun-exposed human eyelid skin found mutation burdens comparable to those in many tumours, with drivers of squamous cell carcinoma under strong positive selection in a substantial fraction of ordinary skin cells. Similar clonal expansions appear in normal oesophagus, endometrium and blood. In blood the phenomenon has a name and a prognosis: clonal haematopoiesis, present in a rising share of people through later life, carries an elevated risk of blood cancer and, unexpectedly, of cardiovascular death.[^somaticclones] Mutation is therefore necessary and nowhere near sufficient. What separates a large mutant clone from a tumour is the tissue context that lets it keep going, which is the same context that ages.

The capabilities a cell must acquire were codified by Hanahan and Weinberg in 2000 and extended twice since. The 2022 revision added phenotypic plasticity, non-mutational epigenetic reprogramming, polymorphic microbiomes, and senescent cells as further dimensions.[^hanahan2022] Two of those overlap directly with the aging literature: the epigenetic changes that [[epigenetic-clock|clocks]] measure, and the arrested cells described under [[cellular-senescence]].

## Peto's paradox

If every cell division carries a risk of transformation, a whale should be riddled with tumours and a mouse should be almost immune. Peto noticed in 1975 that cancer incidence across species tracks neither body size nor lifespan, an observation now called Peto's paradox. Large, long-lived animals must therefore have evolved additional suppression. Some of it has been found: African elephants carry about twenty copies of the tumour-suppressor gene *TP53* against a human single copy, and their cells commit to apoptosis more readily after DNA damage.[^peto] Naked mole-rats, one of the animals discussed under [[negligible-senescence]], resist tumours by other means again.

The paradox is genuinely encouraging for the field: cancer suppression is evolvable, adjustable, and evidently not at its ceiling in humans. It is also a caution. Everything comparative biology has turned up so far is germline architecture built over millions of years, not a module that can be added to an adult body.

## Where longevity work meets oncology

Most proposals for slowing or reversing aging touch cancer risk somewhere, and usually in both directions.

**Telomerase.** Nearly all human tumours restore telomere maintenance, so the attrition described under [[telomeres-and-telomerase]] is a real tumour-suppressive barrier, and the division limit [[leonard-hayflick|Hayflick]] found in cultured human cells is one of the things standing between a mutant clone and a tumour. Mendelian randomization studies find that inherited longer telomeres raise the risk of several cancers. Adding telomerase across an adult body means pushing on a trait that selection appears to have already compromised.

**Senescence.** The arrest is antagonistic pleiotropy in its clearest form. It stops damaged cells from becoming tumours; the secretory phenotype it produces then promotes tumour growth in the surrounding tissue. [[senolytics|Clearing senescent cells]] therefore has a plausible anti-cancer rationale and a plausible pro-cancer one, and the human trials run so far have been too small and too short to speak to either.

**Reprogramming.** Continuous expression of the [[yamanaka-factors]] produces teratomas in mice, which is why the field works on [[partial-reprogramming|transient, partial protocols]] and why the dose-and-duration question is a safety question rather than an optimisation one. [[epigenetic-reprogramming]] is the intervention class in which the cancer risk is most explicitly the rate-limiting problem.

**Growth signalling.** People with growth hormone receptor deficiency have been reported to develop cancer and diabetes at strikingly low rates without living longer,[^guevara2011] which is the cleanest human hint that dialling down growth signalling trades disease incidence against something else. [[rapamycin]] inhibits a pathway that oncology already targets, and [[metformin]]'s observational cancer signal is one of the reasons it was proposed as a geroprotector at all.

**Delivered genes and cells.** Insertional mutagenesis is the classic hazard of [[somatic-gene-therapy]], and it is not historical: in January 2024 the US Food and Drug Administration required a boxed warning about T-cell malignancies on every approved CAR-T product.[^fda2024] The [[crispr-off-target-effects|off-target and structural consequences]] of editing carry the same concern into the CRISPR era.

> [!debate] Would slowing aging reduce cancer?
> The geroscience prediction is that it would, because a younger tissue environment suppresses clonal expansion and a competent immune system removes transformed cells. The counter-argument is that several mechanisms proposed as geroprotective — telomere maintenance, senescence clearance, restored stem-cell proliferation — are the same mechanisms tumours exploit. No human trial has been powered to settle it, and the mouse evidence points both ways depending on the strain and the intervention.

## What has actually worked

The US cancer death rate has fallen by about a third from its 1991 peak, an estimated 4.8 million deaths averted.[^siegel2026] Most of that is not new biology. It is the collapse in smoking, and after it screening and better conventional treatment.

Immunotherapy is the genuine mechanistic addition. Checkpoint blockade moved from a survival signal in metastatic melanoma to standard care across several tumour types within a decade, and engineered T cells produced remission rates in relapsed paediatric leukaemia that chemotherapy had not.[^immuno] Both are also the source of the safety signals above, and neither works for most patients with most cancers.

Screening is where the field's own evidence is least comfortable. South Korea's thyroid cancer incidence rose sharply after ultrasound screening spread, with no matching change in thyroid cancer mortality, the clearest documented case of overdiagnosis in oncology. Multi-cancer early detection tests, which look for tumour-derived DNA in blood, are the current version of the same hope and the same question; the first large prospective study reported test performance and diagnostic workload rather than any effect on deaths.[^screening] Detection is not the same as benefit, a distinction that recurs throughout [[consumer-blood-testing]] and [[aging-biomarkers|biomarker]] work.

## Open problems

Metastasis, not the primary tumour, is what kills in the large majority of cases, and it remains the least tractable part of the disease. Resistance is an evolutionary process rather than a pharmacological one, which is why durable responses are rarer than initial ones. Prevention research attracts a small share of cancer funding, which many in the field argue is the wrong allocation given that most of the mortality decline so far came from people not starting to smoke rather than from anything given to a patient.

Survivorship is the problem success created. Chemotherapy and radiation damage gonadal tissue, so freezing eggs, sperm or embryos through [[in-vitro-fertilisation|IVF]] before treatment is now routine for patients of reproductive age. Cytotoxic therapy also drives cells into the arrested state described above, which is one proposed source of the accelerated functional decline reported in long-term survivors.

For the arguments this wiki tracks, the unresolved question is narrower and sharper. Nobody has shown, in humans, that an intervention aimed at aging changes cancer incidence in either direction. Until a trial is run long enough and large enough to measure it, claims about [[longevity-escape-velocity|radical life extension]] are claims about a body whose dominant late-life failure mode has not been addressed. Cancer is reported to be an uncommon cause of death among the [[maximum-human-lifespan|oldest people on record]], which some read as evidence that late-life cancer risk plateaus and others read as evidence that they died of something faster first.

## See also

- [[hallmarks-of-aging]]
- [[cellular-senescence]]
- [[telomeres-and-telomerase]]
- [[geroscience-hypothesis]]
- [[immunosenescence]]
- [[senolytics]]
- [[somatic-gene-therapy]]
- [[compression-of-morbidity]]

## References

[^armitage1954]: `paper` Armitage, P. and Doll, R. "The Age Distribution of Cancer and a Multi-stage Theory of Carcinogenesis." *British Journal of Cancer*, 1954. {The six-or-seven-stage figure is inferred from the slope of incidence against age, not from any identified molecular step.}
[^siegel2026]: `report` Siegel, R. L., Kratzer, T. B., Wagle, N. S., Sung, H. and Jemal, A. "Cancer statistics, 2026." *CA: A Cancer Journal for Clinicians*, 2026; and Bray, F. et al. "Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries." *CA: A Cancer Journal for Clinicians*, 2024. {Projections for the current year are modelled from registry data several years old, and the GLOBOCAN world figures are estimates built from uneven national registration.}
[^olshansky1990]: `paper` Olshansky, S. J., Carnes, B. A. and Cassel, C. "In Search of Methuselah: Estimating the Upper Limits to Human Longevity." *Science*, 1990. {A cause-elimination life table for the United States around 1990; the arithmetic assumes the eliminated deaths are redistributed to other causes at then-current rates.}
[^somaticclones]: `paper` Martincorena, I. et al. "High burden and pervasive positive selection of somatic mutations in normal human skin." *Science*, 2015; and Jaiswal, S. et al. "Age-Related Clonal Hematopoiesis Associated with Adverse Outcomes." *New England Journal of Medicine*, 2014. {The skin study sampled sun-exposed eyelid tissue from four people, which is the high end of somatic mutation burden rather than a typical organ.}
[^hanahan2022]: `paper` Hanahan, D. "Hallmarks of Cancer: New Dimensions." *Cancer Discovery*, 2022. {A single-author review; the four additions are proposed as emerging dimensions rather than agreed by any consensus body.}
[^peto]: `paper` Peto, R., Roe, F. J. C., Lee, P. N., Levy, L. and Clack, J. "Cancer and ageing in mice and men." *British Journal of Cancer*, 1975; and Abegglen, L. M. et al. "Potential Mechanisms for Cancer Resistance in Elephants and Comparative Cellular Response to DNA Damage in Humans." *JAMA*, 2015. {Elephant cancer rates come from zoo necropsy records, and the apoptosis comparison is in cultured lymphocytes.}
[^guevara2011]: `paper` Guevara-Aguirre, J. et al. "Growth hormone receptor deficiency is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans." *Science Translational Medicine*, 2011. {A small Ecuadorian cohort; disease incidence fell but lifespan did not rise.}
[^fda2024]: `regulator` US Food and Drug Administration. "FDA Requires Boxed Warning for T cell Malignancies Following Treatment with BCMA-Directed or CD19-Directed Autologous Chimeric Antigen Receptor T cell Immunotherapies." Safety communication, January 2024.
[^immuno]: `paper` Hodi, F. S. et al. "Improved Survival with Ipilimumab in Patients with Metastatic Melanoma." *New England Journal of Medicine*, 2010; and Maude, S. L. et al. "Tisagenlecleucel in Children and Young Adults with B-Cell Lymphoblastic Leukemia." *New England Journal of Medicine*, 2018. {In the melanoma trial median survival rose from roughly six months to roughly ten; the durable responders are a minority tail.}
[^screening]: `paper` Ahn, H. S., Kim, H. J. and Welch, H. G. "Korea's Thyroid-Cancer 'Epidemic' — Screening and Overdiagnosis." *New England Journal of Medicine*, 2014; and Schrag, D. et al. "Blood-based tests for multicancer early detection (PATHFINDER): a prospective cohort study." *The Lancet*, 2023. {PATHFINDER is a single-arm study of test performance and diagnostic workup, not a trial of outcomes.}

==============================================================================
ARTICLE: casgevy
TITLE:   Casgevy
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/casgevy
SOURCE:  https://futurehumanwiki.com/raw/casgevy
==============================================================================

---
title: "Casgevy"
slug: "casgevy"
type: "intervention"
status: "emerging"
horizon: "present"
trl: 9
categories: ["genetics"]
tags: ["crispr", "sickle cell disease", "gene therapy", "haemoglobin", "drug pricing", "clinical trials"]
summary: "The first approved CRISPR-based medicine, which disrupts a fetal-haemoglobin repressor in a patient's own blood stem cells to treat sickle cell disease and beta thalassaemia."
updated: "2026-07-27"
humanEvidence: "In single-arm trials with no control group, most sickle cell recipients went at least twelve months without a vaso-occlusive crisis and most thalassaemia recipients stopped transfusions; follow-up is short."
access: "Approved in the United Kingdom in November 2023 and the United States shortly after; the US list price is about 2.2 million dollars, delivery is confined to authorised transplant centres, and early uptake was slow."
reversibility: "irreversible"
issues: ["The Casgevy and Lyfgenia list prices are stated without a source.", "The claim that uptake was slow in the first two years carries no source."]
---

```infobox
{
  "caption": "Approved gene-editing therapy",
  "rows": [
    { "label": "Generic name", "value": "Exagamglogene autotemcel" },
    { "label": "Developers", "value": "Vertex and CRISPR Therapeutics" },
    { "label": "Target", "value": "BCL11A erythroid enhancer" },
    { "label": "Editing tool", "value": "Cas9 ribonucleoprotein", "link": "/wiki/crispr-cas9" },
    { "label": "Route", "value": "Ex vivo autologous HSC infusion" },
    { "label": "First approval", "value": "UK, November 2023" },
    { "label": "US list price", "value": "About $2.2 million" },
    { "label": "Indications", "value": "Sickle cell disease, beta thalassaemia" }
  ]
}
```

**Casgevy** (exagamglogene autotemcel) is a one-time cell therapy in which a patient's own blood-forming stem cells are edited outside the body with [[crispr-cas9]] and returned, raising their production of fetal haemoglobin enough to compensate for a defective adult haemoglobin gene. Authorised in the United Kingdom in November 2023 and in the United States across December 2023 and January 2024, it is the first medicine based on CRISPR editing to reach regulatory approval anywhere. It treats two inherited disorders of the beta-globin gene: sickle cell disease and transfusion-dependent beta thalassaemia.

## How it works

Casgevy does not repair the mutation that causes either disease. It works around it.

Human beings make two different beta-like globin chains over a lifetime. Fetal haemoglobin, containing gamma-globin, dominates before birth and is switched off in the months afterwards, when adult beta-globin takes over. People who carry a beta-globin mutation are therefore healthy in utero and become ill only once the switch completes. A minority of people carry variants that keep fetal haemoglobin switched on into adulthood — hereditary persistence of fetal haemoglobin — and if they also inherit sickle mutations, their disease is markedly milder. That natural experiment defined the therapeutic target.

The switch is enforced by the transcription factor BCL11A, identified as the repressor of fetal haemoglobin by Stuart Orkin's group in 2008.[^sankaran2008] Knocking out BCL11A entirely is unacceptable — it is needed in B cells and in the brain — but the gene carries an enhancer, active only in the erythroid lineage, whose disruption lowers BCL11A in red cell precursors and nowhere else.[^bauer2013] A saturating mutagenesis screen mapped the critical bases within that enhancer, giving a small target that a single guide RNA can hit.[^canver2015]

The manufacturing sequence is that of a stem cell transplant with an editing step inserted. Stem cells are mobilised from the marrow with plerixafor and collected by apheresis, often over several cycles. In the factory, CD34+ cells are electroporated with a Cas9 ribonucleoprotein complex carrying the guide; the protein enters as a preassembled complex rather than being encoded on a viral genome, so it acts within hours and is then degraded. Nothing persists in the cell, which is why no [[aav-vectors|viral vector]] is involved and why immunity to a capsid is not a consideration. The edited cells are frozen and tested. The patient then receives myeloablative busulfan to empty the marrow niche, and the edited cells are infused.

The edit itself is a blunt one — a double-strand break repaired by error-prone end joining, producing small insertions and deletions that happen to destroy an enhancer motif. Competing programmes aim at the same biology with gentler chemistry: [[base-editing]] can install the exact single-base changes found in people with naturally persistent fetal haemoglobin, and [[epigenome-editing]] could in principle silence the enhancer without breaking DNA at all. Correcting the sickle mutation directly, which [[prime-editing]] makes conceivable, would be the more complete solution and is further from the clinic.

## Development history

```timeline
[
  { "year": "2008", "title": "BCL11A identified", "text": "Orkin's laboratory shows that BCL11A is the developmental-stage repressor that silences fetal haemoglobin after birth." },
  { "year": "2013–2015", "title": "The enhancer as a target", "text": "The erythroid-specific BCL11A enhancer is characterised and then dissected base by base with Cas9, defining a narrow, lineage-restricted target." },
  { "year": "2019", "title": "First patients dosed", "text": "Victoria Gray in the United States and a beta thalassaemia patient in Germany receive edited cells in the CLIMB trials." },
  { "year": "2021", "title": "First results published", "text": "The New England Journal of Medicine reports elimination of vaso-occlusive crises and transfusion independence in the first treated patients." },
  { "year": "2023", "title": "Approval", "text": "The UK MHRA authorises the therapy in November; the FDA follows in December for sickle cell disease, and in January 2024 for beta thalassaemia." },
  { "year": "2024–2025", "title": "Reimbursement fights", "text": "European and UK health-technology bodies negotiate managed-access arrangements; uptake remains slow and confined to authorised transplant centres." }
]
```

## Clinical results

In the pivotal single-arm trials, nearly all evaluable sickle cell patients went at least twelve consecutive months without a severe vaso-occlusive crisis, the painful episodes of blocked microcirculation that define the disease and drive its mortality.[^frangoul2024] Most thalassaemia patients became independent of the regular red cell transfusions they had needed since childhood.[^locatelli2024] Fetal haemoglobin typically rose to a large fraction of total haemoglobin and stayed there, and edited alleles persisted in the marrow, indicating that true long-term stem cells had been modified rather than short-lived progenitors.

Two caveats belong beside those numbers. The trials had no control arm, so the comparison is with each patient's own history rather than with a randomised alternative. And follow-up is short relative to the claim being made. A therapy sold as a cure needs decades of surveillance to establish that it is one, particularly given a theoretical risk of clonal expansion from an edited stem cell.

> [!caution] What "cure" means here
> The causative mutation is untouched. Every red cell a treated patient makes still carries it, and any child they have inherits it exactly as before. Casgevy suppresses the phenotype by reactivating a fetal gene; it does not correct the genome's error, and it changes nothing about inheritance in the way [[germline-editing]] would.

## The conditioning burden

The editing is the easy part. The hard part is busulfan.

To make room for edited cells, the existing marrow must be destroyed with high-dose chemotherapy. That carries weeks of neutropenia and infection risk, mucositis, hospitalisation typically measured in a month or more, a high probability of permanent infertility, and a small long-term risk of secondary malignancy. Patients are advised to bank gametes beforehand, which links the therapy to the reproductive questions covered under [[embryo-selection]] and fertility preservation. For a young adult with severe disease the trade may be worth it; for a person with moderate disease it plainly is not, and this is why eligibility is drawn narrowly.

Removing conditioning is the field's main objective. Antibody-based marrow clearance and, further out, editing haematopoietic stem cells directly inside the body using [[lipid-nanoparticles]] targeted to CD117-bearing cells would eliminate the chemotherapy step entirely. Both have shown promise in animals. Neither is an approved treatment as of 2026.

## Price and access

The US list price is about $2.2 million, alongside roughly $3.1 million for Lyfgenia, the lentiviral therapy approved the same day. Total cost of care — mobilisation, apheresis, conditioning, inpatient stay, and follow-up — is higher still. England's health-technology assessor initially declined to recommend the therapy on cost-effectiveness grounds and later agreed managed-access arrangements, first for beta thalassaemia and then for sickle cell disease.

> [!stat] The mismatch
> Roughly 100,000 people in the United States live with sickle cell disease. Global prevalence is in the millions, and the overwhelming majority of affected births occur in sub-Saharan Africa and India, where apheresis and transplant infrastructure is scarce or absent.[^gbd2023]

Uptake in the first two years after approval was slow. Patients must travel to an authorised centre, undergo mobilisation that can fail, and accept myeloablation; payers must find the money; and each course consumes scarce transplant-unit capacity. The result is a therapy that works and that almost nobody receives — an unusually stark instance of the pattern described in [[access-and-inequality]] and [[somatic-gene-therapy]] more generally. It has also sharpened arguments about the governance of who benefits from publicly funded discovery, since the biology underlying the target was worked out largely in academic laboratories.

## Outlook

Casgevy's importance is partly symbolic. It established that a CRISPR-edited cell product can pass a regulator's benefit-risk review, and it made the [[crispr-off-target-effects|off-target analysis package]] for an editing therapy a known, negotiable quantity rather than an open question. Five years after the [[he-jiankui-affair]] had made CRISPR in humans synonymous with recklessness, the approval demonstrated what the ordinary route through [[governance-of-genome-editing|regulatory review]] looks like when it is followed. What follows depends less on editing chemistry than on delivery and logistics. If in vivo editing of blood stem cells works in humans, the same biological target could in principle be addressed with an injection in a district hospital. If it does not, the therapy remains a very expensive procedure available in a few dozen buildings worldwide, and the disease it treats will continue to be managed, for almost everyone who has it, with hydroxyurea and transfusion.

## See also

- [[crispr-cas9]]
- [[somatic-gene-therapy]]
- [[base-editing]]
- [[lipid-nanoparticles]]
- [[germline-editing]]
- [[access-and-inequality]]
- [[crispr-off-target-effects]]
- [[governance-of-genome-editing]]

## References

[^sankaran2008]: `paper` Sankaran, V.G. et al. "Human fetal hemoglobin expression is regulated by the developmental stage-specific repressor BCL11A." *Science*, 2008.
[^bauer2013]: `paper` Bauer, D.E. et al. "An erythroid enhancer of BCL11A subject to genetic variation determines fetal hemoglobin level." *Science*, 2013.
[^canver2015]: `paper` Canver, M.C. et al. "BCL11A enhancer dissection by Cas9-mediated in situ saturating mutagenesis." *Nature*, 2015.
[^frangoul2024]: `paper` Frangoul, H. et al. "Exagamglogene Autotemcel for Severe Sickle Cell Disease." *New England Journal of Medicine*, 2024. {A single-arm trial with no control group; the endpoint was freedom from vaso-occlusive crises, not survival or organ damage.}
[^locatelli2024]: `paper` Locatelli, F. et al. "Exagamglogene Autotemcel for Transfusion-Dependent β-Thalassemia." *New England Journal of Medicine*, 2024. {Also single-arm; transfusion independence is measured against each patient's own prior requirement rather than a randomised comparator.}
[^gbd2023]: `paper` GBD 2021 Sickle Cell Disease Collaborators. "Global, regional, and national prevalence and mortality burden of sickle cell disease, 2000–2021." *The Lancet Haematology*, 2023. {A modelled burden estimate rather than a count; national figures depend heavily on newborn screening coverage.}

==============================================================================
ARTICLE: cellular-senescence
TITLE:   Cellular senescence
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/cellular-senescence
SOURCE:  https://futurehumanwiki.com/raw/cellular-senescence
==============================================================================

---
title: "Cellular senescence"
slug: "cellular-senescence"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity"]
tags: ["aging", "senescence", "sasp", "cell biology", "cancer", "mechanisms"]
summary: "A stable arrest of cell division, triggered by telomere attrition or stress, in which the arrested cell stays metabolically active and secretes inflammatory factors."
updated: "2026-07-27"
humanEvidence: "The arrest was first described in cultured human fibroblasts and the markers appear in aged human tissue; every result showing that clearing senescent cells extends life is from mice, and randomized senolytic trials have not yet shown clinical benefit."
issues: ["The Unity Biotechnology phase 2 osteoarthritis result is stated without a source."]
---

```infobox
{
  "caption": "Cell state in biology of aging",
  "rows": [
    { "label": "First described", "value": "1961" },
    { "label": "Described by", "value": "Leonard Hayflick, Paul Moorhead" },
    { "label": "Defining feature", "value": "Stable growth arrest" },
    { "label": "Common markers", "value": "p16INK4a, p21, SA-β-gal" },
    { "label": "Key output", "value": "SASP" },
    { "label": "Therapeutic angle", "value": "Senolytics, senomorphics" }
  ]
}
```

**Cellular senescence** is a stable exit from the cell cycle in which a cell stops dividing but does not die, remains metabolically active, resists apoptosis, and alters what it secretes. It is one of the [[hallmarks-of-aging]] and one of the clearest cases of antagonistic pleiotropy in mammalian biology: the same arrest that prevents damaged cells from becoming tumours degrades the tissue around them when senescent cells accumulate with age.

## Discovery and definition

Leonard Hayflick and Paul Moorhead reported in 1961 that normal human diploid fibroblasts divide a limited number of times in culture and then stop, contradicting the then-standard belief that vertebrate cells were immortal under good conditions.[^hayflick1961] The limit is now attributed largely to progressive [[telomeres-and-telomerase|telomere shortening]], but replicative exhaustion turned out to be only one route into the state.

Senescence has no single definitive marker, which is a persistent methodological problem. Investigators typically require several features together: absence of proliferation markers, expression of the cyclin-dependent kinase inhibitors p16INK4a or p21CIP1, senescence-associated β-galactosidase activity detected at pH 6, loss of nuclear Lamin B1, and evidence of a persistent DNA damage response.[^dimri1995] Each of these appears in non-senescent contexts. p16 expression, in particular, can be induced in macrophages as a reversible physiological response, which confounds any attempt to quantify senescent burden in tissue by p16 alone.[^hall2017]

> [!note] Terminology
> "Senescence" in this article means the cellular state. It is unrelated in mechanism to organismal senescence — the whole-body decline the word denotes in demography and evolutionary biology — although the two are often conflated in popular coverage.

## Triggers

Replicative senescence follows telomere attrition and the DNA damage response it provokes at uncapped chromosome ends. Oncogene-induced senescence is triggered by aberrant proliferative signalling; Serrano and colleagues showed in 1997 that expressing oncogenic *RAS* in normal human cells drives them into arrest rather than transformation, establishing senescence as a tumour-suppressive barrier.[^serrano1997] Stress-induced premature senescence follows oxidative damage, [[mitochondrial-dysfunction|mitochondrial]] impairment, irradiation or [[proteostasis|proteotoxic]] stress without telomere involvement. Therapy-induced senescence is a common outcome of cytotoxic chemotherapy and contributes to the long-term functional deficits seen in cancer survivors.

Senescence also has programmed roles outside pathology. Two independent groups reported in 2013 that transient senescent populations appear at defined sites during mammalian embryonic development, including the apical ectodermal ridge, where they shape tissue patterning before being cleared by macrophages.[^munoz2013] Senescence is therefore not a failure mode that evolution overlooked; it is a used mechanism whose persistence in aged tissue is the problem.

## The senescence-associated secretory phenotype

The feature that makes senescent cells matter beyond their own arrest is the senescence-associated secretory phenotype, or SASP: a program of secreted interleukins, chemokines, growth factors, and matrix-remodelling proteases characterized by Coppé, Campisi and colleagues in 2008 at the [[buck-institute|Buck Institute]] and Lawrence Berkeley National Laboratory.[^coppe2008] SASP composition varies with the trigger and the cell type; there is no canonical SASP. Cytosolic DNA sensing through the cGAS–STING pathway is one of its principal upstream activators, which links the secretory program to genome instability and to leaked mitochondrial DNA.

The SASP reinforces arrest in the secreting cell, recruits immune cells to clear it, and induces senescence in neighbours — a paracrine spread that lets a small number of senescent cells alter a large volume of tissue. Its chronic version is one of the main proposed sources of [[inflammaging|sterile chronic inflammation]] in old animals, and it impairs the function of nearby progenitors, connecting senescence to [[stem-cell-exhaustion|stem cell exhaustion]]. Transplanting a modest number of senescent cells into young mice is sufficient to produce measurable physical dysfunction that spreads beyond the injection site.[^xu2018]

## Beneficial roles

Senescent cells are not simply debris. Beyond tumour suppression and development, they participate in wound healing: in mice, senescent fibroblasts and endothelial cells appearing at wound sites secrete PDGF-AA and accelerate closure, and animals unable to mount that response heal more slowly.[^demaria2014] Senescence of activated hepatic stellate cells limits liver fibrosis in mouse models. Indiscriminate clearance is therefore not obviously good. Work using genetic ablation of p16-high cells has reported that removing certain populations — including liver sinusoidal endothelial cells — causes tissue damage rather than rejuvenation.[^grosse2020]

## From cell-culture curiosity to drug target

For four decades senescence was treated as an artifact of culture with uncertain relevance in vivo. Two developments changed that. First, reliable in vivo markers and reporter mice made senescent cells visible in tissue. Second, transgenic systems allowed their selective killing.

```timeline
[
  { "year": "1961", "title": "Replicative limit reported", "text": "Hayflick and Moorhead show human fibroblasts stop dividing after a finite number of doublings." },
  { "year": "1995", "title": "A usable marker", "text": "Dimri and colleagues describe senescence-associated β-galactosidase, allowing senescent cells to be stained in tissue." },
  { "year": "2008", "title": "SASP characterized", "text": "Coppé, Campisi and colleagues define the secretory phenotype and show senescent cells act on their neighbours." },
  { "year": "2011", "title": "Clearance benefits mice", "text": "Baker and colleagues use the INK-ATTAC transgene to kill p16-positive cells in a progeroid mouse, delaying several age-related pathologies." },
  { "year": "2015", "title": "First senolytic drugs", "text": "Zhu and colleagues identify dasatinib and quercetin as agents that preferentially kill senescent cells by disabling their anti-apoptotic defences." },
  { "year": "2016", "title": "Lifespan effect in normal mice", "text": "Clearance of p16-positive cells extends median lifespan and improves healthspan in naturally aged mice." }
]
```

The transgenic results — first in a progeroid strain, then in normally aged animals — established that senescent cells are causally involved in mouse aging phenotypes rather than merely present.[^baker] That finding motivated the search for drugs with the same effect, which became the field of [[senolytics]]. Progress from there has been slower than the mouse work implied. Early human studies have been small, open-label, and focused on surrogate measures such as senescent-cell burden in fat biopsies; a phase 2 trial of a locally injected senolytic for osteoarthritis run by Unity Biotechnology did not beat placebo, and a randomized trial of intermittent dasatinib plus quercetin in postmenopausal women missed its primary bone-turnover endpoint.[^farr2024] Senomorphics, which suppress the SASP without killing the cell, are the alternative approach; [[rapamycin]] and JAK inhibitors both blunt SASP output in animal models.

> [!debate] The disagreement
> Whether senescent cells are a primary driver of aging or one downstream consequence among many is unsettled. Clearance experiments show they are causal for specific phenotypes in mice. They do not show that senescence sits upstream of the other hallmarks, and the mouse strains used are not a straightforward model of human aging.

## Open problems

Quantifying senescent burden in living humans is not currently possible with any validated assay, which blocks the obvious trial design: measure burden, remove cells, measure again, follow outcomes. The NIH's Cellular Senescence Network was created in part to build a tissue atlas that would support such measures, and defining senescence in a way that is specific enough for [[aging-biomarkers|biomarker]] use remains its central task.

Heterogeneity is the deeper issue. Senescent cells derived from different tissues by different triggers share arrest but differ in secretome, in the anti-apoptotic pathways they depend on, and probably in whether they help or harm. A senolytic that kills one subtype may leave another untouched, which is a plausible explanation for the gap between mouse results and human trial outcomes. Whether the state is truly irreversible is also in question: [[partial-reprogramming|transient reprogramming]] can restore proliferative capacity to senescent cells in culture, which suggests the arrest is enforced epigenetically rather than written into the genome.

The clinical question that follows is which diseases to target. Senescent-cell involvement has been argued for idiopathic pulmonary fibrosis, osteoarthritis, diabetic kidney disease, atherosclerosis and Alzheimer's disease, and small early-phase studies exist in several of these. The [[geroscience-hypothesis|geroscience]] version of the argument is more ambitious: that clearing senescent cells would delay several such conditions at once, and so would demonstrate a change in the aging process rather than in one disease. Testing that claim requires an outcome measure the field does not yet have, and in its absence senescence therapeutics are assessed one indication at a time on ordinary clinical endpoints rather than on [[biological-age]] or [[healthspan]].

## See also

- [[senolytics]]
- [[hallmarks-of-aging]]
- [[inflammaging]]
- [[telomeres-and-telomerase]]
- [[stem-cell-exhaustion]]
- [[aging-biomarkers]]

## References

[^hayflick1961]: `paper` Hayflick, L., Moorhead, P. S. "The serial cultivation of human diploid cell strains." *Experimental Cell Research*, 1961. {Human cells in culture; the doubling limit is a property of cells in a dish and was never measured in a living person.}
[^dimri1995]: `paper` Dimri, G. P. et al. "A biomarker that identifies senescent human cells in culture and in aging skin in vivo." *PNAS*, 1995.
[^hall2017]: `paper` Hall, B. M. et al. "p16(Ink4a) and senescence-associated β-galactosidase can be induced in macrophages as part of a reversible response to physiological stimuli." *Aging*, 2017. {Mouse macrophages responding to ordinary stimuli, which is why p16 staining alone cannot be read as a count of senescent cells.}
[^serrano1997]: `paper` Serrano, M., Lin, A. W., McCurrach, M. E., Beach, D., Lowe, S. W. "Oncogenic ras provokes premature cell senescence associated with accumulation of p53 and p16INK4a." *Cell*, 1997.
[^munoz2013]: `paper` Muñoz-Espín, D. et al. "Programmed cell senescence during mammalian embryonic development." *Cell*, 2013.
[^coppe2008]: `paper` Coppé, J.-P. et al. "Senescence-associated secretory phenotypes reveal cell-nonautonomous functions of oncogenic RAS and the p53 tumor suppressor." *PLoS Biology*, 2008.
[^xu2018]: `paper` Xu, M. et al. "Senolytics improve physical function and increase lifespan in old age." *Nature Medicine*, 2018.
[^demaria2014]: `paper` Demaria, M. et al. "An essential role for senescent cells in optimal wound healing through secretion of PDGF-AA." *Developmental Cell*, 2014.
[^grosse2020]: `paper` Grosse, L. et al. "Defined p16High senescent cell types are indispensable for mouse healthspan." *Cell Metabolism*, 2020.
[^farr2024]: `paper` Farr, J.N. et al. "Effects of intermittent senolytic therapy on bone metabolism in postmenopausal women: a randomized controlled trial." *Nature Medicine*, 2024. {The trial reported signals in prespecified subgroups, which is hypothesis-generating rather than a demonstration of benefit.}
[^baker]: `paper` Baker, D. J. et al. "Clearance of p16Ink4a-positive senescent cells delays ageing-associated disorders." *Nature*, 2011; and Baker, D. J. et al. "Naturally occurring p16Ink4a-positive cells shorten healthy lifespan." *Nature*, 2016. {Both use an engineered transgene to kill p16-positive cells in mice, not a drug; the 2011 strain is progeroid and the 2016 one normally aged.}

==============================================================================
ARTICLE: closed-loop-life-support
TITLE:   Closed-loop life support
PORTAL:  Space & Extreme Environments
URL:     https://futurehumanwiki.com/wiki/closed-loop-life-support
SOURCE:  https://futurehumanwiki.com/raw/closed-loop-life-support
==============================================================================

---
title: "Closed-loop life support"
slug: "closed-loop-life-support"
type: "technology"
status: "emerging"
horizon: "2030s"
trl: 5
categories: ["space"]
tags: ["life support", "spaceflight", "recycling", "bioregenerative", "mars", "closed ecology"]
summary: "Engineering that regenerates a crew's air, water, and food from their own waste, reducing the resupply mass that otherwise makes long-duration spaceflight impossible."
updated: "2026-07-27"
humanEvidence: "Crews have lived for months on regenerated air and water aboard the ISS and in sealed ground facilities such as BIOS-3; no flown system has closed the food loop, and essentially every calorie eaten in orbit was launched."
access: "Nothing to obtain: these are one-off agency and research systems, and flight-qualified hardware exists only aboard the ISS and in a handful of sealed ground facilities."
reversibility: "reversible"
issues: ["The 98 percent ISS water-recovery figure appears three times without a source."]
---

```infobox
{
  "caption": "Spacecraft and habitat subsystem",
  "rows": [
    { "label": "Function", "value": "Air, water and food regeneration" },
    { "label": "Water recovery on ISS", "value": "~98%" },
    { "label": "Oxygen source", "value": "Water electrolysis" },
    { "label": "Food loop closure", "value": "Effectively zero" },
    { "label": "Largest closed test", "value": "Biosphere 2, 1991–93" },
    { "label": "Readiness", "value": "Air and water flown; food unsolved" }
  ]
}
```

**Closed-loop life support** is the engineering of a habitat that regenerates the consumables its crew uses — oxygen, water, and in the fully closed case food — from metabolic and system waste, rather than carrying or resupplying them. It is a mass problem before it is a biology problem. An unrecycled crew consumes roughly five kilograms of oxygen, water and food per person per day, so a three-year Mars mission for four people implies a resupply mass no launch architecture can deliver. Every fraction of a loop that closes converts consumable mass into hardware mass and power, and the design question is where that trade turns favourable.

```keyfacts
[
  { "value": "~5 kg", "label": "Consumables per person per day", "note": "oxygen, water and food before any recycling" },
  { "value": "~98%", "label": "ISS water recovery", "note": "reached after the brine processor entered service" },
  { "value": "14.2%", "label": "Oxygen concentration in Biosphere 2", "note": "down from 20.9% over sixteen months, requiring injection" }
]
```

## How it works

Four loops are usually treated separately.

The **atmosphere loop** removes carbon dioxide, supplies oxygen, controls trace contaminants and manages pressure and humidity. On the International Space Station, carbon dioxide is captured on regenerable zeolite sorbent beds and oxygen is produced by electrolysing water. A Sabatier reactor closes part of the carbon loop by combining captured carbon dioxide with hydrogen from electrolysis to make methane and water; the methane is vented, so hydrogen, and therefore water, leaves the system permanently. Closing that last step requires methane pyrolysis to recover the hydrogen, which has been developed but not operationally flown at scale.

The **water loop** is the most closed in practice. Urine is processed by vacuum distillation, condensate and hygiene water are collected, and everything passes through multifiltration and catalytic oxidation. The addition of a brine processor, which extracts water from the residual concentrate the distillation leaves behind, raised station recovery to roughly 98 per cent.

The **food loop** is not closed at all. Every calorie consumed in orbit since 1961 has been launched from Earth. Small plant growth facilities aboard the station produce leafy greens, radishes and peppers in quantities that are nutritionally and psychologically welcome and metabolically trivial.

The **waste loop** is largely open. Solid waste is compacted and disposed of; the carbon, nitrogen and phosphorus it contains are not recovered.

## Development history

```timeline
[
  { "year": "1961–1972", "title": "Open-loop era", "text": "Early crewed vehicles carry all consumables and vent carbon dioxide overboard, an approach adequate for missions measured in days." },
  { "year": "1972–1984", "title": "BIOS-3", "text": "The Institute of Biophysics in Krasnoyarsk operates a sealed 315-cubic-metre facility with algal cultivators and wheat phytotrons, running crews for up to six months with near-complete air and water regeneration and roughly half their food produced internally." },
  { "year": "1989", "title": "MELiSSA begins", "text": "ESA starts a multi-decade programme to build a five-compartment microbial and plant loop that converts crew waste back into oxygen, water and edible biomass." },
  { "year": "1991–1993", "title": "Biosphere 2", "text": "Eight people are sealed into a 1.27-hectare structure in Arizona for two years. Oxygen falls steadily, food runs short, and the mission becomes the field's most-cited cautionary case." },
  { "year": "2008–2010", "title": "ISS water and carbon systems", "text": "The urine processor and water recovery system enter service, followed by a Sabatier reactor that recovers water from metabolic carbon dioxide." },
  { "year": "2017–2018", "title": "Lunar Palace 1", "text": "A Beijing facility runs volunteers through a 370-day sealed experiment with plant and insect production, reporting very high closure." },
  { "year": "2021–2023", "title": "Oxygen from Mars", "text": "The MOXIE instrument on Perseverance produces oxygen from atmospheric carbon dioxide by solid-oxide electrolysis, demonstrating in-situ resource use at gram scale." }
]
```

## Current state

ISS life support is a hybrid physicochemical system with essentially no biology in the loop, and it works. Water recovery near 98 per cent and oxygen generation from recovered water mean the station's consumable resupply is dominated by food. That is a satisfactory position for a vehicle in low Earth orbit with regular cargo flights, and an inadequate one for Mars.

Reliability rather than closure is the operational pain point. The oxygen generation and urine processing assemblies have both required repeated maintenance and part replacement, and the spares logistics that supports them is itself a resupply stream. A system that fails every few months is acceptable when replacement parts are ninety minutes away, and not otherwise. Assessments of readiness in this domain frequently note that the relevant [[technology-readiness-level]] for Mars is not about whether a component works but whether it works unattended for a thousand days.

In-situ resource use changes the arithmetic where it is available. The Martian atmosphere is over ninety-five per cent carbon dioxide, and MOXIE demonstrated oxygen production from it at gram scale on the surface.[^hoffman2022] Subsurface water ice would supply hydrogen. A surface base that can mine its own oxygen and water does not need the same degree of loop closure as a transit vehicle, which is one reason mission architectures front-load risk into the cruise phase — the same phase in which [[radiation-hardening-humans|radiation dose]] and [[microgravity-adaptation|deconditioning]] also peak, and the phase that proposals for [[human-hibernation|crew torpor]] target.

## Bioregenerative systems

Closing the food loop requires organisms. Two long-running programmes define the field.

BIOS-3 in Krasnoyarsk demonstrated in the 1970s and 1980s that a sealed facility with algal cultivators and hydroponic wheat could regenerate air and water essentially completely and supply a substantial share of the crew's calories, with small crews and runs of up to six months.[^salisbury1997] Its results remain among the most credible in the literature precisely because the system was small, instrumented, and honest about what it did not close.

ESA's MELiSSA takes a compartmentalised microbial approach: anaerobic liquefaction of waste, photoheterotrophic and nitrifying stages that convert the products into nitrate, a photoautotrophic compartment growing cyanobacteria, and a higher-plant compartment producing food and oxygen for the crew. A pilot plant in Barcelona has operated compartments in series with animal crew analogues. Progress has been steady and slow, which is what a system with this many coupled biological stages should be expected to produce.

Crop area is the binding constraint on full food closure. Estimates cluster in the range of tens of square metres of continuously cropped area per person, and the power demand for artificial lighting typically dominates the entire habitat's energy budget. Cultured animal cells have been proposed as a compact protein source using the methods of [[tissue-engineering]], and engineered microorganisms — including strains built with the tools described in [[synthetic-genomes]] and biocontained by the synthetic auxotrophy of [[recoded-organisms]] — appear in most forward-looking architectures. None has flown.

## Limitations

Small closed ecosystems are unstable in ways large ones are not. The buffering capacity of the Earth's atmosphere and oceans is enormous relative to any perturbation a population of eight can generate; the buffering capacity of a sealed structure is not. Biosphere 2 illustrated this comprehensively. Atmospheric oxygen fell from the normal 20.9 per cent to about 14.2 per cent over sixteen months and had to be injected. The cause was traced to microbial respiration in organic-rich soils generating carbon dioxide that was then absorbed by uncured concrete, so oxygen disappeared without a corresponding rise in carbon dioxide to signal where it had gone.[^severinghaus1994] Food production fell short, the crew lost weight, most vertebrate and pollinator species died, and ants and cockroaches came to dominate the fauna.

The lesson usually drawn is that Biosphere 2 failed. The more useful lesson is that it failed in a way nobody predicted, through a concrete-chemistry interaction with a soil microbial community, and that closed-system engineering is dominated by such couplings.

Other persistent problems are more mundane. Volatile organic compounds accumulate and must be catalytically oxidised. Biofilms form in water lines and have caused recurring maintenance on the station. Crop pathogens in a monoculture with no natural enemies can propagate through a hydroponic system rapidly. Menu monotony degrades crew intake in ways that show up as weight loss in every long-duration analogue.

> [!caution] Closure percentages are not comparable
> A system described as "98 per cent closed" may be quoting water recovery, total mass balance, or oxygen regeneration, and may or may not count food. Because food dominates the unclosed mass in every flown system, headline closure figures that exclude it can differ by an order of magnitude from figures that include it. Comparisons between facilities generally require reading the definitions.

## Risks

The dominant risk is coupled failure. Physicochemical and biological subsystems share water, air and power, so a fault in one propagates. Redundancy is expensive in mass and does not help against a common-mode failure such as a contaminant that poisons both the crop and the microbial compartments.

Introducing engineered organisms adds a containment question. A bioregenerative loop is a deliberately constructed ecosystem, and organisms designed for it would eventually be released, intentionally or otherwise, into a planetary environment — a scenario that has driven planetary protection policy for decades and that shares structure with the concerns raised in [[mirror-life]]. How much weight the [[precautionary-principle]] should carry when the alternative is a crew running out of food is not a question the literature has resolved.

## Outlook

The near-term trajectory is incremental: better reliability, methane pyrolysis to close the hydrogen loop, higher-yield crop chambers, and improved trace contaminant control. None of that produces a closed system, and no agency's current Mars architecture assumes one; the baseline is high water and oxygen recovery with all food carried.

Full closure only becomes mandatory when resupply is unavailable: a permanent settlement, or the multi-generation voyages considered under [[generation-ship-biology]]. At that point the problem stops being spacecraft engineering and becomes ecology, with all the stability questions that implies and none of the redundancy the Earth provides. It is not obvious that a system small enough to launch can be stable for centuries, and the two-year Biosphere 2 record remains the longest human test of the proposition. Whether the answer is better engineering, larger systems, or adapting the crew instead — the argument made in [[pantropy]] — is unsettled, and bears directly on any serious account of the [[future-of-humanity]] beyond Earth.

## See also

- [[space-medicine]]
- [[generation-ship-biology]]
- [[pantropy]]
- [[microgravity-adaptation]]
- [[human-hibernation]]
- [[recoded-organisms]]
- [[existential-risk]]
- [[technology-readiness-level]]

## References

[^severinghaus1994]: `paper` Severinghaus, J.P., Broecker, W.S., Dempster, W.F., MacCallum, T., Wahlen, M. "Oxygen loss in Biosphere 2." *Eos, Transactions American Geophysical Union*, 1994. {The concrete sink was identified after the fact; oxygen fell without the matching rise in carbon dioxide that would have shown where it went.}
[^salisbury1997]: `paper` Salisbury, F.B., Gitelson, J.I., Lisovsky, G.M. "Bios-3: Siberian Experiments in Bioregenerative Life Support." *BioScience*, 1997.
[^hoffman2022]: `paper` Hoffman, J.A. et al. "Mars Oxygen ISRU Experiment (MOXIE) — Preparing for human Mars exploration." *Science Advances*, 2022.

==============================================================================
ARTICLE: cochlear-implant
TITLE:   Cochlear implant
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/cochlear-implant
SOURCE:  https://futurehumanwiki.com/raw/cochlear-implant
==============================================================================

---
title: "Cochlear implant"
slug: "cochlear-implant"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["cybernetics", "bodies"]
tags: ["hearing", "sensory restoration", "implants", "deafness", "neuroprosthetics", "disability"]
summary: "An implanted device that restores a functional sense of hearing by stimulating the auditory nerve with electrical pulses, bypassing the ear's damaged hair cells."
updated: "2026-07-27"
humanEvidence: "More than 700,000 people have been implanted; most postlingually deafened adults reach open-set sentence recognition in quiet, and a prospective multicentre cohort found earlier-implanted children gained spoken language faster."
access: "Approved and routinely reimbursed in high-income countries; device, surgery and years of audiological programming run to tens of thousands of dollars, and most of the world's profoundly deaf people never receive one."
reversibility: "difficult"
issues: ["The 700,000 recipient count and the near-one-million estimate carry no source.", "The optical cochlear implant work in rodents is described without a citation."]
---

```infobox
{
  "caption": "Auditory neuroprosthesis",
  "rows": [
    { "label": "Type", "value": "Sensory neuroprosthesis" },
    { "label": "Key developers", "value": "House, Clark, Hochmair" },
    { "label": "First multichannel implants", "value": "1977–1978, Vienna and Melbourne" },
    { "label": "First US approval", "value": "1984 (adults)" },
    { "label": "Paediatric approval", "value": "1990 (age 2 and over)" },
    { "label": "Electrodes", "value": "12–24 intracochlear" },
    { "label": "Recipients", "value": "700,000+ worldwide" },
    { "label": "Readiness", "value": "Routine clinical care" }
  ]
}
```

**Cochlear implants** are surgically implanted devices that restore a usable sense of hearing by
delivering electrical pulses directly to the auditory nerve, replacing the function of the inner
ear's hair cells rather than amplifying sound as a hearing aid does. They are the most widely
deployed neuroprosthesis in existence and the only one that routinely restores a lost sense well
enough for most adult recipients to hold a conversation. They are also the subject of the oldest
live ethical dispute in [[neuroprosthetics]], because a substantial part of the Deaf community
regards deafness as a linguistic identity rather than a deficit to be corrected.

```keyfacts
[
  { "value": "~8", "label": "Effective spectral channels", "note": "independent channels saturate well below the 12–24 physical electrodes" },
  { "value": "9 months", "label": "Lowest US-approved implantation age", "note": "lowered from 12 months for some devices around 2020" },
  { "value": "700,000+", "label": "Devices implanted worldwide", "note": "registry estimate as of the late 2010s; likely near one million by 2026" }
]
```

```figure
{"key": "cochlear-implant-generations", "caption": "Four processor generations. The external hardware has changed far more than the implanted electrode array it drives."}
```

## How it works

Sound reaches a microphone on a behind-the-ear or off-the-ear processor. The processor splits the
signal into a bank of frequency channels, extracts the slowly varying amplitude envelope of each,
and converts those envelopes into a schedule of biphasic current pulses. The schedule is
transmitted through the intact skin by radio-frequency coupling across a magnetically aligned coil
pair to a receiver-stimulator seated in a recess milled into the temporal bone. From there a wire
bundle runs to an electrode array threaded through the round window into the scala tympani, roughly
one to one-and-a-half turns along the cochlear spiral.

The array exploits tonotopy. The cochlear base responds to high frequencies and the apex to low
ones, so an electrode's position determines the pitch a stimulated population of spiral ganglion
neurons signals. High-frequency channels drive basal electrodes, low-frequency channels drive
apical ones, and the auditory system interprets the result as sound with approximately correct
spectral structure.

The decisive engineering insight was interleaving. Because current spreads through conductive
perilymph, simultaneous stimulation on neighbouring electrodes sums unpredictably. Continuous
interleaved sampling, introduced in 1991, fires only one electrode at a time in rapid rotation,
which sharply reduced channel interaction and produced a step change in speech scores.[^wilson1991]
Most modern strategies are refinements of that idea, often selecting only the few highest-energy
channels in each cycle.

What the encoding discards is temporal fine structure — the rapid waveform detail within each
frequency band that carries pitch, timbre, and much of the information used to separate one talker
from another. A cochlear implant transmits envelope, not waveform. This single design choice
explains most of the device's characteristic failures.

The device is an encoder, not a decoder: it writes information into the nervous system rather than
reading intent out of it, which puts it on the harder side of the asymmetry described under
[[brain-computer-interface]] and [[neural-decoding]] research. Its success despite that is the
strongest existing evidence that a crude artificial code can be learned by a sensory system given
enough exposure.

## Development history

```timeline
[
  { "year": "1957", "title": "First direct auditory nerve stimulation", "text": "André Djourno and Charles Eyriès stimulate the auditory nerve of a deaf patient in Paris, who perceives crude sound sensations." },
  { "year": "1961–1972", "title": "Single-channel devices", "text": "William House in Los Angeles implants single-electrode devices and, with the 3M company, brings a wearable version to patients; most researchers doubt that useful speech understanding is achievable." },
  { "year": "1977–1978", "title": "Multichannel implants", "text": "Ingeborg and Erwin Hochmair in Vienna and Graeme Clark's group in Melbourne implant multi-electrode arrays; Clark's first recipient, Rod Saunders, demonstrates speech perception." },
  { "year": "1984–1985", "title": "Regulatory approval", "text": "The US Food and Drug Administration approves a single-channel device for adults, followed by the multichannel Nucleus system." },
  { "year": "1990", "title": "Children approved", "text": "US approval extends to children aged two and over, shifting the field toward early implantation and provoking organized objection from Deaf associations." },
  { "year": "1991", "title": "Continuous interleaved sampling", "text": "Blake Wilson and colleagues publish a non-simultaneous stimulation strategy that substantially improves open-set speech recognition." },
  { "year": "2000s", "title": "Bilateral and hybrid devices", "text": "Bilateral implantation becomes common, and electric-acoustic 'hybrid' devices preserve residual low-frequency hearing in the implanted ear." },
  { "year": "2020s", "title": "Age limits fall further", "text": "Approved implantation ages drop below one year for some devices, and gene therapies for specific genetic deafness enter first-in-human trials." }
]
```

## Outcomes

Postlingually deafened adults — people who lost hearing after acquiring spoken language — do best.
Most reach open-set sentence recognition in quiet, meaning they can understand speech without
lipreading or a closed set of choices, and many use the telephone. Group averages conceal enormous
variance: outcomes range from near-normal scores to little more than sound awareness, and the
established predictors (duration of deafness, aetiology, surviving neural population, age,
cognition) together explain only part of the spread. Predicting an individual's result before
surgery remains unreliable.

In children, timing dominates. A prospective multicentre US cohort found that children implanted
earlier gained spoken language faster than those implanted later, with the youngest group closest
to the trajectory of hearing peers.[^niparko2010] Congenital deafness leaves auditory cortex
without patterned input during a developmental window, and implantation after that window produces
poorer speech perception. This is the empirical basis for the push toward infant implantation, and
the reason the ethical dispute cannot be deferred to the child's own later decision.

Bilateral implantation improves sound localization and gives a modest advantage for speech in
noise. It does not restore normal binaural hearing, because the two processors are not
synchronized at the level of temporal fine structure.

## Limitations

The functional ceiling is spectral. Adding physical electrodes does not add independent channels
past a point, because current from one contact excites neurons near its neighbours. Experiments
with acoustic simulations in normal-hearing listeners and with electrode-deactivation in implant
users converge on roughly eight effective channels for most recipients — enough for speech in quiet,
which is highly redundant, and not enough for the tasks that require spectral
detail.[^friesen2001] The same saturation limits the [[retinal-implant]], for the same physical
reason.

The consequences are consistent across recipients:

- **Speech in background noise** degrades much faster than in normal hearing.
- **Music** is generally reported as thin or unpleasant; melody recognition without rhythmic cues
  is poor, though many users retain rhythm perception.
- **Talker identity, prosody, and tonal-language lexical tone** are carried largely by fine
  structure and transmit poorly.

Surgical risk is low but real: facial nerve injury, vestibular disturbance, loss of any residual
acoustic hearing in the implanted ear, and device failure requiring reimplantation. Recipients
carry an elevated risk of bacterial meningitis, identified in the early 2000s and associated
particularly with one discontinued electrode design; pneumococcal vaccination is standard practice.
Devices constrain magnetic resonance imaging, a limitation shared with
[[deep-brain-stimulation]] hardware, though newer receiver magnets are designed for compatibility.

Cost and access are the larger practical limits. Device, surgery, and the years of audiological
programming and rehabilitation that follow run to tens of thousands of dollars, and the great
majority of implants are performed in high-income countries. The distribution problem that
[[access-and-inequality]] describes for expensive medicine generally is visible here in an
unusually stark form: the technology is mature, and most of the world's profoundly deaf people
still do not receive it.

## The Deaf community objection

The National Association of the Deaf issued a position paper in 1991 opposing routine paediatric
implantation, and revised it in 2000 into a more accommodating statement that nonetheless insists
on sign language access and on Deaf people's role in decisions about deaf children.[^nad2000] The
core argument is not that implants fail. It is that deafness is a cultural and linguistic condition
rather than a defect, that more than nine in ten deaf children are born to hearing parents who make
the decision without contact with Deaf adults, and that a device installed in infancy is a choice
made about an identity, not merely about an ear.

Two empirical points sharpen the dispute. A cochlear implant does not make a child hearing; it
produces degraded electrical hearing that is switched off at night and in water, and outcomes vary
widely. And children denied sign language during the years spent waiting for spoken language to
develop risk language deprivation if the implant underperforms. Advocates of bilingual bimodal
upbringing argue that signing does not impede spoken-language development, which weakens the
either-or framing that dominated the 1990s debate.

This is the reference case for the expressivist objection examined in
[[disability-rights-and-enhancement]], and it recurs whenever a technology in
[[human-enhancement]] is offered as a correction to a way of being that some people do not regard
as broken. It also bears on [[morphological-freedom]], since the person modified is not the person
consenting, and it supplies the concrete case that abstract arguments in
[[bioethics-of-enhancement]] tend to lack.

> [!debate] Whose decision
> The medical case for early implantation rests on a developmental window that closes before a child
> can consent. The cultural case rests on the claim that the intervention alters an identity the
> child might have chosen. Both are correct about their own premises, which is why the disagreement
> has not resolved in thirty years.

## Outlook

Three lines of work could change the device's ceiling. Gene therapy for specific genetic deafness
has produced the most striking recent results: delivery of a functional *OTOF* gene by
[[aav-vectors]] restored hearing in children with a rare form of congenital deafness caused by
otoferlin mutations, reported from Chinese and US groups from 2024.[^lv2024] This is
[[somatic-gene-therapy]] restoring biological hearing rather than substituting for it, but it
applies only where hair cells survive and a single gene is at fault. Hair-cell regeneration by small
molecules has so far failed in controlled trials.

Optical stimulation is the most direct attack on the spectral bottleneck. Light does not spread
through perilymph the way current does, so an optical cochlear implant using
[[optogenetics]] to make spiral ganglion neurons light-sensitive could in principle deliver many
more independent channels. Work in rodents by Tobias Moser's group has demonstrated the mechanism;
no human optical cochlear implant exists, and it would require gene delivery to the inner ear as a
prerequisite.

The third line is unglamorous and probably more consequential in the near term: fully implantable
processors with no external component, better automatic scene analysis borrowed from hearing-aid
signal processing, and closing the access gap. Proposals to route non-auditory information through
the same electrode array — infrasound, ultrasound, or arbitrary data streams — belong to
[[sensory-augmentation]] rather than to hearing restoration, and none has been attempted in a
person. Whether the auditory system can be given more than
about eight usable channels through an electrode array at all remains the field's open question,
and no result so far suggests that it can.

## See also

- [[neuroprosthetics]]
- [[retinal-implant]]
- [[sensory-augmentation]]
- [[disability-rights-and-enhancement]]
- [[deep-brain-stimulation]]
- [[speech-neuroprosthesis]]
- [[optogenetics]]
- [[brain-computer-interface]]

## References

[^wilson1991]: `paper` Wilson, B. S., Finley, C. C., Lawson, D. T., Wolford, R. D., Eddington, D. K., Rabinowitz, W. M. "Better speech recognition with cochlear implants." *Nature*, 1991.

[^friesen2001]: `paper` Friesen, L. M., Shannon, R. V., Başkent, D., Wang, X. "Speech recognition in noise as a function of the number of spectral channels: comparison of acoustic hearing and cochlear implants." *Journal of the Acoustical Society of America*, 2001.

[^niparko2010]: `paper` Niparko, J. K. et al. "Spoken language development in children following cochlear implantation." *JAMA*, 2010. {A prospective observational cohort, not a randomized trial; families who choose earlier implantation differ from those who choose later.}

[^nad2000]: `statement` National Association of the Deaf. "NAD Position Statement on Cochlear Implants." 2000. {The association speaking for itself, revising its own more strongly worded 1991 paper; it states a community position, not outcome data.}

[^lv2024]: `paper` Lv, J. et al. "AAV1-hOTOF gene therapy for autosomal recessive deafness 9: a single-arm trial." *The Lancet*, 2024. {A single-arm trial in a handful of children with one rare genetic form of deafness in which hair cells survive.}

==============================================================================
ARTICLE: colossal-biosciences
TITLE:   Colossal Biosciences
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/colossal-biosciences
SOURCE:  https://futurehumanwiki.com/raw/colossal-biosciences
==============================================================================

---
title: "Colossal Biosciences"
slug: "colossal-biosciences"
type: "organization"
status: "contested"
horizon: "late 2020s"
categories: ["organizations", "genetics"]
tags: ["de-extinction", "genome editing", "conservation", "ancient dna", "biotech", "mammoth"]
summary: "A Texas biotechnology company founded in 2021 to recreate extinct species by editing the genomes of living relatives, and to spin out conservation technology from that work."
updated: "2026-07-27"
issues: ["The dire wolf edit counts and the reported valuation are stated without a source."]
---

```infobox
{
  "caption": "Private biotechnology company",
  "rows": [
    { "label": "Founded", "value": "2021" },
    { "label": "Founders", "value": "Ben Lamm, George Church", "link": "/wiki/george-church" },
    { "label": "Headquarters", "value": "Dallas, Texas" },
    { "label": "Chief science officer", "value": "Beth Shapiro" },
    { "label": "Flagship programmes", "value": "Mammoth, thylacine, dodo" },
    { "label": "Reported valuation", "value": "Over $10 billion, 2025" },
    { "label": "Spin-outs", "value": "Form Bio, Breaking" }
  ]
}
```

**Colossal Biosciences** is a privately held biotechnology company founded in 2021 by the entrepreneur Ben Lamm and the geneticist [[george-church]] to pursue [[de-extinction]]: producing animals resembling extinct species by editing the genomes of their closest living relatives. Its stated flagship targets are the woolly mammoth, the thylacine and the dodo. It is the best-funded organization ever to work on the problem and also the most criticized, chiefly over whether an edited proxy animal can properly be described as a de-extincted species.

## Overview

Colossal does not clone extinct animals: no viable cell has ever been recovered from a species that was already extinct, and the only extinct animal ever cloned, the bucardo, came from tissue banked while its last individual was still alive, as [[de-extinction]] describes. Its method is comparative: sequence ancient DNA to identify the genetic differences between the extinct species and a living relative, select a subset of those differences judged to underlie the phenotype of interest, introduce them into the relative's cells using [[crispr-cas9]] and related tools including [[base-editing]], and produce an animal by somatic cell nuclear transfer — the technique behind [[human-cloning|Dolly and her successors]] — or by germline transmission. For the mammoth the living relative is the Asian elephant; for the thylacine, the fat-tailed dunnart; for the dodo, the Nicobar pigeon. Church had set out the general programme years before the company existed, alongside proposals for [[recoded-organisms|recoded genomes]] and other large-scale genome engineering.[^church2012]

The company also frames itself as a conservation technology business, arguing that the tools built for de-extinction — genome assembly from degraded samples, multiplex editing, cloning, artificial gestation — apply directly to endangered species. That claim is the pivot on which most of the debate about Colossal turns.

## History

Lamm, a serial software entrepreneur, and Church, whose Harvard laboratory had been editing elephant cells with mammoth variants since the mid-2010s, incorporated the company in 2021. It raised successive rounds at rapidly rising valuations, reported at over $10 billion in early 2025, funded by venture investors and by individuals with an interest in conservation and in the publicity the projects attract.

Beth Shapiro, an ancient-DNA researcher and author of a well-known skeptical book on de-extinction, joined as chief science officer in 2024 — a hire that gave the company scientific credibility and drew criticism from colleagues who read her earlier work as inconsistent with the company's marketing.[^shapiro2015]

```timeline
[
  { "year": "2021", "title": "Founded", "text": "Ben Lamm and George Church launch the company with the woolly mammoth as its flagship target and a stated goal of calves before the end of the decade." },
  { "year": "2022", "title": "Thylacine programme", "text": "A second programme targets the thylacine, using the fat-tailed dunnart as the host species, with a laboratory at the University of Melbourne." },
  { "year": "2023", "title": "Dodo programme", "text": "A bird programme is announced, requiring primordial germ cell techniques that do not yet work in pigeons." },
  { "year": "2025", "title": "The woolly mouse", "text": "Mice edited with several mammoth-associated coat and metabolism variants are announced as a demonstration of multiplex editing in a live animal." },
  { "year": "2025", "title": "The dire wolf claim", "text": "Three grey wolf pups carrying, on the company's account, about twenty edits at fourteen genes are presented as de-extincted dire wolves, drawing immediate and widespread scientific objection." },
  { "year": "2025", "title": "Further targets", "text": "The company announces work on the giant moa in partnership with New Zealand interests, extending the bird programme." }
]
```

## Programmes

### Mammoth

The mammoth project aims to produce an Asian elephant carrying a set of edits associated with cold adaptation — hair, subcutaneous fat, haemoglobin oxygen affinity at low temperature, and temperature-sensing channels. Church's laboratory reported editing elephant cells with mammoth variants before the company existed. The company has stated a target of first calves in the late 2020s.

Three obstacles are severe. Asian elephants are endangered, and using them as surrogates for a 22-month pregnancy raises welfare and regulatory problems that Colossal proposes to solve with an [[artificial-womb|ex utero]] gestation system that does not exist for any large mammal; the [[ectogenesis]] literature suggests the gap between supporting a late-gestation lamb and gestating an elephant from implantation is not incremental. Elephant cloning has never been demonstrated. And the number of edits required for a genuinely mammoth-like animal is far larger than the number so far attempted, with each additional edit carrying the [[crispr-off-target-effects|off-target and structural-rearrangement risks]] that multiplex editing accumulates.

### Thylacine

The thylacine, extinct since 1936, has a comparatively recent and well-sequenced genome, and its living relatives are small marsupials with short gestation. Marsupial reproduction is in some respects more tractable than placental: young are born at an early developmental stage and complete development in the pouch. The genetic distance between the thylacine and the dunnart is nonetheless very large, considerably larger than between mammoth and elephant.

### Birds

The dodo and moa programmes face a different obstacle. Birds cannot be cloned by nuclear transfer, so edits must be made in primordial germ cells and transmitted through a chimeric host. The technique works in chickens and has not been established in pigeons or ratites.

### The dire wolf announcement

In 2025 Colossal announced three pups, named Romulus, Remus and Khaleesi, produced from grey wolf cells carrying, on the company's account, roughly twenty edits across fourteen genes, and described them as dire wolves. Scientific objection was immediate and near-uniform: dire wolves belong to a lineage that diverged from grey wolves several million years ago and differ at a very large number of loci, so an animal with twenty engineered changes is a grey wolf with a modified phenotype. Critics also noted that some edits were chosen to produce a white coat, a trait for which there is no evidence in the fossil record and which appears to reflect popular culture rather than palaeontology.

The company's defence rests on a functional species concept: if an animal fills the ecological role and displays the relevant traits, the label is defensible. That position has few adherents among systematists, and the episode is now the standard illustration of what critics mean when they say de-extinction is a marketing category.

> [!caution] What "de-extinction" delivers
> No extinct species has been restored. What exists are living animals of extant species carrying a small number of edits derived from an extinct relative's genome. These animals lack the extinct species' full genome, its gut microbiome, its learned behaviour and its ecological context. Whether they are useful is a separate question from whether the name is accurate.

## Funding and spin-outs

Colossal is venture funded, with reported valuations rising sharply across successive rounds. It has spun out two companies: Form Bio, a computational biology platform, and Breaking, which works on enzymatic degradation of plastics. The spin-outs are cited by the company as evidence that the de-extinction work generates transferable technology, and by critics as evidence that the de-extinction work is a fundraising vehicle for other businesses.

The company reports a conservation portfolio that includes cloning of red wolves from cell lines derived from a hybrid coyote population, work on cane-toad-toxin resistance in the northern quoll, and genetic rescue projects in several other species. This work is closer to established conservation genetics than the flagship programmes and has attracted less objection.

## Criticism

Four distinct objections recur.

**Taxonomic.** An edited proxy is not the extinct species. Species are defined by ancestry and by whole genomes, not by a handful of traits, and calling the result a mammoth or a dire wolf misstates what has been achieved. The IUCN's own guidance on the subject uses the term "proxy" throughout and is explicit that the products of such work are not the extinct species restored.[^iucn2016]

**Conservation opportunity cost.** Ecologists have argued that resources spent on de-extinction would save more biodiversity if spent on habitat protection, and that the existence of a claimed technological fix weakens the political case for prevention. It is the same moral-hazard structure that the [[precautionary-principle]] literature identifies in other domains. The counter-argument is that Colossal's funding comes from investors who would not otherwise fund conservation at all.

**Animal welfare.** Elephant surrogacy, repeated cloning attempts with high embryonic loss rates, and the creation of animals with no conspecifics or natural habitat all raise welfare questions that the company has addressed in general terms.

**Communication.** The company announces results through press releases and documentary partnerships ahead of peer review, which makes independent evaluation slow and shapes public understanding in ways the scientific record later has to correct. This is the criticism that has done the most damage to its standing among researchers, including some sympathetic to the underlying science.

## Outlook

Colossal's technical contributions are real and separable from its claims. Multiplex editing at scale in mammalian cells, ancient genome assembly from heavily degraded material, and cloning in difficult species are all genuine capabilities, and the same toolkit underlies [[xenotransplantation]] pig engineering, parts of [[gene-drive]] research, and the [[governance-of-genome-editing|governance debates]] about who may release an engineered organism into a shared environment. Whether an elephant calf carrying mammoth edits is born in the late 2020s is the concrete test the company has set itself, and gestation remains the binding constraint.

The deeper question the company has forced is definitional and will outlast it. If a species is a lineage, no amount of editing restores one, and the entire enterprise is misnamed. If a species is a set of functions in an ecosystem, then editing living relatives is a legitimate route to restoring something valuable, and the objection is semantic. Conservation biology has not settled which it means, and Colossal has made that unsettled question expensive.

## See also

- [[de-extinction]]
- [[george-church]]
- [[crispr-cas9]]
- [[xenotransplantation]]
- [[gene-drive]]
- [[synthetic-genomes]]
- [[artificial-womb]]
- [[human-cloning]]

## References

[^shapiro2015]: `book` Shapiro, B. *How to Clone a Mammoth: The Science of De-Extinction*. Princeton University Press, 2015. {Written nine years before its author became the company's chief science officer, and argues that edited proxies are not restored species.}
[^church2012]: `book` Church, G. and Regis, E. *Regenesis: How Synthetic Biology Will Reinvent Nature and Ourselves*. Basic Books, 2012.
[^iucn2016]: `report` IUCN Species Survival Commission. *IUCN SSC Guiding Principles on Creating Proxy Species for Conservation Benefit*. International Union for Conservation of Nature, 2016. {Guidance from a conservation body with no regulatory authority; it uses "proxy" throughout and sets a conservation-benefit test rather than a taxonomic one.}

==============================================================================
ARTICLE: compression-of-morbidity
TITLE:   Compression of morbidity
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/compression-of-morbidity
SOURCE:  https://futurehumanwiki.com/raw/compression-of-morbidity
==============================================================================

---
title: "Compression of morbidity"
slug: "compression-of-morbidity"
type: "concept"
status: "contested"
horizon: "present"
categories: ["longevity", "society"]
tags: ["aging", "healthspan", "public health", "disability", "demography", "health policy"]
summary: "The hypothesis that chronic illness can be postponed into a shorter interval before death, so that added years of life are healthy ones."
updated: "2026-07-27"
humanEvidence: "Disability rates in the United States and several European countries fell through the 1980s and 1990s and then stalled, while years lived with diagnosed chronic disease rose; no national population has been shown to compress morbidity."
issues: ["The National Long Term Care Survey disability decline is stated without a source."]
---

```infobox
{
  "caption": "Hypothesis in public health and gerontology",
  "rows": [
    { "label": "Proposed by", "value": "James F. Fries" },
    { "label": "First described", "value": "1980" },
    { "label": "Published in", "value": "New England Journal of Medicine" },
    { "label": "Core claim", "value": "Illness onset postponable faster than death" },
    { "label": "Rival scenarios", "value": "Expansion, dynamic equilibrium" },
    { "label": "Status", "value": "Evidence mixed to unfavorable" }
  ]
}
```

**Compression of morbidity** is the hypothesis that the age at which chronic illness and disability begin can be pushed back faster than the age of death, shortening the period of poor health at the end of life. It was set out by the Stanford physician James F. Fries in 1980 and has shaped how health systems and gerontologists think about the value of longer life ever since.[^fries1980] Whether it has actually occurred in any national population remains disputed.

## The 1980 hypothesis

Fries argued from two premises. The first was that human lifespan has a natural ceiling — he took roughly 85 years as the average biological limit — so that the survival curve would become increasingly rectangular as premature death was eliminated, with most people dying in a narrow band near that age. The second was that the onset of chronic disease is more modifiable than the age of death, being driven substantially by behavior: smoking, inactivity, diet, and injury.

If both held, then improving those behaviors would push illness onset rightward while the death age stayed roughly fixed, squeezing morbidity into a shorter terminal interval. The clinical picture Fries described is of a person who remains functional into their eighties and then declines rapidly over months.

The first premise has not held. [[maximum-human-lifespan|Maximum lifespan]] shows no sign of the fixed ceiling Fries assumed, and mean life expectancy in the longest-lived countries has passed 85 for women. This does not by itself refute the hypothesis; it removes the mechanism Fries relied on to keep the death age still while the onset age moved.

## Three scenarios

```compare
{
  "columns": ["Compression", "Expansion", "Dynamic equilibrium"],
  "rows": [
    { "label": "Associated with", "values": ["Fries, 1980", "Gruenberg, 1977", "Manton, 1982"] },
    { "label": "Lifespan assumption", "values": ["Fixed ceiling near 85", "None required", "None required"] },
    { "label": "Years lived with disease", "values": ["Fall", "Rise", "Rise"] },
    { "label": "Severity of that disease", "values": ["Unchanged", "Unchanged or worse", "Falls"] },
    { "label": "Net effect on function", "values": ["Better", "Worse", "Roughly neutral"] },
    { "label": "Care-cost implication", "values": ["Lower lifetime cost", "Higher lifetime cost", "Shifted, not reduced"] }
  ]
}
```

The expansion thesis predates Fries. Ernest Gruenberg's "failures of success" argument held that medicine had become better at preventing death from chronic disease than at preventing the disease itself, so survival gains would accumulate as years lived in illness.[^gruenberg1977] Kenneth Manton's dynamic-equilibrium model split the difference: prevalence of chronic conditions rises because people survive with them, but average severity falls because progression slows, leaving aggregate function roughly stable.

These are not merely rival predictions. They imply different things about what medicine is doing. Compression implies prevention is working; expansion implies rescue is working and prevention is not.

## Evidence

The empirical record is mixed and depends heavily on which outcome is measured.

Measures of *disability* — difficulty with activities of daily living, institutionalization, mobility limitation — improved in the United States and several European countries through the 1980s and 1990s. Analyses of the US National Long Term Care Survey found sustained declines in chronic disability among older adults over that period, consistent with compression.

Measures of *disease* moved the other way. Studies tracking years of life lived with diagnosed chronic conditions have generally found expansion: as mortality from cardiovascular disease fell, the number of years spent living with cardiovascular disease rose. Crimmins and Beltrán-Sánchez examined both types of outcome in US data and concluded that the period had produced expansion of morbidity rather than compression, with gains in life expectancy accompanied by more years with disease and, on some measures, more years with functional loss.[^crimmins2011]

The disability gains also stalled. Rising obesity and diabetes prevalence in cohorts entering old age after 2000 halted and in places reversed the earlier improvement, and self-reported pain and mobility limitation in middle-aged cohorts have worsened in several high-income countries. Whether the widespread treatment of obesity with [[glp-1-receptor-agonists]] changes that trajectory is not yet answerable from population data.

Where compression does appear cleanly is in selected, self-selected groups. A long-running comparison of members of a runners' club with community controls found both lower mortality and a delay in the onset of disability of more than a decade, with disability curves that diverged rather than converging.[^chakravarty2008] It remains the strongest single piece of evidence that [[exercise-and-aging|sustained physical activity]] compresses rather than merely postpones. Cohort studies of religious communities with low rates of smoking and drinking, including the Californian Seventh-day Adventists later folded into the [[blue-zones]] literature, show similar patterns. These groups are not randomly assigned, so the effect is an upper bound on what behavior could achieve, not an estimate of it.

> [!debate] The disagreement in one line
> Compression advocates point to declining disability rates in the late twentieth century; critics point out that those declines coincided with rising disease prevalence and have since stopped, and that the runners and Adventists who show the clearest compression differ from the general population in ways no policy can replicate.

## Why the evidence is hard to read

Four measurement problems recur.

Diagnostic thresholds move. Hypertension, diabetes and chronic kidney disease have all been redefined downward over the period in question, converting healthy people into diseased ones by definition and mechanically producing "expansion".

Self-report drifts. Rising awareness and screening increase reported morbidity independent of any change in underlying health.

Disability depends on environment. Better housing, mobility aids and workplace accommodation reduce measured disability without changing the underlying impairment, so the same body can score as disabled in one decade and not the next.

And period-versus-cohort confusion is endemic. Most published claims compare cross-sections of different birth cohorts at the same age, which conflates aging effects with cohort effects such as childhood nutrition and smoking history.

## Implications

Which scenario obtains determines the fiscal arithmetic of population aging. Under compression, longer lives are cheaper per person because the expensive interval of dependency stays short. Under expansion, each added year of life expectancy carries an added year of care costs, and long-term care rather than acute medicine becomes the binding constraint. Under dynamic equilibrium, costs shift from hospitals to community care without falling. The distinction also reframes the demographic objection discussed under [[overpopulation-and-longevity]]: a population living longer in good function is not the same burden as one living longer in dependency, and the two are routinely conflated.

The scenarios carry ethical weight as well. A long terminal interval of dependency is the situation that assisted-dying law was written for, and expansion of morbidity enlarges the population facing it — a point taken up in [[right-to-die-and-right-to-live]]. Whether the therapies that might compress morbidity reach the people whose morbidity is longest is a separate question, treated under [[access-and-inequality]]; national statistics agencies that publish both figures generally find a wider gap in healthy life expectancy between the most and least deprived areas than in life expectancy itself.

This is the pivot on which the [[longevity-dividend]] argument turns. The economic case for investing in aging biology assumes compression: that delaying the [[hallmarks-of-aging|underlying processes]] of aging delays disease onset across the board rather than extending survival with disease. The [[geroscience-hypothesis]] makes the same assumption explicit, and it is why geroscience trials favor composite endpoints counting multiple age-related conditions rather than mortality alone, the design [[nir-barzilai]] built into the proposed TAME trial. Interventions such as [[senolytics]] and [[rapamycin]] are evaluated in animals on precisely this criterion — whether function is preserved, not merely whether survival is longer.

The counterexample sits in the animal literature: some life-extending manipulations lengthen the frail interval rather than shortening it, an outcome discussed under [[healthspan]]. If human interventions behaved that way, life extension would deliver expansion by design.

## Outlook

Compression remains a policy target more than an observed fact. The strongest version — that morbidity can be squeezed against a fixed wall of natural death — depends on a premise that demographic data have not supported. The weaker and more useful version, that the proportion of life spent in poor health can be reduced, is testable and has not yet been achieved at population scale anywhere.

Whether it becomes achievable depends on something Fries did not consider: whether the onset of multiple age-related diseases shares an upstream driver that can be shifted at once. If it does, compression follows more or less automatically. If chronic diseases are largely independent, postponing them one at a time will keep producing what the record so far shows, which is longer lives containing more illness. Settling that question requires [[aging-biomarkers|measures]] the field does not yet have and trials of a length nobody has yet funded.

## See also

- [[healthspan]]
- [[longevity-dividend]]
- [[geroscience-hypothesis]]
- [[maximum-human-lifespan]]
- [[longevity-escape-velocity]]
- [[exercise-and-aging]]
- [[overpopulation-and-longevity]]
- [[blue-zones]]

## References

[^fries1980]: `paper` Fries, J.F. "Aging, Natural Death, and the Compression of Morbidity." *New England Journal of Medicine*, 1980.
[^gruenberg1977]: `paper` Gruenberg, E.M. "The Failures of Success." *Milbank Memorial Fund Quarterly*, 1977.
[^crimmins2011]: `paper` Crimmins, E.M. and Beltrán-Sánchez, H. "Mortality and Morbidity Trends: Is There Compression of Morbidity?" *The Journals of Gerontology: Series B*, 2011.
[^chakravarty2008]: `paper` Chakravarty, E.F., Hubert, H.B., Lingala, V.B. and Fries, J.F. "Reduced Disability and Mortality Among Aging Runners: A 21-Year Longitudinal Study." *Archives of Internal Medicine*, 2008. {Fries, who proposed the hypothesis, is an author, and the runners were recruited from a running club rather than randomized.}

==============================================================================
ARTICLE: connectomics
TITLE:   Connectomics
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/connectomics
SOURCE:  https://futurehumanwiki.com/raw/connectomics
==============================================================================

---
title: "Connectomics"
slug: "connectomics"
type: "technology"
status: "emerging"
horizon: "2030s"
trl: 5
categories: ["minds"]
tags: ["connectome", "electron microscopy", "neuroscience", "brain mapping", "segmentation", "neural circuits"]
summary: "The mapping of complete wiring diagrams of nervous systems, from macroscale fibre tracts down to every synapse in a volume of tissue."
updated: "2026-07-27"
humanEvidence: "Synapse-resolution mapping in humans has reached one cubic millimetre of fixed cortex from a single person; complete connectomes exist only for a nematode and a fly, and no mammalian brain has been mapped in full."
access: "Nothing to purchase as a service: the published datasets are freely downloadable, and the imaging and segmentation are done by a handful of consortium-scale groups."
reversibility: "irreversible"
issues: ["The sequencing-based barcoding approaches are described without a citation."]
---

```infobox
{
  "caption": "Field of neuroscience and imaging",
  "rows": [
    { "label": "Term coined", "value": "2005" },
    { "label": "Primary method", "value": "Volume electron microscopy" },
    { "label": "Resolution needed", "value": "~4–8 nm in-plane" },
    { "label": "Complete connectomes", "value": "C. elegans, larval and adult fly" },
    { "label": "Largest mammalian volume", "value": "~1 mm³ of cortex" },
    { "label": "Data per mm³", "value": "Roughly a petabyte" }
  ]
}
```

**Connectomics** is the mapping of the complete set of connections in a nervous system. The term
covers two very different enterprises that share a name: macroscale connectomics, which uses
diffusion MRI to trace fibre bundles between brain regions in living people, and synapse-resolution
connectomics, which uses electron microscopy on fixed tissue to identify every neuron and every
synapse in a volume. The second is the one relevant to [[whole-brain-emulation]] and to circuit
neuroscience, and it is destructive, slow, and — for the first time in the mid-2020s — producing
complete brains.

The word was proposed in 2005, by analogy with the genome, for a comprehensive structural
description of a nervous system.[^sporns2005] The practice is much older: the complete wiring
diagram of *Caenorhabditis elegans*, reconstructed by hand from electron micrographs over more than
a decade, was published in 1986 and remains the only nervous system whose full anatomy has been
known for long enough to test what such knowledge is worth.[^white1986]

## Scales and methods

Macroscale connectomes derive from diffusion-weighted MRI, which infers the orientation of white
matter tracts from the directional diffusion of water. Resolution is on the order of a millimetre,
tractography algorithms are prone to false positives at fibre crossings, and the output is a
region-by-region matrix rather than a circuit. The NIH Human Connectome Project, launched in 2010,
produced the reference datasets for this scale and made population-level comparison of connectivity
routine. Macroscale maps are clinically useful — they guide electrode placement in
[[deep-brain-stimulation]] and inform hypotheses about psychiatric disorders as disorders of
connectivity — but they are three orders of magnitude too coarse to identify a synapse.

Synapse-resolution connectomics works from tissue that has been chemically fixed, stained with
osmium and other heavy metals, embedded in resin, and sectioned or milled. Three imaging strategies
dominate: serial-section transmission electron microscopy, in which ultrathin sections are collected
on tape or grids and imaged in parallel; focused ion beam scanning electron microscopy, which mills
away the block face between images and gives isotropic resolution at lower throughput; and
serial block-face scanning electron microscopy using a diamond knife inside the microscope chamber.
All require in-plane resolution of a few nanometres to resolve synaptic vesicles and the ~20 nm gap
between membranes.

Reconstruction, not imaging, was long the bottleneck. Tracing neurites by hand through thousands of
sections is prohibitively slow; the field became tractable when convolutional segmentation models,
notably the flood-filling networks developed at Google, reduced the error rate enough that human
effort shifted from tracing to proofreading.[^januszewski2018] Even so, the adult fly connectome
required a large community of proofreaders working over years.

## Development history

```timeline
[
  { "year": "1986", "title": "C. elegans", "text": "White, Southgate, Thomson and Brenner publish the nematode's complete nervous system: 302 neurons and on the order of 7,000 connections, reconstructed by hand from electron micrographs over more than a decade." },
  { "year": "2005", "title": "The word", "text": "Sporns, Tononi and Kötter propose 'connectome' for a comprehensive structural description of a nervous system, by analogy with the genome." },
  { "year": "2010", "title": "Human Connectome Project", "text": "The NIH funds large-scale diffusion and functional MRI mapping in healthy adults, establishing macroscale connectomics as a standard method." },
  { "year": "2018", "title": "Automated segmentation matures", "text": "Flood-filling networks push automated neurite tracing to error rates low enough that reconstruction of large volumes becomes practical." },
  { "year": "2020", "title": "Fly hemibrain", "text": "A Janelia Research Campus team releases a dense reconstruction of about 25,000 neurons in the central brain of Drosophila, with synaptic connectivity." },
  { "year": "2023", "title": "Larval fly brain", "text": "Winding and colleagues publish the complete synaptic connectome of a Drosophila larva, roughly 3,000 neurons — the first whole brain of an insect." },
  { "year": "2024", "title": "Human cortical millimetre", "text": "Shapson-Coe, Lichtman, Jain and colleagues release H01, a cubic millimetre of human temporal cortex containing tens of thousands of cells and around 150 million synapses in about 1.4 petabytes of imagery." },
  { "year": "2024", "title": "Adult fly connectome", "text": "The FlyWire consortium publishes the wiring diagram of an entire adult fruit fly brain, roughly 140,000 neurons, together with cell-type annotation." },
  { "year": "2025", "title": "MICrONS mouse cortex", "text": "A cubic millimetre of mouse visual cortex is published with structural reconstruction co-registered to prior functional imaging of the same neurons." }
]
```

## What has been mapped

Complete connectomes exist for *C. elegans* — including both sexes and several developmental
stages — for the *Drosophila* larva, whose roughly 3,000 neurons were mapped in 2023,[^winding2023]
and for the adult fruit fly. The fly is the important milestone: an animal with courtship, learning,
navigation, and flight, whose full synaptic wiring is now a downloadable
dataset.[^dorkenwald2024] Partial but dense reconstructions exist for larval zebrafish and for the
mouse and human cortical volumes noted above; the human sample, a cubic millimetre of temporal
cortex, occupies about 1.4 petabytes and contains on the order of 150 million
synapses.[^shapsoncoe2024]

For mammals the unit of achievement is still the cubic millimetre. The MICrONS mouse dataset is the
most informative because the same neurons were imaged functionally, using calcium indicators, before
the tissue was fixed and sectioned; it is therefore possible to ask how a cell's tuning relates to
its inputs. That combination, not the wiring alone, is what circuit neuroscience wanted.

```keyfacts
[
  { "value": "~140,000", "label": "Neurons in the adult fly connectome", "note": "FlyWire, 2024 — the first whole brain of a behaviourally complex animal" },
  { "value": "~1.4 PB", "label": "Data for 1 mm³ of human cortex", "note": "H01 dataset, 2024" },
  { "value": "~10⁶", "label": "Cubic millimetres in a human brain", "note": "Linear extrapolation puts a whole-brain dataset near a zettabyte" }
]
```

## Throughput, cost, and the scaling problem

The human-brain arithmetic is unforgiving. A cubic millimetre at electron-microscope resolution
generates on the order of a petabyte; a human brain is roughly a million cubic millimetres. Linear
extrapolation gives a dataset near a zettabyte, comparable to a substantial fraction of global data
storage, before any of it is segmented. Imaging time scales similarly. Progress on this is real —
multi-beam electron microscopes, faster milling, better compression, and cheaper segmentation have
each moved throughput by large factors — but the gap is six orders of magnitude, and no announced
instrument closes it.

Alternative strategies aim to sidestep the imaging cost entirely. Sequencing-based approaches
developed in Anthony Zador's group tag individual neurons with random RNA barcodes and read out
projections by sequencing rather than imaging, trading synapse-level detail for throughput.
[[george-church]] and collaborators have proposed related molecularly annotated schemes. Synchrotron X-ray
tomography can survey large volumes of stained tissue at coarser resolution without sectioning
them, which is useful for targeting where to slice. None of these yields a synapse-level map of a mammalian
brain.

## Why a wiring diagram is not a mind

The connectome is a static structural description. It does not carry synaptic sign, strength, or
short-term dynamics; it does not record which neuromodulators a cell responds to; it does not
capture extrasynaptic signalling, gap-junction coupling strength, or gene expression. Cornelia
Bargmann and Eve Marder made the argument early and directly: the same anatomy supports many
functional configurations, and knowing the anatomy constrains but does not determine the
computation.[^bargmann2013] This is the same underdetermination that separates a connectome from a
theory of the [[neural-correlates-of-consciousness]], and it is why possession of a wiring diagram
does not by itself decide questions about [[substrate-independence]] or [[machine-consciousness]].

The nematode is the standing test of what a wiring diagram is worth, and the test has not been
passed. Its 302 neurons have been mapped since 1986, and four decades of effort have not produced a
simulation that reproduces the animal's behavioural repertoire — the missing quantities being
exactly synaptic sign and strength and the extrasynaptic neuropeptide signalling that reconfigures
the circuit. No connectome of any animal has yet been used to reproduce that animal's behaviour.

A widely cited demonstration of the limit comes from outside biology. Eric Jonas and Konrad Kording
applied standard neuroscience analyses, including a complete connectome, to a 6502 microprocessor
running video games and found that the methods failed to recover how the chip works.[^jonas2017]
The point is not that connectomes are useless — they are the substrate on which every circuit
hypothesis is built — but that structure underdetermines function.

> [!caution] One brain, one time point
> Every published whole-animal connectome is a snapshot. The fly datasets each come from a single
> individual fixed at a single moment; the nematode reconstruction was assembled from sections of
> more than one animal, so it is a composite rather than any one worm's wiring. Variation between
> individuals of the same species is substantial in mammals and non-trivial even in flies, and
> synapses turn over on timescales of days. A connectome is a snapshot, not a constant.

## Applications

Circuit neuroscience is the immediate beneficiary. Fly connectome data has already been used to
predict responses of unrecorded neurons and to identify circuits for specific behaviours, with
predictions tested by [[optogenetics]] and electrophysiology. In mammals, dense reconstruction has
revealed connectivity rules — which cell types contact which, and how selectively — that sparse
recording could not establish. Structural priors of this kind feed back into applied work:
population models used in [[neural-decoding]] and in [[brain-computer-interface]] research, and
circuit models of the hippocampus underlying attempts at a [[memory-prosthesis]], all depend on
assumptions about connectivity that dense reconstruction can now check. Connectomic methods have
begun to be applied to cortical [[organoids]] as well, where the question is how far self-organized
tissue reproduces the wiring statistics of a real cortex.

Connectomics also underwrites the argument for [[brain-preservation]]: if the connectome plus
molecular annotation is the information that matters, then a preservation method that demonstrably
retains it converts death into a storage problem. Whether that "if" is true is exactly what the
Bargmann–Marder objection challenges, and it bears directly on [[cryonics]], on
[[mind-uploading]], and on what would have to survive for the preserved person to survive with it,
a question treated under [[personal-identity-and-continuity]]. Nothing in current connectomics
settles which level of detail an emulation would need.

## Outlook

The stated near-term targets are a whole mouse brain at synapse resolution — an effort several
groups and funders, including the US BRAIN Initiative, have framed as the next milestone — and
larger human cortical volumes with molecular labelling. A mouse brain is roughly a thousand times
the volume already reconstructed and would be the first mammalian connectome in full. Whether it
arrives in the 2030s depends less on microscopes than on segmentation cost and on whether funders
sustain a project whose output is a dataset rather than a discovery.

## See also

- [[whole-brain-emulation]]
- [[mind-uploading]]
- [[brain-preservation]]
- [[neural-decoding]]
- [[neural-correlates-of-consciousness]]
- [[optogenetics]]
- [[organoids]]
- [[george-church]]

## References

[^white1986]: `paper` White, J. G., Southgate, E., Thomson, J. N. and Brenner, S. "The structure of the nervous system of the nematode *Caenorhabditis elegans*." *Philosophical Transactions of the Royal Society B*, 1986. {The reconstruction was assembled by hand from sections of more than one animal, so it is a composite rather than one individual's wiring.}

[^sporns2005]: `paper` Sporns, O., Tononi, G. and Kötter, R. "The human connectome: a structural description of the human brain." *PLoS Computational Biology*, 2005.

[^januszewski2018]: `paper` Januszewski, M. et al. "High-precision automated reconstruction of neurons with flood-filling networks." *Nature Methods*, 2018.

[^shapsoncoe2024]: `paper` Shapson-Coe, A. et al. "A petavoxel fragment of human cerebral cortex reconstructed at nanoscale resolution." *Science*, 2024. {The tissue was removed during epilepsy surgery in one person, so the sample is a single individual carrying a neurological diagnosis.}

[^dorkenwald2024]: `paper` Dorkenwald, S. et al. "Neuronal wiring diagram of an adult brain." *Nature*, 2024.

[^winding2023]: `paper` Winding, M. et al. "The connectome of an insect brain." *Science*, 2023.

[^bargmann2013]: `paper` Bargmann, C. I. and Marder, E. "From the connectome to brain function." *Nature Methods*, 2013. {A commentary rather than a study, arguing from small invertebrate circuits in which the same anatomy produces different outputs under neuromodulation.}

[^jonas2017]: `paper` Jonas, E. and Kording, K. P. "Could a neuroscientist understand a microprocessor?" *PLoS Computational Biology*, 2017. {A deliberate methodological provocation applied to a 6502 chip; no nervous system was studied and the analogy is the authors' own argument.}

==============================================================================
ARTICLE: consumer-blood-testing
TITLE:   Consumer blood testing
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/consumer-blood-testing
SOURCE:  https://futurehumanwiki.com/raw/consumer-blood-testing
==============================================================================

---
title: "Consumer blood testing"
slug: "consumer-blood-testing"
type: "technology"
status: "contested"
horizon: "present"
trl: 9
categories: ["longevity", "society"]
tags: ["diagnostics", "screening", "overdiagnosis", "preventive medicine", "biomarkers", "direct-to-consumer"]
summary: "The sale of clinical laboratory panels directly to asymptomatic consumers as preventive medicine, where accurate assays meet a screening logic that guidelines reject."
updated: "2026-07-27"
humanEvidence: "The individual assays are standard clinical chemistry validated in people; no randomized trial has shown that broad panels in asymptomatic adults reduce illness or death, and trials of periodic general health checks found no mortality benefit."
access: "Sold online in most US states and in several other countries without a physician referral; single panels cost tens to hundreds of dollars and annual memberships several hundred, almost never reimbursed."
reversibility: "context"
issues: ["Prices are approximate and change frequently.", "Regulation outside the United States is covered only in passing."]
---

```infobox
{
  "caption": "Direct-to-consumer diagnostic service",
  "rows": [
    { "label": "Type", "value": "Clinical laboratory testing" },
    { "label": "Sold as", "value": "Preventive screening" },
    { "label": "Typical panel", "value": "20 to 100+ analytes" },
    { "label": "Ordering", "value": "Consumer-initiated, physician-signed" },
    { "label": "Oversight", "value": "Lab certification, not clinical utility" },
    { "label": "Guideline support", "value": "None for broad panels" },
    { "label": "Readiness", "value": "TRL 9" }
  ]
}
```

**Consumer blood testing** is the sale of clinical laboratory panels directly to people who have no
symptoms and no physician's referral, marketed as preventive or proactive medicine. The assays
themselves are ordinary hospital chemistry, run in certified laboratories, and mostly accurate. What
is contested is the inference drawn from them: whether measuring dozens or hundreds of analytes in a
person who feels well yields information that makes them healthier, or mainly yields
abnormal-looking numbers.

```keyfacts
[
  { "value": "5%", "label": "Healthy results outside a reference interval", "note": "by construction: the interval is set to the central 95% of a reference population" },
  { "value": "~64%", "label": "Chance of at least one flagged result", "note": "on 20 statistically independent tests in a person with nothing wrong" },
  { "value": "0", "label": "Broad panels recommended for asymptomatic adults", "note": "screening bodies grade tests one at a time" }
]
```

## What is sold

The transaction is uniform across sellers. A consumer orders a panel online; a contracted physician
network signs the requisition where state law still requires a clinician's order; blood is drawn at a
commercial service centre, by a mobile phlebotomist, or through a mail-in capillary kit. Results
arrive in an application rather than a consultation, colour-coded against a range and accompanied by
generated advice on diet, [[dietary-supplements]], sleep and training.

Quest and Labcorp, the dominant American reference laboratories, sell consumer-initiated versions of
the tests they run for hospitals. Everlywell built its business on mail-in kits aimed at single
questions; InsideTracker pairs a modest panel with algorithmic nutrition advice; Function Health,
co-founded by the physician and author Mark Hyman, sells an annual membership covering on the order
of a hundred biomarkers. The category has extended into whole-body MRI screening sold on comparable
terms by companies such as Prenuvo.

Panels routinely include analytes with no established screening indication in healthy adults: vitamin D,
extended thyroid batteries, high-sensitivity C-reactive protein sold as a readout of
[[inflammaging]], ferritin, sex-hormone panels, and add-ons for
[[telomeres-and-telomerase|telomere length]] or [[epigenetic-clock|methylation age]]. Many sellers
report results against a proprietary "optimal" band narrower than the laboratory's own reference
interval, which mechanically increases the share of results flagged as actionable.

> [!note] What certification covers
> A CLIA-certified laboratory has demonstrated that it measures what it claims to measure, to a
> stated precision, under audit. Certification says nothing about whether the test should have been
> ordered. Consumer marketing routinely answers the accuracy question while appearing to answer the
> utility one.

## Development history

```timeline
[
  { "year": "2013–2014", "title": "Theranos enters retail", "text": "Theranos opens wellness centres inside Walgreens stores, first in California and then across the Phoenix area, promising a full laboratory panel from a fingerstick sample." },
  { "year": "2015", "title": "Direct-access testing broadens", "text": "Arizona enacts a law, lobbied for by Theranos, letting residents order any laboratory test without a physician's order; a Wall Street Journal investigation later that year reports that the company runs only a small fraction of the tests it offers on its own device." },
  { "year": "2016", "title": "Regulators act", "text": "The US agency that certifies clinical laboratories revokes the certificate of Theranos's California facility and bars Elizabeth Holmes from operating a lab for two years. The company voids or corrects large numbers of previously reported results." },
  { "year": "2019–2021", "title": "Reference labs go direct", "text": "Quest and Labcorp launch consumer-facing platforms selling much of their own catalogue without a physician visit." },
  { "year": "2020", "title": "The pandemic normalizes home collection", "text": "Emergency authorizations for at-home sample collection, including one granted to Everlywell, make mail-in specimen kits a familiar consumer product." },
  { "year": "2022", "title": "Fraud conviction", "text": "Holmes is convicted in January on four counts of defrauding investors, acquitted on the counts relating to patients, and sentenced later that year to more than eleven years." },
  { "year": "2023–2026", "title": "Longevity-branded panels", "text": "Membership services selling a hundred or more analytes a year, framed around biological aging rather than disease, become a prominent part of the market." }
]
```

Theranos is the field's cautionary anchor, and it is routinely misremembered. The fraud lay not in
selling consumers their own laboratory data but in the specific, false claim that a full panel could
be run from a few drops of capillary blood on a proprietary device.[^carreyrou2018] Most samples the
company processed went to conventional commercial analysers. The demand it identified outlived it,
which is why the same retail model returned within a few years under companies making no device
claims at all.

## The arithmetic of reference ranges

The central scientific objection to broad panels is not about assay quality. A reference interval
describes a population; it is not a boundary between health and disease. By convention a laboratory
sets it to cover the central 95 percent of a defined healthy reference group,[^clsi] so roughly one
result in twenty from a person with nothing wrong falls outside it by construction.

Test many analytes at once and the flags accumulate. On twenty statistically independent tests the
probability that a completely healthy person has at least one out-of-range value is about 64
percent; on a hundred, an all-normal report is the unusual outcome. Real analytes are correlated, so
the true figure is lower than the naive calculation, but the direction is not in doubt. A panel wide
enough to be marketed as comprehensive is wide enough that most healthy customers will see something
flagged.

Intervals carry a second problem: each is drawn from whichever population a laboratory sampled on
whichever platform it ran, so the same blood can be flagged by one laboratory and normal in another.

## What screening evidence supports

Bodies that evaluate screening do so one test at a time, against evidence that finding the condition
early changes an outcome. The US Preventive Services Task Force grades each proposed test on that
basis and issues an "insufficient evidence" statement where the evidence does not settle whether
testing helps or harms, as it has for vitamin D screening in asymptomatic adults.[^uspstfvitd] No
screening body recommends broad multi-analyte panels in people without symptoms, and the
professional-society Choosing Wisely lists target several of the tests such panels sell most heavily.

The closest direct test of the strategy is the trial literature on periodic general health checks,
which bundled bloodwork with examination. A Cochrane review of those randomized trials found no
effect on all-cause, cardiovascular or cancer mortality while documenting increases in diagnoses and
in treatment.[^krogsboll] That is a result about a programme rather than about any single analyte,
and it remains the nearest evidence to the claim consumer testing makes.

> [!debate] Population range or personal baseline
> The strongest case for repeat testing is not the cross-sectional flag but the within-person trend:
> an individual varies around their own set point within a band narrower than the population
> interval, so a real drift can be visible while every value stays technically normal. Personal
> reference intervals are a long-standing idea in laboratory medicine, and the same argument underpins
> [[continuous-glucose-monitoring]] and [[wearable-health-sensors]]. Missing is any trial showing
> that acting on such trends improves outcomes in healthy people.

## Multi-cancer early detection

A distinct and more consequential category shares the same retail channel. Multi-cancer early
detection tests, of which Galleri is the most prominent, look for tumour-derived DNA fragments and
methylation signatures in blood, reporting a cancer signal and a predicted tissue of origin. As
laboratory-developed tests they have fallen outside routine premarket review; a 2024 US rule that
would have changed that was vacated by a federal district court the following year.[^fdaldt2024]

The performance data are genuinely interesting and genuinely insufficient. In a published prospective
study of Galleri in adults aged 50 and older, fewer than half of the participants with a positive
signal proved to have cancer, and resolution took months of imaging and procedures for many of
them.[^schrag2023] A large randomized trial in England enrolled on the order of 140,000 participants
and was designed around a reduction in late-stage diagnoses rather than deaths. As of mid-2026 no
multi-cancer test is recommended for routine screening by a national body, and none has been shown
to reduce cancer mortality.

> [!caution] A different risk profile
> A false positive on a cholesterol panel costs a repeat draw. A false positive on a multi-cancer
> test can commit a well person to imaging, endoscopy and biopsy while the signal's source is
> hunted, and a true positive for a cancer that would never have caused symptoms commits them to
> treatment for it. Both the potential benefit and the potential harm are larger than for ordinary
> panels, and the harms are at present the better characterised of the two.

## Harms and the diagnostic cascade

Overdiagnosis is the organising concept. Screening a healthy population finds indolent disease that
would never have surfaced, and because the person is then treated, every such case enters the
statistics as a life saved. Thyroid cancer in South Korea is the clearest documented case: incidence
rose roughly fifteenfold as ultrasound screening spread, with essentially no change in mortality
from the disease.[^ahn2014] That is an argument for grading each test against outcomes rather than
against detection, not against screening as such.

The proximate harm from a flagged panel is the cascade. An out-of-range value invites a confirmatory
test, then imaging, then an incidental finding on that imaging, then a procedure — a sequence in
which each step is locally reasonable and the endpoint was not warranted by the starting point.
Diagnostic labels also carry an effect independent of treatment, documented since a 1978 study found
absenteeism rose among steelworkers after they were told they were hypertensive.[^haynes1978]

The reversibility of the technology is unusual here. The draw leaves nothing behind and testing can
simply be stopped; a result cannot be un-known, and the workup it sets off can end in a biopsy, a
diagnosis, or a permanent notation that no later normal test removes.

Two structural harms round out the list. Companies that do not bill insurers may sit outside the
health-privacy rules governing clinical records, accumulating longitudinal biological data under
consumer-software terms, which connects the practice to [[genetic-discrimination]] and
[[mental-privacy]]. And because the panels are paid out of pocket, they route clinician attention
toward people who can afford them, an inversion of need familiar from
[[access-and-inequality]].

## Outlook

Consumer testing is embedded in longevity practice, where panels serve as the entry point to
[[biohacking]] self-experimentation, to off-label prescribing of drugs such as [[metformin]] and
[[glp-1-receptor-agonists]], and to composite [[biological-age|biological age]] scores whose
underlying [[aging-biomarkers|biomarker validity]] is unresolved.

Three developments would change the assessment. The first is randomized evaluation of the strategy
rather than the assay: whether people given a broad panel end up healthier than people not given one.
The second is a shift from population intervals to individual baselines with published reliability,
which would make repeat testing statistically defensible and is where
[[human-digital-twins|integrated personal modelling]] points. The third is triage that reduces
rather than multiplies downstream procedures, the promise attached to
[[nanoparticle-diagnostics|nanoscale assays]].

Until then the field faces an uncomfortable comparison. The actions most reliably recommended after
a consumer panel — [[exercise-and-aging|structured exercise]], [[sleep-and-longevity|sleep]],
weight, alcohol, smoking —
are the ones recommended without it, and the [[healthspan]] case built on the
[[geroscience-hypothesis]] does not obviously require a hundred analytes to make. A test that
changes what someone does is worth its cost. Most of these have not been shown to.

## See also

- [[aging-biomarkers]]
- [[biological-age]]
- [[epigenetic-clock]]
- [[quantified-self]]
- [[wearable-health-sensors]]
- [[biohacking]]
- [[access-and-inequality]]
- [[compression-of-morbidity]]

## References

[^carreyrou2018]: `book` Carreyrou, J. *Bad Blood: Secrets and Lies in a Silicon Valley Startup*. Knopf, 2018. {Carreyrou's Wall Street Journal reporting from 2015 is the primary source for the gap between the fingerstick claim and the commercial analysers actually used.}
[^clsi]: `report` Clinical and Laboratory Standards Institute. *Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory* (EP28-A3c). {The origin of the convention that a reference interval spans the central 95 percent of a defined healthy reference population.}
[^uspstfvitd]: `report` U.S. Preventive Services Task Force. "Screening for Vitamin D Deficiency in Adults: US Preventive Services Task Force Recommendation Statement." *JAMA*, 2021. {Graded as insufficient evidence to weigh benefits against harms, which is not the same as a finding of no benefit.}
[^krogsboll]: `paper` Krogsbøll, L.T., Jørgensen, K.J. & Gøtzsche, P.C. "General health checks in adults for reducing morbidity and mortality from disease." *Cochrane Database of Systematic Reviews*, 2019. {Covers periodic checks combining bloodwork and examination, not the specific commercial panels sold today.}
[^schrag2023]: `paper` Schrag, D. et al. "Blood-based tests for multicancer early detection (PATHFINDER): a prospective cohort study." *The Lancet*, 2023. {A single-arm study of test performance and diagnostic workup, not a trial of outcomes.}
[^fdaldt2024]: `regulator` U.S. Food and Drug Administration. "Medical Devices; Laboratory Developed Tests." Final rule, 2024. {Vacated by a federal district court in 2025, leaving laboratory-developed tests outside routine premarket review.}
[^ahn2014]: `paper` Ahn, H.S., Kim, H.J. & Welch, H.G. "Korea's Thyroid-Cancer 'Epidemic' — Screening and Overdiagnosis." *New England Journal of Medicine*, 2014.
[^haynes1978]: `paper` Haynes, R.B. et al. "Increased absenteeism from work after detection and labeling of hypertensive patients." *New England Journal of Medicine*, 1978. {The rise followed the label rather than the treatment, which is what makes the study a reference point for diagnostic labelling.}

==============================================================================
ARTICLE: continuous-glucose-monitoring
TITLE:   Continuous glucose monitoring
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/continuous-glucose-monitoring
SOURCE:  https://futurehumanwiki.com/raw/continuous-glucose-monitoring
==============================================================================

---
title: "Continuous glucose monitoring"
slug: "continuous-glucose-monitoring"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["longevity", "cybernetics"]
tags: ["diabetes", "biosensors", "metabolic health", "wearables", "self-tracking", "measurement"]
summary: "A wearable sensor that reports glucose in the fluid between cells every few minutes, well established in diabetes and of unproven value to people without it."
updated: "2026-07-27"
humanEvidence: "Randomized trials in type 1 and insulin-treated type 2 diabetes show lower HbA1c and less hypoglycemia; no trial in people without diabetes has shown a benefit on any health outcome."
access: "Prescribed and widely reimbursed for insulin-treated diabetes in wealthy countries; over-the-counter sensors have been sold in the United States since 2024, paid out of pocket."
reversibility: "reversible"
issues: ["Reimbursement and over-the-counter status outside the US are not covered.", "Sensor-wear durations are given as ranges rather than per device."]
---

```infobox
{
  "caption": "Wearable biosensor",
  "rows": [
    { "label": "Measures", "value": "Interstitial fluid glucose" },
    { "label": "Sampling interval", "value": "1–5 minutes" },
    { "label": "Sensor life", "value": "Roughly 10–15 days" },
    { "label": "First FDA approval", "value": "1999 (retrospective)" },
    { "label": "Real-time systems", "value": "Mid-2000s" },
    { "label": "Over-the-counter clearance", "value": "2024 (US)" },
    { "label": "Main manufacturers", "value": "Dexcom, Abbott, Medtronic" },
    { "label": "Readiness", "value": "TRL 9" }
  ]
}
```

**Continuous glucose monitoring** uses a coin-sized sensor worn on the skin, with a filament sitting
in the tissue beneath it, to report glucose concentration in interstitial fluid every one to five
minutes. In type 1 diabetes and in insulin-treated type 2 diabetes the technology is established,
reimbursed, and supported by randomized trials. Since 2024 the same class of sensor has been sold
over the counter in the United States to adults who do not use insulin, including people without
diabetes, a market in which no trial has shown that wearing one improves health.

```keyfacts
[
  { "value": "5–15 min", "label": "Typical lag behind blood glucose", "note": "diffusion into interstitial fluid plus algorithmic smoothing" },
  { "value": "2024", "label": "First over-the-counter monitors cleared in the US", "note": "Dexcom's Stelo, then Abbott's Lingo" },
  { "value": ">70%", "label": "Consensus time-in-range target", "note": "for most adults with diabetes, 70–180 mg/dL" }
]
```

## How it works

An applicator drives a flexible filament a few millimetres into subcutaneous tissue and leaves an
adhesive patch on the skin. The filament is an enzyme electrode: an immobilized enzyme, usually
glucose oxidase, reacts with glucose diffusing in from the surrounding fluid, and the electronics
read the resulting current, which rises and falls with concentration. A transmitter sends a value to
a phone every one to five minutes. Current consumer sensors are factory-calibrated, need a warm-up
period after insertion before they display anything, and are discarded after roughly ten to fifteen
days.

The crucial detail is what the sensor is standing in. Glucose reaches interstitial fluid by
diffusion out of capillaries, so the interstitial value trails the blood value by something on the
order of five to fifteen minutes, and manufacturers' smoothing algorithms add their own delay. When
glucose is stable the two track closely. When it is moving fast, during exercise, in the half hour
after a meal, or while treating a low, the displayed number describes the recent past. This is why
the trend arrow on the screen carries more decision-relevant information than the digit beside it,
and why insulin users are taught to treat a falling reading differently from a flat one at the same
value.

The clinically useful outputs are aggregate rather than instantaneous: how much of the day is spent
in a target band, how variable the readings are, how much time is spent low overnight. The same
signal also closes a control loop. Paired with an insulin pump and a dosing algorithm, a monitor
becomes the sensor half of an automated insulin delivery system, the sense-and-respond architecture
also used by closed-loop [[deep-brain-stimulation]] and proposed for
[[medical-nanorobots|nanoscale therapeutic devices]].

> [!note] Terminology
> Time in range is the share of the day spent between 70 and 180 mg/dL (3.9–10.0 mmol/L). The
> glucose management indicator converts mean sensor glucose into an HbA1c-like percentage; it is an
> estimate, not a laboratory value, and the two can disagree. The ambulatory glucose profile is the
> standard one-page report that stacks two weeks of traces into a single median-and-percentile day.

## Development history

```timeline
[
  { "year": "1999", "title": "First CGM approved", "text": "The FDA approves MiniMed's system, which records glucose blind and is downloaded at a clinic days later. It is a diagnostic tool for physicians, not a display for patients." },
  { "year": "2001", "title": "GlucoWatch reaches the market", "text": "Cygnus wins approval for a wrist device that draws fluid through the skin by reverse iontophoresis. Skin irritation and unreliable readings sink it, and it is off the market within a few years." },
  { "year": "2005–2006", "title": "Real-time systems arrive", "text": "Medtronic's Guardian RT and DexCom's STS put a live number, a trend arrow, and alarms in the user's hand for the first time." },
  { "year": "2008", "title": "Randomized benefit demonstrated", "text": "A JDRF-funded trial in type 1 diabetes reports improved HbA1c with CGM, with the effect concentrated in adults aged 25 and over." },
  { "year": "2014", "title": "FreeStyle Libre launches in Europe", "text": "Abbott's factory-calibrated, scan-to-read sensor cuts cost and eliminates fingerstick calibration, and drives the first mass adoption outside intensive insulin users." },
  { "year": "2016", "title": "Dosing from the sensor allowed", "text": "The FDA permits the Dexcom G5 to be used for insulin dosing without a confirmatory fingerstick, making the sensor the primary measurement rather than an adjunct. Calibration fingersticks are still required." },
  { "year": "2017", "title": "Trials confirm the effect on injections", "text": "The DIAMOND and GOLD trials show HbA1c improvement in type 1 diabetes managed with multiple daily injections rather than pumps. FreeStyle Libre is approved in the US." },
  { "year": "2018", "title": "Interoperable sensors defined", "text": "The Dexcom G6 is authorized as an integrated CGM, a device category built so that sensors, pumps, and algorithms from different manufacturers can be combined." },
  { "year": "2024", "title": "Over-the-counter sensors", "text": "The FDA clears Dexcom's Stelo and then Abbott's Lingo for sale without a prescription to adults who do not use insulin." }
]
```

Each step turned less on the electrochemistry than on trust in it. The 1999 system was blinded
because nobody was prepared to let a patient act on the number; permitting insulin dosing from a
sensor reading in 2016 was the decision that made the automated pumps, and eventually the retail
shelf, possible.

## Evidence in diabetes

For people who take insulin, a reading changes an action. The 2008 JDRF trial found improved
glycemic control in adults but not in the children and young adults enrolled alongside them, a
split usually attributed to how much the younger groups actually wore the sensor.[^jdrf2008] The
DIAMOND trial extended the result to adults on injections rather than pumps, the larger population,
with a modest but consistent HbA1c reduction.[^diamond2017] In type 2 diabetes treated with basal
insulin in ordinary primary care, the MOBILE trial found a similar direction of
effect.[^mobile2021] Less time spent hypoglycemic is a second and arguably more important benefit,
since severe lows are the acute danger of insulin therapy and often go unnoticed overnight.

An international consensus group in 2019 converted the raw traces into targets that trials and
clinics could share, chiefly the recommendation that most adults with diabetes spend more than 70%
of the day between 70 and 180 mg/dL.[^battelino2019] Time in range is not a validated surrogate for
complications in the way HbA1c is, but it correlates with it and captures variability that a
three-month average hides.

> [!debate] Where the indication ends
> The strongest evidence is in people whose reading changes an insulin dose. In type 2 diabetes
> managed with diet, [[metformin]], or a [[glp-1-receptor-agonists|GLP-1 receptor agonist]], trials
> are fewer, effects on HbA1c smaller, and reviewers disagree about whether continuous wear is worth
> the cost against periodic testing. That boundary, rather than the wellness market, is where most
> of the live clinical argument sits.

## Use without diabetes

In 2024 the FDA cleared the first monitors sold without a prescription, Dexcom's Stelo followed by
Abbott's Lingo, both indicated for adults who do not use insulin.[^fda2024] The consumer proposition
is that watching one's own curve teaches which foods, workouts, and bad nights produce which
responses, and that flattening it is a route to [[healthspan|longer healthy life]]. Companies
including Levels and Signos sell subscriptions built on Dexcom and Abbott hardware, adding food
logging, meal scores, and coaching rather than sensors of their own. The sensor joins the other
[[wearable-health-sensors|consumer physiological sensors]] as a fixture of [[biohacking]] and of the
[[quantified-self|self-tracking]] culture that surrounds it, alongside
[[consumer-blood-testing|direct-to-consumer blood panels]] and
[[dietary-supplements|supplement regimens]].

The scientific basis is thinner than the marketing. The most-cited support is a 2015 Weizmann
Institute study which showed that postprandial glucose responses to identical meals differ markedly
between people, and that an algorithm trained on personal data could design meals that blunted those
responses.[^zeevi2015] The endpoint there was the glucose curve itself. The larger PREDICT study
later reported the same between-person spread and, importantly, substantial variation within the
same person eating the same food on different days.[^berry2020] Reference data from people without
diabetes wearing blinded sensors show that normal glucose tolerance means spending nearly all of the
day in a narrow band, with brief post-meal excursions that a consumer app may flag as
alarming.[^shah2019] A monitor that pathologizes normal physiology has a specificity problem, not a
sensitivity one.

> [!caution] What a spike does not mean
> No threshold for a post-meal glucose excursion has been validated as a risk factor in a person
> with normal glucose tolerance. The same meal, eaten by the same person on two days, can produce
> visibly different curves through sleep, prior activity, stress, sensor site, and simple noise. A
> number that moves is not the same as a number that matters, and the inference from a curve to a
> health outcome runs through evidence that does not exist yet.

## Accuracy and its limits

Sensor accuracy is reported as mean absolute relative difference (MARD): the average percentage gap
between sensor readings and a reference measurement taken at the same time. Manufacturer-reported
MARD for current sensors sits in the high single digits, close to the reproducibility of fingerstick
meters. The figure is easy to over-read: it depends on the reference method, on how many comparisons
fell during rapid change, and on the distribution of glucose values in the study population, so
numbers from different manufacturers' studies are not directly comparable. Accuracy is generally
worse in the low range, which is where errors matter most.

Physical artifacts are common. Lying on a sensor compresses the tissue around it and can produce a
false low overnight, the best-known failure mode among users. Some sensors are affected by common
substances: high-dose vitamin C has been reported to raise readings on certain devices, and
acetaminophen interfered with older Dexcom generations. Adhesive failure, insertion-site
variability, and the first day of a sensor's life all contribute noise that a smooth on-screen line
conceals.

## Risks and second-order effects

The physical risks are minor: adhesive dermatitis, occasional infection, rarely a retained filament
fragment. The interesting risks are behavioural. Clinicians have raised concerns that continuous
feedback encourages food restriction and anxiety in people predisposed to disordered eating, and
alarm fatigue is a documented problem even among those with a clinical reason to wear a sensor.
Attention spent on curves is also attention not spent on [[exercise-and-aging|structured exercise]],
which has randomized evidence on function in older adults, or on [[sleep-and-longevity|sleep]],
whose mortality epidemiology is far larger than anything assembled for glucose curves in people
without diabetes.

A glucose trace also reveals meal timing, sleep, illness, and activity, and it sits with
manufacturers and app companies under commercial privacy policies rather than the rules that govern
clinical records. Cost pushes in the other direction: sensors are a recurring out-of-pocket
expense, so a technology with proven benefit in diabetes is distributed partly by ability to pay, an
instance of the pattern examined under [[access-and-inequality]].

## Outlook

Hardware is moving toward longer wear and more analytes. Senseonics' implanted fluorescence-based
sensor runs for months to a year rather than days, and manufacturers have announced sensors
reporting ketones alongside glucose, which matters for users of automated insulin delivery. Non-invasive optical
measurement remains the field's perennial promise; the FDA warned in 2024 against smartwatches and
rings claiming to measure blood glucose without piercing the skin, and no such device has been
authorized.

The more consequential question is interpretive. Sensors generate dense longitudinal metabolic data
on large numbers of people outside clinical research, the kind of input that
[[human-digital-twins|individualized physiological models]] require. Whether that yields anything
beyond diabetes depends on studies that have not been run: trials in people without diabetes,
powered on health outcomes rather than glucose curves, long enough to show whether early detection
of impaired glucose tolerance changes anything. Until those exist, wellness monitoring occupies
the same position as consumer [[epigenetic-clock|methylation age tests]] and other
[[aging-biomarkers|biomarker panels]] sold on the promise of measuring
[[biological-age|biological aging]]: a real measurement of a real quantity, with an unestablished
link between moving the number and living better. That gap is the standing challenge to the
[[geroscience-hypothesis|geroscience]] programme as a whole, and glucose is simply its most widely
worn instance.

## See also

- [[wearable-health-sensors]]
- [[consumer-blood-testing]]
- [[quantified-self]]
- [[glp-1-receptor-agonists]]
- [[metformin]]
- [[biohacking]]
- [[aging-biomarkers]]
- [[human-digital-twins]]

## References

[^jdrf2008]: `paper` Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group. "Continuous Glucose Monitoring and Intensive Treatment of Type 1 Diabetes." *New England Journal of Medicine*, 2008. {The benefit was significant in adults aged 25 and over; the younger age groups, who wore the sensors less, did not show it.}
[^diamond2017]: `paper` Beck, R.W. et al. "Effect of Continuous Glucose Monitoring on Glycemic Control in Adults With Type 1 Diabetes Using Insulin Injections: The DIAMOND Randomized Clinical Trial." *JAMA*, 2017.
[^mobile2021]: `paper` Martens, T. et al. "Effect of Continuous Glucose Monitoring on Glycemic Control in Patients With Type 2 Diabetes Treated With Basal Insulin: A Randomized Clinical Trial." *JAMA*, 2021.
[^fda2024]: `regulator` U.S. Food and Drug Administration. "FDA Clears First Over-the-Counter Continuous Glucose Monitor." News release, 2024. {The clearance covers the device and its labelled population; it is not a finding that wearing one improves any health outcome.}
[^battelino2019]: `paper` Battelino, T. et al. "Clinical Targets for Continuous Glucose Monitoring Data Interpretation: Recommendations From the International Consensus on Time in Range." *Diabetes Care*, 2019.
[^zeevi2015]: `paper` Zeevi, D. et al. "Personalized Nutrition by Prediction of Glycemic Responses." *Cell*, 2015. {The endpoint was the postprandial glucose curve itself over a short period, not weight, disease incidence, or any other health outcome.}
[^berry2020]: `paper` Berry, S.E. et al. "Human postprandial responses to food and potential for precision nutrition." *Nature Medicine*, 2020.
[^shah2019]: `paper` Shah, V.N. et al. "Continuous Glucose Monitoring Profiles in Healthy Nondiabetic Participants: A Multicenter Prospective Study." *Journal of Clinical Endocrinology and Metabolism*, 2019. {Useful mainly as a reference range: it describes what sensors read in people without diabetes, not what any reading predicts.}

==============================================================================
ARTICLE: crispr-cas9
TITLE:   CRISPR–Cas9
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/crispr-cas9
SOURCE:  https://futurehumanwiki.com/raw/crispr-cas9
==============================================================================

---
title: "CRISPR–Cas9"
slug: "crispr-cas9"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["genetics", "foundations"]
tags: ["crispr", "genome editing", "molecular biology", "gene therapy", "nobel prize", "biotechnology"]
summary: "A bacterial defence system repurposed as a programmable DNA-cutting tool, in which a short guide RNA directs the Cas9 enzyme to a chosen sequence in almost any genome."
updated: "2026-07-27"
humanEvidence: "One approved medicine, Casgevy, edits a patient's own blood stem cells outside the body; in vivo editing has durably lowered the disease-causing protein in transthyretin amyloidosis, and heritable editing is unapproved in every country with a relevant law."
access: "Reagents are ordinary catalogue items sold to any laboratory; the one approved medicine built on the platform treats two inherited blood disorders and is priced in the millions of dollars per patient."
reversibility: "irreversible"
issues: ["The 2025 Federal Circuit ruling in the patent dispute carries no source."]
---

```infobox
{
  "caption": "Genome editing platform",
  "rows": [
    { "label": "Type", "value": "RNA-guided DNA nuclease" },
    { "label": "Source organism", "value": "Streptococcus pyogenes" },
    { "label": "Described", "value": "2012" },
    { "label": "Key figures", "value": "Doudna, Charpentier, Zhang", "link": "/wiki/jennifer-doudna" },
    { "label": "Nobel Prize", "value": "Chemistry, 2020" },
    { "label": "First approved medicine", "value": "Casgevy, 2023", "link": "/wiki/casgevy" },
    { "label": "Readiness", "value": "TRL 9" }
  ]
}
```

**CRISPR–Cas9** is a two-part molecular machine — a short guide RNA and a DNA-cutting protein — that bacteria use to destroy invading viral genomes and that biologists have repurposed to cut a chosen sequence in almost any organism. Its advantage over earlier editing platforms is that the target is specified by an interchangeable twenty-nucleotide RNA rather than by a redesigned protein, so retargeting takes days instead of months. Cutting, however, is only the first step. What an edit becomes depends on how the cell repairs the break, and that repair is the part biologists control least.

```figure
{"key": "crispr-cas9-complex", "caption": "Cas9 holding a guide RNA against the unwound target strand. The enzyme does the cutting; the RNA decides where."}
```

## Origins in bacterial immunity

The sequences now called CRISPR — clustered regularly interspaced short palindromic repeats — were first noticed in 1987 by Yoshizumi Ishino's group at Osaka University, who found an odd series of repeats downstream of a gene they were sequencing for unrelated reasons and could not explain them.[^ishino1987] Through the 1990s and early 2000s the same architecture turned up in many bacteria and archaea. Francisco Mojica at the University of Alicante catalogued the arrays, coined the acronym with Ruud Jansen, and proposed that the sequences between the repeats, called spacers, matched viral and plasmid DNA, implying an adaptive immune function.

```timeline
[
  { "year": "1987", "title": "Repeats observed", "text": "Ishino's group at Osaka University reports unusual repeated sequences in Escherichia coli while sequencing an unrelated gene. Their function is unknown." },
  { "year": "2005", "title": "Spacers matched to viruses", "text": "Mojica and, independently, other groups notice that CRISPR spacers correspond to phage and plasmid sequences, suggesting an immune role." },
  { "year": "2007", "title": "Immunity demonstrated", "text": "Barrangou and colleagues at Danisco show experimentally that Streptococcus thermophilus acquires new spacers after phage challenge and becomes resistant." },
  { "year": "2011", "title": "tracrRNA identified", "text": "Charpentier's group finds a second small RNA required for CRISPR RNA maturation in the type II system, completing the picture of the Cas9 machinery." },
  { "year": "2012", "title": "Programmable cleavage", "text": "Jinek and colleagues show Cas9 can be aimed at a chosen DNA sequence and that its two natural RNAs can be fused into one engineered guide; Gasiunas and colleagues independently report the same cleavage biochemistry." },
  { "year": "2013", "title": "Editing in human cells", "text": "Groups led by Zhang, Church and others adapt the system to mammalian cells within months of the 2012 papers." },
  { "year": "2020", "title": "Nobel Prize", "text": "Charpentier and Doudna share the Nobel Prize in Chemistry for the development of a method for genome editing." }
]
```

The experimental proof came in 2007, when Rodolphe Barrangou, Philippe Horvath and colleagues at the food-ingredients company Danisco challenged *Streptococcus thermophilus* with bacteriophage and showed that surviving bacteria had incorporated fragments of the phage genome into their CRISPR array and become resistant.[^barrangou2007] Subsequent work established that the arrays are transcribed and processed into short CRISPR RNAs, and that these guide Cas proteins to matching DNA.

The step that made the system a tool came in 2012. Martin Jinek, Krzysztof Chylinski, Jennifer Doudna, Emmanuelle Charpentier and co-workers showed that the type II protein Cas9 is a dual-RNA-guided endonuclease, that its two RNAs could be fused into a single chimeric guide, and that changing twenty nucleotides of that guide redirected cleavage to a new sequence.[^jinek2012] Virginijus Šikšnys's group published closely related biochemistry the same year.[^gasiunas2012] Within a year, several laboratories had the system cutting in human and mouse cells.[^cong2013][^mali2013]

## How it works

The system has two jobs that are worth separating, because they fail in different ways. The first is finding one twenty-base address in a genome of three billion. The second is doing something useful once it arrives. Cas9 performs the first job well and the second not at all: it makes a break and leaves the consequences to the cell.

### Targeting

The guide RNA carries a spacer of roughly twenty nucleotides complementary to the intended target. Cas9 does not scan the genome by reading that sequence directly. It first searches for a short motif adjacent to the target called the protospacer adjacent motif, or PAM — for the commonly used *Streptococcus pyogenes* enzyme, the three bases 5'-NGG-3'. Only after binding a PAM does Cas9 locally unwind the double helix and test whether the guide RNA can pair with the adjacent strand. A match holds the complex in place and triggers a conformational change that activates the enzyme; a mismatch lets it release and continue searching.

This two-stage check explains both the specificity and the limits of the system. The PAM requirement means not every site in a genome is addressable, though engineered variants with relaxed PAM preferences have widened the accessible fraction considerably. It also explains why mismatches distant from the PAM are tolerated more often than mismatches close to it, which is the root of most [[crispr-off-target-effects|off-target activity]].

### Cutting and repair

Cas9 has two nuclease domains, HNH and RuvC, which cut the two DNA strands to leave a blunt double-strand break a few base pairs from the PAM. The cell then repairs it, and the repair pathway determines the outcome. Non-homologous end joining and related pathways rejoin the ends directly and frequently insert or delete a few bases, which shifts the reading frame and disables the gene. This is the easy, high-efficiency outcome, and it is why CRISPR is far better at breaking genes than at fixing them.

Precise replacement requires homology-directed repair, which copies a supplied template. HDR operates mainly in dividing cells during S and G2 phase, competes poorly against end joining, and is inefficient in the post-mitotic cells that matter most in the nervous system, heart and muscle. Much of the field's subsequent engineering — [[base-editing]], [[prime-editing]] and [[epigenome-editing]] — exists to get useful changes without depending on a double-strand break at all.

> [!key] Why the break matters
> A double-strand break is a genuine chromosomal injury. The cell's response to it, not the nuclease, produces the edit — which is why the same guide RNA can give clean knockout in one cell type and a mixture of deletions, inversions and unedited alleles in another.

## Adaptation and the wider toolbox

The Cas9 protein from *S. pyogenes* is about 1,360 amino acids, and its coding sequence sits uncomfortably close to the roughly 4.7 kilobase packaging limit of [[aav-vectors|adeno-associated viral vectors]]. Compact orthologues such as the *Staphylococcus aureus* enzyme were adopted partly for that reason. Other CRISPR effectors have distinct properties: Cas12a recognises T-rich PAMs and leaves staggered cuts, while Cas13 targets RNA rather than DNA and underpins several diagnostic assays. Sequence models have begun to generate Cas nuclease variants that were not found in any bacterium, an application of [[ai-protein-design]] to the editing toolkit itself. Catalytically dead Cas9, with both nuclease domains disabled, retains its RNA-programmed DNA binding and became the docking module for the transcriptional and epigenetic tools described elsewhere on this wiki.

Delivery, not cutting, is the practical constraint. Ex vivo work usually electroporates a preassembled Cas9–guide ribonucleoprotein into cells outside the body, where dosing is controlled and the protein degrades within hours. In vivo work depends on [[lipid-nanoparticles]] or viral vectors, and inherits their tropism: nanoparticles reach hepatocytes easily and other tissues with difficulty.

## Current state

As of 2026 the technology is routine in research and has begun to produce approved medicine. [[casgevy]], an ex vivo therapy that disrupts an enhancer of *BCL11A* in a patient's own blood stem cells to raise fetal haemoglobin, was authorised in the United Kingdom and United States across late 2023 and early 2024 for sickle cell disease and transfusion-dependent beta thalassemia; it was the first approved CRISPR medicine anywhere.[^frangoul2021] In vivo editing has also reached patients: a lipid-nanoparticle-delivered Cas9 targeting the *TTR* gene in the liver produced durable reductions in the disease-causing protein in transthyretin amyloidosis.[^gillmore2021]

Beyond human therapy, Cas9 is standard equipment for making animal models, for genome-wide knockout screens, for engineering donor pigs in [[xenotransplantation]], for the edited proxy species pursued under the banner of [[de-extinction]], and for the population-scale constructs described under [[gene-drive|gene drives]]. Crops edited rather than transgenically modified have reached market in several jurisdictions under regulatory regimes that treat small deletions differently from inserted foreign DNA.

## Limitations

Efficiency varies enormously by cell type, target site and delivery route, and the number reported in a paper is usually the best of several guides in a permissive cell line. Precise correction remains hard in non-dividing tissue. Editing is stochastic across a cell population, so a treated tissue is a mosaic of edited and unedited cells, and for many conditions nobody knows what fraction must be corrected to matter clinically.

Measurement is a limitation in its own right. Editing is usually reported as the percentage of sequencing reads carrying an indel at the target, a figure that says nothing about how those edits are distributed. The same aggregate number can describe a population in which most cells carry one disrupted allele and one intact one, or a smaller population of fully edited cells among untouched neighbours, and for a recessive condition the two are clinically very different. Amplicon sequencing also cannot see alleles it fails to amplify, so the rearrangements described below are systematically undercounted by the assay used to declare success.

Immunology imposes another constraint. Because both *S. pyogenes* and *S. aureus* are common human pathogens, a substantial share of people carry pre-existing antibodies and T cells against the corresponding Cas9 proteins, which complicates repeat in vivo dosing.[^charlesworth2019]

## Risks

The best-characterised hazards are genomic. Guides can cut at sequences resembling the target, and the resulting edits are unintended and permanent. On-target damage is arguably the larger problem: repair of a Cas9 break can produce kilobase-scale deletions, inversions and complex rearrangements that short-amplicon sequencing does not detect.[^kosicki2018] Breaks can also trigger a p53-mediated damage response, which selects against successfully edited cells and, in principle, for cells with impaired p53.

The governance questions differ sharply between somatic and heritable use. [[somatic-gene-therapy]] affects only the treated person and is regulated as medicine. [[germline-editing|Heritable editing]] passes changes to descendants, cannot be consented to by the person affected, and after the [[he-jiankui-affair]] of 2018 is prohibited or unapproved in essentially every country with a relevant law — a situation examined in [[governance-of-genome-editing]]. The ease of the technique also makes it a standing example in debates over [[dual-use-research]], where the self-governance model of the [[asilomar-conference]] is repeatedly invoked and repeatedly disputed.

> [!caution] Contested
> Claims that CRISPR is "precise" describe the targeting step, not the outcome. Guide binding is specific; the repair that follows is not, and characterising on-target rearrangements requires long-read or single-cell sequencing that many published studies did not perform.

## The patent dispute

Ownership of the eukaryotic application has been litigated for more than a decade. The University of California, the University of Vienna and Charpentier filed first; the Broad Institute filed later and obtained early grants through accelerated examination. United States interference proceedings turned on who first conceived a working system in eukaryotic cells, and in 2022 the Patent Trial and Appeal Board awarded priority to the Broad. On appeal in 2025 the Federal Circuit vacated that decision and returned the case to the board, so the American position remains unsettled as of 2026. European outcomes have differed, partly on procedural grounds relating to priority claims. The 2020 Nobel Prize in Chemistry to Charpentier and Doudna, which omitted Feng Zhang and [[george-church]] among others, drew comment precisely because it did not track the patent fight.

## Outlook

Cas9 is now the least interesting part of the problem. The active frontier is delivery to tissues other than liver and blood, editors that write rather than break, and the regulatory question of how to evaluate a therapy manufactured for one patient. Efforts to reduce the cost of bespoke editing bear on whether the technology narrows or widens [[access-and-inequality|existing inequities in access]]. Whether cutting-based editing remains the dominant approach at all is genuinely open: for many indications, a tool that changes one base or silences a gene without severing a chromosome may prove both safer and easier to approve.

## See also

- [[base-editing]]
- [[prime-editing]]
- [[crispr-off-target-effects]]
- [[casgevy]]
- [[germline-editing]]
- [[jennifer-doudna]]
- [[gene-drive]]
- [[governance-of-genome-editing]]

## References

[^ishino1987]: `paper` Ishino, Y., Shinagawa, H., Makino, K., Amemura, M., Nakata, A. "Nucleotide sequence of the iap gene, responsible for alkaline phosphatase isozyme conversion in Escherichia coli, and identification of the gene product." *Journal of Bacteriology*, 1987.
[^barrangou2007]: `paper` Barrangou, R. et al. "CRISPR provides acquired resistance against viruses in prokaryotes." *Science*, 2007.
[^jinek2012]: `paper` Jinek, M., Chylinski, K., Fonfara, I., Hauer, M., Doudna, J. A., Charpentier, E. "A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity." *Science*, 2012.
[^gasiunas2012]: `paper` Gasiunas, G., Barrangou, R., Horvath, P., Šikšnys, V. "Cas9–crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria." *Proceedings of the National Academy of Sciences*, 2012.
[^cong2013]: `paper` Cong, L. et al. "Multiplex genome engineering using CRISPR/Cas systems." *Science*, 2013.
[^mali2013]: `paper` Mali, P. et al. "RNA-guided human genome engineering via Cas9." *Science*, 2013.
[^charlesworth2019]: `paper` Charlesworth, C. T. et al. "Identification of preexisting adaptive immunity to Cas9 proteins in humans." *Nature Medicine*, 2019. {Antibodies and T cells were measured in blood from healthy donors; whether they cause harm in a treated patient was not tested.}
[^kosicki2018]: `paper` Kosicki, M., Tomberg, K., Bradley, A. "Repair of double-strand breaks induced by CRISPR–Cas9 leads to large deletions and complex rearrangements." *Nature Biotechnology*, 2018. {The large deletions were visible only because the authors used long-range assays; standard short-amplicon sequencing does not detect them.}
[^frangoul2021]: `paper` Frangoul, H. et al. "CRISPR-Cas9 gene editing for sickle cell disease and β-thalassemia." *New England Journal of Medicine*, 2021. {An early report on the first two patients, one with each disease, rather than the pivotal trial result.}
[^gillmore2021]: `paper` Gillmore, J. D. et al. "CRISPR-Cas9 in vivo gene editing for transthyretin amyloidosis." *New England Journal of Medicine*, 2021. {A small first-in-human dose-escalation report; the endpoint was serum transthyretin concentration, not any clinical outcome.}

==============================================================================
ARTICLE: cryonics
TITLE:   Cryonics
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/cryonics
SOURCE:  https://futurehumanwiki.com/raw/cryonics
==============================================================================

---
title: "Cryonics"
slug: "cryonics"
type: "technology"
status: "contested"
horizon: "indefinite"
trl: 2
categories: ["longevity", "bodies"]
tags: ["cryopreservation", "vitrification", "life extension", "death", "cryobiology", "revival"]
summary: "The practice of preserving legally dead people at cryogenic temperatures in the hope that future medicine can repair and revive them."
updated: "2026-07-27"
humanEvidence: "Fewer than a thousand people have been cryopreserved and none has been revived; no mammal has been recovered from cryogenic temperature, and the strongest whole-organ result is a vitrified rat kidney rewarmed and transplanted."
access: "Sold by a handful of member organizations; Alcor's minimum funding is $200,000 whole-body and $80,000 neuro, the Cryonics Institute charges several times less, and most members pay by life insurance."
reversibility: "irreversible"
issues: ["The membership and patient counts are given without a source.", "The 1990s California pre-mortem case is described without a name or citation."]
---

```infobox
{
  "caption": "Preservation practice and speculative technology",
  "rows": [
    { "label": "First patient", "value": "James Bedford, 1967" },
    { "label": "Named by", "value": "Robert Ettinger, 1960s" },
    { "label": "Method", "value": "Vitrification at −196 °C" },
    { "label": "Legal prerequisite", "value": "Pronouncement of death" },
    { "label": "People preserved", "value": "Fewer than 1,000 worldwide" },
    { "label": "Revival demonstrated", "value": "No" },
    { "label": "Mainstream reception", "value": "Rejected by cryobiology societies" }
  ]
}
```

**Cryonics** is the preservation of a legally dead human body or brain at cryogenic temperatures, on the premise that the damage of dying and of preservation itself may one day be reversible by medicine that does not yet exist. It is not a treatment and makes no claim to work now. It is a wager that the information constituting a person survives the process, and that some future technology will be able to read or restore it. The practice has been closely associated with [[transhumanism]] since both emerged in the 1960s and 1970s, and remains marginal to mainstream medicine.

```keyfacts
[
  { "value": "1967", "label": "First human cryopreservation", "note": "James Bedford, still maintained today" },
  { "value": "−196 °C", "label": "Storage temperature", "note": "liquid nitrogen; no chemical change on this timescale" },
  { "value": "0", "label": "Mammals revived from cryopreservation", "note": "one rat kidney has been recovered and transplanted" }
]
```

## How it works

The procedure begins at legal death, which in most jurisdictions means cardiopulmonary arrest and a physician's pronouncement. Nothing may be done before that point. This constraint shapes everything: the ideal case is a patient who dies expectedly, with a standby team present, and the worst case is an unwitnessed death discovered days later.

### Stabilization

Immediately after pronouncement, the team restores circulation mechanically and begins external cooling with ice, sometimes with drugs intended to limit reperfusion injury and blood clotting. The aim is to reduce the interval of warm ischemia during which cells continue metabolizing without oxygen. Ischemic injury before perfusion is generally regarded, including by cryonics organizations themselves, as the largest source of damage in real cases.

### Cryoprotection

Blood is washed out and replaced by a cryoprotectant solution perfused through the carotid and vertebral arteries, or through the whole vasculature for whole-body cases. Concentrations are ramped upward over hours. The chemistry is the crux of the field: at high enough concentrations, the solution does not freeze as the temperature falls but becomes progressively more viscous until it forms a glass.

### Vitrification and cooldown

Vitrification avoids ice entirely. Ice crystals grow between cells, tear membranes, and concentrate solutes to damaging levels; a vitrified solution has no crystalline structure and no phase change. The cost is toxicity, because the agents that suppress ice — dimethyl sulfoxide, ethylene glycol, formamide and their mixtures — are themselves harmful to cells at the concentrations required. Modern solutions such as those developed at 21st Century Medicine are engineered to minimize toxicity at vitrifying concentrations, and their formulation is the main technical advance of the past twenty-five years.

Below roughly −124 °C the solution passes its glass transition and becomes mechanically brittle. Continued cooling to liquid nitrogen temperature produces thermal stress and, in whole organs and bodies, fracturing. Some organizations have proposed intermediate-temperature storage just below the glass transition to avoid this, at the cost of a more complex and failure-prone storage system.

> [!key] Why temperature is not the hard part
> Storage at −196 °C is essentially free of chemical change: at that temperature, reactions that would degrade tissue take longer than the age of the universe. The unsolved problems are all in the transitions — getting cryoprotectant into a dying brain, and, in any revival scenario, getting it out again without killing what remains.

## Development history

```timeline
[
  { "year": "1962–1964", "title": "The idea in print", "text": "Robert Ettinger's The Prospect of Immortality argues that freezing the dead is rational if future medicine might repair them. Evan Cooper independently publishes a similar argument and founds the first advocacy society." },
  { "year": "1967", "title": "First patient", "text": "James Bedford, a retired psychology professor, is cryopreserved in California. His body was later transferred to Alcor, where it remains." },
  { "year": "1972–1976", "title": "Surviving organizations founded", "text": "Alcor and the Cryonics Institute are established, the latter by Ettinger himself, with trust structures intended to outlast their founders." },
  { "year": "1979", "title": "The Chatsworth failure", "text": "Bodies stored by the Cryonics Society of California are found thawed and decomposed after the organization ran out of money. The scandal nearly ended the field and permanently shaped how later organizations structure funding." },
  { "year": "c. 2001", "title": "Vitrification adopted", "text": "Cryonics organizations switch from freezing to vitrification using low-toxicity cryoprotectant mixtures, displacing ice formation as the dominant damage mechanism in well-perfused tissue." },
  { "year": "2009", "title": "A vitrified organ transplanted", "text": "Greg Fahy's group reports a rabbit kidney vitrified, rewarmed and transplanted that sustained the animal, the first such result for a whole mammalian organ." },
  { "year": "2023", "title": "Nanowarming a rat kidney", "text": "A Minnesota group reports vitrified rat kidneys rewarmed by inductive heating of infused magnetic nanoparticles, then transplanted, restoring life-sustaining function." }
]
```

## The argument

The case for cryonics does not require that current methods preserve a person perfectly. It requires a weaker claim, formalized by Ralph Merkle as *information-theoretic death*:[^merkle1994] a person is dead in the final sense only when the structures encoding their memory and personality have been degraded past any possible inference, not merely past the reach of present medicine. On this view the relevant question about a cryopreserved brain is whether the information is still there in some recoverable arrangement, not whether the tissue is alive.

Two revival routes follow, and they make very different demands. The biological route requires rewarming without ice damage, removing cryoprotectant, and repairing accumulated injury cell by cell — a task usually attributed to hypothetical [[medical-nanorobots]] and thus dependent on the [[molecular-assembler|molecular manufacturing]] programme that [[eric-drexler]] set out and that has not materialized. The informational route abandons the body: the preserved brain is sectioned and scanned, its [[connectomics|connectome]] and molecular state reconstructed, and the person instantiated as a [[whole-brain-emulation|whole brain emulation]]. That route is the explicit goal of [[brain-preservation]] as a separate field, and it inherits every objection to [[mind-uploading]], including the [[substrate-independence]] premise and the question of whether the result would be the same person at all — the subject of [[personal-identity-and-continuity]].

> [!debate] What the disagreement is really about
> Almost no cryobiologist disputes that vitrification suppresses ice, or that −196 °C halts decay. The dispute is about whether the ischemia, toxicity and fracturing that occur in real cases leave enough structure behind to matter, and whether "enough structure" is even a well-defined criterion in the absence of a theory of how memory is physically encoded.

## Evidence

The relevant cryobiology has advanced, though not in the direction of human revival.

Small organisms survive vitrification routinely. Human embryos and oocytes are vitrified and thawed as standard IVF practice with high survival rates, which is the basis of egg banking discussed under [[reproductive-longevity]]; vitrified *C. elegans* have been reported to retain a learned association after revival, which is at least a demonstration that a memory trace can survive the process in a 302-neuron animal.[^vitamore2015]

Whole mammalian organs are the frontier. Fahy's group vitrified rabbit kidneys with the M22 solution, rewarmed them, and transplanted one that supported the animal's life.[^fahy2009] Rewarming was the bottleneck: heating a large mass fast enough and uniformly enough to avoid devitrification and cracking is harder than cooling it. That problem yielded partially in 2023, when a group reported vitrified rat kidneys rewarmed by inductively heating infused magnetic nanoparticles and then transplanted with restored function.[^han2023] These results matter for organ banking and the [[organ-shortage]] problem regardless of what one thinks of cryonics.

No brain has been recovered. No mammal has been revived from cryogenic temperature. The gap between a rat kidney and an adult human brain is several orders of magnitude in mass, and the brain is the one organ for which "restored function" cannot be assessed by perfusion pressure and urine output.

## Current state

Fewer than a thousand people are in cryopreservation worldwide, held by a small number of organizations: Alcor in Arizona, the Cryonics Institute in Michigan, Tomorrow Bio in Europe, and smaller operations elsewhere including Australia's first case in 2024. Several thousand more are signed up as members. Growth has been slow and roughly linear for decades, not exponential.

[[alcor]], which was led from 2011 to 2020 by the philosopher [[max-more]], sets minimum funding at $200,000 for whole-body preservation and $80,000 for neuropreservation — the head alone, on the reasoning that a future capable of reviving anyone can also grow a body, an assumption that leans on [[lab-grown-organs]]; the Cryonics Institute charges several times less than Alcor for whole-body. Most members fund it with a life insurance policy naming the organization as beneficiary, which converts a large capital cost into a modest premium. Whether that makes the practice broadly available or merely cheap for the already insured is one instance of the [[access-and-inequality]] problem that runs through life-extension technology.

The membership has always been small and unrepresentative, skewing technical, male, and American. Among the better-known patients is the futurist [[fm-2030]], preserved at Alcor in 2000.

The structural risk is organizational rather than technical. Patients must be maintained for an unknown period, possibly centuries, by institutions with no revenue from them after the initial payment. Alcor's response is a patient care trust holding invested funds separately from operating accounts, sized so that liquid nitrogen and maintenance can be paid from returns. Whether any private institution can survive on that basis for two hundred years is untested, and the Chatsworth failure is the field's own worked example of what happens when one does not.

## Legal status

Cryonics is legal in most countries but exists in a category the law does not otherwise recognize. Preservation may begin only after death is pronounced, which guarantees a period of ischemia and rules out preserving a patient whose brain is being destroyed by a progressive disease while they are still alive. A challenge to that rule in California in the early 1990s, brought by a man with a brain tumor who sought pre-mortem cryopreservation, was rejected by the courts — a decision that sits awkwardly beside both [[right-to-die-and-right-to-live|assisted-dying law]] and the [[morphological-freedom]] arguments made elsewhere in this field.

Patients are usually handled under anatomical gift or body-donation law, giving them the legal status of remains rather than persons. Estate law offers no mechanism for a preserved person to retain assets, which is why some members establish personal revival trusts of uncertain enforceability. An English High Court ruling in 2016 permitted a terminally ill 14-year-old's wish to be cryopreserved to be carried out, in a judgment that dealt with parental dispute over disposal of a body rather than with cryonics itself.

## Criticism

The Society for Cryobiology has distanced itself from cryonics for decades, at one point barring practitioners from membership, and its position is that the practice is not supported by science and should not be described as medicine. The core objections are specific rather than general.

Cryoprotectant toxicity is not eliminated, only reduced, and the doses reaching brain tissue in a real case are unknown and unmeasurable. Perfusion is uneven; ischemic tissue swells and blocks capillaries, so parts of the brain may receive little cryoprotectant and freeze rather than vitrify. Fracturing at cryogenic temperature is documented and unresolved for whole organs. And the field's central premise, that structure preserves the person, is an assumption about the physical basis of memory that neuroscience has not established — synaptic connectivity is clearly necessary, but whether it is sufficient, or whether molecular states that fixation and cooling do not preserve also matter, is open.

A separate criticism concerns epistemics rather than biology. Cryonics is unfalsifiable in practice: no experiment run today can show that a preserved patient will not be revived in 2300, which makes the claim immune to disconfirmation and therefore, critics argue, more like a belief than a hypothesis. Proponents reply that the relevant decision is a bet under uncertainty with an asymmetric payoff, a framing that critics in turn regard as a rebranding of Pascal's wager.

> [!caution] What is and is not demonstrated
> Vitrification of small tissue volumes is routine laboratory practice. Long-term storage without chemical change is established physics. Revival of a cryopreserved mammal, recovery of a cryopreserved brain, and repair of ischemic and toxic injury at cellular scale are all undemonstrated, and the last requires technology that does not exist.

## Outlook

The field's near-term progress is more likely to come from organ banking than from cryonics itself. If vitrification and nanowarming scale from rat kidneys to human hearts and livers, transplant medicine acquires a supply chain it currently lacks, and cryonics acquires evidence that a large vitrified mammalian organ can be brought back. That would not demonstrate revival of a person, but it would move one of the field's assumptions from argument to data.

The deeper question is not technical. Cryonics asks whether a structure that is not alive, not conscious, and not repairable by any known means should be treated as a patient rather than a corpse. Nothing in cryobiology answers that, and nothing will until either a mammal is recovered or a preserved brain is scanned and shown to contain what it was supposed to contain. Until then the honest statement of the odds is that they are unknown and that the people making the bet are not in a position to collect on it themselves.

## See also

- [[brain-preservation]]
- [[alcor]]
- [[mind-uploading]]
- [[whole-brain-emulation]]
- [[personal-identity-and-continuity]]
- [[medical-nanorobots]]
- [[longevity-escape-velocity]]
- [[max-more]]

## References

[^fahy2009]: `paper` Fahy, G.M., Wowk, B., Pagotan, R. et al. "Physical and biological aspects of renal vitrification." *Organogenesis*, 2009.
[^han2023]: `paper` Han, Z. et al. "Vitrification and nanowarming enable long-term organ cryopreservation and life-sustaining kidney transplantation in a rat model." *Nature Communications*, 2023.
[^vitamore2015]: `paper` Vita-More, N. and Barranco, D. "Persistence of Long-Term Memory in Vitrified and Revived *Caenorhabditis elegans*." *Rejuvenation Research*, 2015. {The animal has 302 neurons and the assay is a single learned odour association, which is a long way from a mammalian memory.}
[^merkle1994]: `paper` Merkle, R.C. "The Molecular Repair of the Brain." *Cryonics*, 1994. {Published in a cryonics movement magazine rather than a peer-reviewed journal; it proposes a criterion for death and reports no experiment.}

==============================================================================
ARTICLE: cynthia-kenyon
TITLE:   Cynthia Kenyon
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/cynthia-kenyon
SOURCE:  https://futurehumanwiki.com/raw/cynthia-kenyon
==============================================================================

---
title: "Cynthia Kenyon"
slug: "cynthia-kenyon"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "longevity"]
tags: ["aging", "genetics", "c. elegans", "insulin signaling", "foxo", "biogerontology"]
summary: "American molecular geneticist whose 1993 finding that a single daf-2 mutation doubles the lifespan of C. elegans made aging a tractable genetic problem."
updated: "2026-07-27"
issues: ["Kenyon's UCSF appointments and the year of her election to the National Academy of Sciences are unsourced.", "Her account of changing her own diet is drawn from interviews rather than a citable document."]
---

```infobox
{
  "caption": "Molecular geneticist",
  "rows": [
    { "label": "Born", "value": "1954" },
    { "label": "Nationality", "value": "American" },
    { "label": "Education", "value": "PhD, Massachusetts Institute of Technology" },
    { "label": "Postdoctoral", "value": "MRC Laboratory of Molecular Biology, Cambridge" },
    { "label": "Known for", "value": "daf-2 lifespan doubling in C. elegans" },
    { "label": "Field", "value": "Molecular genetics; biology of aging" },
    { "label": "Affiliation", "value": "Calico Life Sciences", "link": "/wiki/calico" },
    { "label": "Elected", "value": "National Academy of Sciences" }
  ]
}
```

**Cynthia Kenyon** is an American molecular geneticist whose laboratory showed in 1993 that a
mutation in a single gene, *daf-2*, roughly doubles the lifespan of the nematode *Caenorhabditis
elegans*, and that the extension depends entirely on a second gene, *daf-16*. The long-lived
animals stayed active rather than merely surviving longer, a detail that shaped how the result was
received. Before that experiment, aging was widely treated as dispersed damage with no control
point and no serious genetics; afterwards it was a pathway. Since 2014 she has been vice president
of aging research at [[calico]].

```figure
{"key": "c-elegans-adult", "caption": "About a millimetre long and transparent, with a fixed somatic cell count in the adult hermaphrodite \u2014 which is why a lifespan screen was tractable at all."}
```

## Career

Kenyon studied biochemistry at the University of Georgia and took a doctorate at the Massachusetts
Institute of Technology with Graham Walker, where she showed that DNA-damaging agents switch on a
specific set of genes in *Escherichia coli*, several of them involved in DNA repair — early
evidence that a cell answers damage with a coordinated transcriptional response. She then joined
Sydney Brenner's group at the MRC Laboratory of Molecular Biology in Cambridge, where *C. elegans*
had been built into a model organism and its cell lineage was being traced cell by cell.

She joined the University of California, San Francisco in the mid-1980s and worked first on how Hox
genes pattern the worm's body. The turn to aging was, by her own account, a deliberate move into a
question the field considered unproductive. At UCSF she was a professor in the department of
biochemistry and biophysics and directed the Hillblom Center for the Biology of Aging. She is a
member of the National Academy of Sciences.

## Why the result was unexpected

Evolutionary theory gave no reason to expect genes dedicated to aging: selection weakens once
reproduction ends, so a programme that destroys an old animal has nothing to maintain it. The
working picture was of damage accumulating across many systems at once — oxidation, protein
cross-linking, somatic mutation — with no lever anywhere. [[caloric-restriction]] was the one
intervention reliably known to extend rodent lifespan, and even that had no mechanism attached. A
molecular biologist who announced plans to find aging genes was assumed to be looking for something
that did not exist.

There had been signals. Michael Klass isolated long-lived worm strains in the early 1980s, and
Thomas Johnson and David Friedman later characterized one of them, *age-1*, reporting a
substantially longer mean lifespan together with reduced fertility.[^friedman1988] The work was
known and did not change the field's mind: a single mutant of unknown mechanism is easy to dismiss
as sickly or slow.

## The 1993 experiment

*daf-2* and *daf-16* were already named for their role in dauer formation, an alternative larval
stage that worms enter under crowding or starvation and can persist in for months. Kenyon's group
used temperature-sensitive partial loss-of-function alleles, which let animals develop normally and
then, as adults, live roughly twice as long as wild type.[^kenyon1993] The mutants moved and fed
like younger worms well past the age at which the controls had died. Mutating *daf-16* abolished the
effect completely: without it, *daf-2* mutants lived an ordinary span.

That genetic dependency mattered more than the doubling. It showed the extension was not a
by-product of being ill or torpid but the output of a regulatory relationship — *daf-2* signalling
normally holds *daf-16* off, and lifting that suppression switches on something that keeps an animal
alive longer. Aging, in at least one animal, had a control point.

> [!key] What changed in 1993
> One mutation, in an otherwise normal animal, in a named pathway, with a clean epistatic
> dependency on a second gene. That combination made the result impossible to file under "sick
> worms live longer", and laboratories with no interest in gerontology began running lifespan
> assays.

## Building out the pathway

In 1997 Gary Ruvkun's laboratory cloned *daf-2* and found that it encodes a receptor of the
insulin/IGF-1 family.[^kimura1997] The same year Kenyon's group and Ruvkun's independently
identified *daf-16* as a forkhead transcription factor, the worm counterpart of the mammalian FOXO
proteins.[^lin1997] What the 1993 experiment had caught was a nutrient-sensing system conserved
across animals rather than a worm peculiarity, which is why the finding travelled.

Her laboratory then established that worm lifespan is regulated between tissues rather than cell by
cell. Removing germline precursor cells extends lifespan, and signals from the reproductive system
act on DAF-16 elsewhere in the body.[^hsin1999] Combining germline removal with a *daf-2* mutation
produced animals living roughly six times the normal span and still moving.[^arantes2003] Sensory
neurons contribute too: worms whose sensory perception is impaired outlive worms that sense their
surroundings normally. Downstream, DAF-16 turned out to control hundreds of genes — chaperones and
other [[proteostasis]] machinery, stress-response and antimicrobial genes, metabolic enzymes — so
the longevity output is itself polygenic, and [[autophagy]] genes are required for it.

One experiment matters particularly for translation. Reducing *daf-2* activity only in adulthood,
after development is complete, is enough to extend lifespan.[^dillin2002] Worm aging is not locked
in by developmental history, which is the premise every adult-onset intervention rests on.

```timeline
[
  { "year": "1983–1988", "title": "First long-lived worm mutants", "text": "Michael Klass isolates strains that outlive wild type; Friedman and Johnson later characterize age-1. The field takes little notice." },
  { "year": "1993", "title": "daf-2 doubles lifespan", "text": "Kenyon, Chang, Gensch, Rudner and Tabtiang report that a daf-2 mutation roughly doubles C. elegans lifespan and that the effect requires daf-16." },
  { "year": "1997", "title": "The genes are cloned", "text": "daf-2 proves to encode an insulin/IGF-1-like receptor and daf-16 a forkhead (FOXO) transcription factor, placing worm longevity in a conserved pathway." },
  { "year": "1999", "title": "Lifespan is signalled between tissues", "text": "Removing germline precursor cells extends lifespan, showing that the reproductive system regulates aging in the rest of the animal." },
  { "year": "2002", "title": "Adult-onset reduction suffices", "text": "Lowering daf-2 activity only after development still extends lifespan, separating aging from developmental programming." },
  { "year": "2003", "title": "IGF-1 receptor mice live longer", "text": "Mice heterozygous for the IGF-1 receptor show a modest lifespan increase, concentrated in females — the mammalian echo of the worm result." },
  { "year": "2008", "title": "FOXO3 and human survival", "text": "Variants near FOXO3 are associated with exceptional longevity in a Japanese-American cohort, and the association replicates in other populations." },
  { "year": "2014", "title": "Kenyon joins Calico", "text": "She moves from UCSF to the aging-research company Google founded the year before, as vice president of aging research." }
]
```

## From worms to mammals

Reduced insulin/IGF-1 signalling extends lifespan in fruit flies and, more modestly, in mice: animals
heterozygous for the IGF-1 receptor live longer, with the effect concentrated in
females,[^holzenberger2003] and dwarf mice with defective growth hormone signalling are among the
longest-lived laboratory strains. The direction is consistent across species and the magnitude
shrinks as the animal gets more complicated. A worm doubles. A mouse gains a fraction of that, often
in one sex only, and the long-lived growth hormone mutants carry small body size and altered glucose
handling with them.

In humans the pathway appears in genetics rather than in trials. Variants near FOXO3 are among the
few associations with exceptional survival that have replicated across populations.[^willcox2008]
People with inherited growth hormone receptor deficiency, who have very low circulating IGF-1, have
been reported to be nearly free of diabetes and cancer without living longer than their
relatives.[^guevara2011] Studies of centenarian families, including the cohorts assembled by
[[nir-barzilai]], keep implicating growth and insulin signalling without yielding an intervention.

## Reception and criticism

The immediate objection in 1993 was that the mutants might simply have diverted into a dauer-like
torpor and be living longer by living less. Kenyon's emphasis on their activity, the cloning of the
genes, and the later demonstration that adult-onset reduction works answered most of that. Within a
decade the subject had funding, dedicated journals and dedicated institutes; the organizing claim
that treating aging itself would postpone several diseases at once was later formalized as the
[[geroscience-hypothesis]].

Two criticisms have not gone away. The first is that the worm's adult soma is post-mitotic with a
fixed cell number, so much of what kills mammals has no counterpart in it: no somatic
[[stem-cell-exhaustion]], no dividing cells to accumulate [[telomeres-and-telomerase|telomere
attrition]], no analogue of the mammalian [[cellular-senescence]] that [[senolytics]] are built to
clear. Several of the [[hallmarks-of-aging]] are not testable in a nematode, and a pathway that sets
the rate of decline there need not bound [[maximum-human-lifespan]]. The second is that *daf-2*
mutants pay for their longevity in ways the headline number hides: reduced brood size, and a
physiology partly resembling the stress-resistant dauer state the genes were named for.

> [!debate] A programme for aging, or a repurposed survival programme
> Kenyon has argued that animals carry a regulatory system that sets the rate of aging and that it
> can be turned down. Critics read the same experiments as a nutrient-sensing switch into a
> stress-resistant, low-growth state that happens to postpone death — conserved and real, but not
> the same claim. The distinction decides whether "the aging pathway" is a drug target or a
> metaphor.

## Legacy

The 1993 paper is much of the reason biogerontology can be funded as molecular biology. It made *C. elegans*
the workhorse of the field, supporting compound screens, standardized lifespan assays and
multi-laboratory replication efforts. Arguments that aging is a treatable condition rather than a
fixed fact of biology, pressed by [[aubrey-de-grey]] and later by [[david-sinclair]], rest on an
experimental anchor that did not exist before 1993, and the institutional build-out that followed
runs from the [[buck-institute]] through Calico to [[altos-labs]]. The [[longevity-dividend]] case for treating aging as a public health target rests
on the premise her worms supplied: that the rate of aging is adjustable at all. Kenyon has framed
the goal as [[healthspan]] rather than years alone, on the grounds that her mutants did not merely
persist, they stayed young longer.

What the pathway has not produced is a human intervention. Nothing acting on insulin/IGF-1 signalling
has been shown to slow human aging; [[rapamycin]] and [[metformin]] work on adjacent nutrient-sensing
biology and remain unproven for that purpose in people; no [[aging-biomarkers|surrogate endpoint]]
has been qualified by a regulator, so no trial design would settle the question quickly; and the
genetic route, whether by [[gene-therapy-for-aging]] or otherwise, has no human-validated target.
The worm result changed what questions the field may ask, not what can be prescribed.

> [!caution] What the worm result does not license
> Kenyon has described cutting sugar and starch from her own diet after work in her laboratory
> linked glucose to shortened worm lifespan through DAF-16. That inference has not been tested in
> people. Three decades on, the pathway has yielded human genetics and no human therapy, and the
> best-supported way to compress late-life decline in humans is still
> [[exercise-and-aging|exercise]].

The open question is the one the 1993 result created. Nematode lifespan can be dialled by a single
receptor; mammalian lifespan responds to the same receptor by tens of percent, with costs in growth,
fertility and metabolism attached. Whether that ratio reflects a translation problem that better
molecules will solve, or a real difference between an animal with a fixed soma and one with dividing
tissue and cancer, is unresolved — and it decides whether the pathway is a route to human
[[negligible-senescence|slowed aging]] or a beautiful piece of worm biology.

## See also

- [[calico]]
- [[geroscience-hypothesis]]
- [[hallmarks-of-aging]]
- [[caloric-restriction]]
- [[nir-barzilai]]
- [[david-sinclair]]
- [[maximum-human-lifespan]]
- [[healthspan]]

## References

[^friedman1988]: `paper` Friedman, D. B. and Johnson, T. E. "A mutation in the age-1 gene in Caenorhabditis elegans lengthens life and reduces hermaphrodite fertility." *Genetics*, 1988. {The first long-lived worm mutant to be mapped; the reduced fertility is why it was initially read as a sickness rather than as slowed aging.}
[^kenyon1993]: `paper` Kenyon, C., Chang, J., Gensch, E., Rudner, A. and Tabtiang, R. "A C. elegans mutant that lives twice as long as wild type." *Nature*, 1993.
[^kimura1997]: `paper` Kimura, K. D., Tissenbaum, H. A., Liu, Y. and Ruvkun, G. "daf-2, an insulin receptor-like gene that regulates longevity and diapause in Caenorhabditis elegans." *Science*, 1997.
[^lin1997]: `paper` Lin, K., Dorman, J. B., Rodan, A. and Kenyon, C. "daf-16: An HNF-3/forkhead family member that can function to double the life-span of Caenorhabditis elegans." *Science*, 1997. {Ruvkun's laboratory reported the same identification independently in Nature the same year.}
[^hsin1999]: `paper` Hsin, H. and Kenyon, C. "Signals from the reproductive system regulate the lifespan of C. elegans." *Nature*, 1999.
[^arantes2003]: `paper` Arantes-Oliveira, N., Berman, J. R. and Kenyon, C. "Healthy animals with extreme longevity." *Science*, 2003.
[^dillin2002]: `paper` Dillin, A., Crawford, D. K. and Kenyon, C. "Timing requirements for insulin/IGF-1 signaling in C. elegans." *Science*, 2002.
[^holzenberger2003]: `paper` Holzenberger, M. et al. "IGF-1 receptor regulates lifespan and resistance to oxidative stress in mice." *Nature*, 2003. {Heterozygous knockout mice; the effect was far smaller than in worms and clearest in females.}
[^willcox2008]: `paper` Willcox, B. J. et al. "FOXO3A genotype is strongly associated with human longevity." *PNAS*, 2008. {An association in long-lived Japanese-American men, since replicated elsewhere; it links the pathway to human survival without showing that altering it would help.}
[^guevara2011]: `paper` Guevara-Aguirre, J. et al. "Growth hormone receptor deficiency is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans." *Science Translational Medicine*, 2011. {A small Ecuadorian cohort; disease incidence fell but lifespan did not rise.}

==============================================================================
ARTICLE: david-sinclair
TITLE:   David Sinclair
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/david-sinclair
SOURCE:  https://futurehumanwiki.com/raw/david-sinclair
==============================================================================

---
title: "David Sinclair"
slug: "david-sinclair"
type: "person"
status: "contested"
horizon: "present"
categories: ["people", "longevity"]
tags: ["aging", "sirtuins", "nad", "epigenetics", "reprogramming", "biogerontology"]
summary: "Australian-American biologist at Harvard known for sirtuin and NAD+ research, the information theory of aging, and disputes over the strength of his public claims."
updated: "2026-07-27"
issues: ["The 2024 Academy for Health and Lifespan Research resignations are described without a source.", "The Sirtris acquisition figure is given qualitatively and unsourced."]
---

```infobox
{
  "caption": "Molecular biologist",
  "rows": [
    { "label": "Born", "value": "26 June 1969, Sydney" },
    { "label": "Nationality", "value": "Australian-American" },
    { "label": "Education", "value": "PhD, University of New South Wales, 1995" },
    { "label": "Known for", "value": "Sirtuins; information theory of aging" },
    { "label": "Field", "value": "Genetics; biology of aging" },
    { "label": "Affiliation", "value": "Harvard Medical School" },
    { "label": "Key book", "value": "Lifespan (2019)" }
  ]
}
```

**David Sinclair** is an Australian-American molecular biologist at Harvard Medical School whose
laboratory has worked on yeast aging, sirtuins, NAD+ metabolism and
[[partial-reprogramming|epigenetic rejuvenation]]. He is among the most publicly visible
biogerontologists and the most criticized within the field, with the criticism aimed less at the
underlying experiments than at how far his public statements travel beyond them.

## Career

Sinclair completed a doctorate in molecular genetics at the University of New South Wales and moved
to MIT as a postdoctoral researcher with Leonard Guarente, where he identified extrachromosomal
ribosomal DNA circles as a driver of replicative aging in budding yeast.[^sg1997] He joined Harvard
Medical School in 1999 and co-directs the Paul F. Glenn Center for Biology of Aging Research there.
He has founded or co-founded a series of companies across the aging sector, including Sirtris
Pharmaceuticals, Life Biosciences and the consumer [[biological-age]] testing firm Tally Health, and
he holds advisory positions across many more.

## Sirtuins and resveratrol

The sirtuins are NAD+-dependent deacetylases; Guarente's laboratory had linked the yeast homolog
SIR2 to lifespan and to the response to [[caloric-restriction]], and the question was whether small
molecules could reproduce that response by activating the mammalian equivalents. In 2003 Sinclair's group reported that resveratrol, a polyphenol found in red wine,
activated SIRT1 and extended yeast lifespan.[^howitz2003] A 2006 paper reported that resveratrol
improved survival in mice fed a high-calorie diet.[^baur2006] Sirtris was founded on this work and
acquired by GlaxoSmithKline in 2008 for a sum in the hundreds of millions of dollars.

The programme then unravelled. Pfizer scientists reported that the apparent SIRT1 activation was an
artifact of the fluorophore attached to the assay substrate, and did not occur with native
peptides.[^pacholec2010] GSK halted development of its resveratrol formulation in 2010 and closed
Sirtris in 2013. Sinclair and colleagues have continued to argue that allosteric sirtuin activation
is real under the right substrate conditions, and the biochemistry remains disputed. No sirtuin
activator has demonstrated an effect on human aging.

> [!caution] Contested
> Resveratrol's lifespan effects have not replicated consistently in mammals. The US National
> Institute on Aging's Interventions Testing Program, which tests compounds across three
> genetically heterogeneous mouse cohorts, did not find a lifespan extension for resveratrol —
> in contrast to its result for [[rapamycin]].

## NAD+ and the information theory of aging

Sinclair's laboratory subsequently focused on NAD+ decline with age, reporting that falling nuclear
NAD+ disrupts communication with mitochondria and that raising it with precursors reverses aspects
of that state in mice.[^gomes2013] This work is part of the case for [[nad-precursors]] as
geroprotectors. Human trials of nicotinamide riboside and nicotinamide mononucleotide have reliably
raised blood NAD+ concentrations; they have not demonstrated functional benefit on aging endpoints.

His broader theoretical claim is the information theory of aging: that aging is driven not by the
accumulation of DNA mutations but by the progressive loss of epigenetic information — cells
forgetting which genes to express — and that the information can in principle be restored because
a backup copy persists. The supporting experiment used mice engineered to sustain repeated
non-mutagenic DNA double-strand breaks, which accelerated epigenetic and physiological aging
markers.[^yang2023] Critics have argued that repeated double-strand breaks are a severe and
non-physiological insult, and that the result shows DNA damage accelerates aging phenotypes rather
than that ordinary aging is epigenetic in origin. The distinction matters because the therapeutic
programme of resetting the epigenome follows only from the stronger claim, and because the
readouts used to demonstrate it are [[epigenetic-clock|DNA methylation clocks]] whose causal status
is itself disputed. See [[epigenetic-reprogramming]] and [[hallmarks-of-aging]].

## Reprogramming work

In 2020 his laboratory reported that transient expression of three Yamanaka factors, omitting Myc,
restored vision in mice after optic nerve crush and in an aged glaucoma model, with the effect
depending on DNA demethylases.[^lu2020] The paper is among the most cited demonstrations that
[[yamanaka-factors|OSK]] can improve function in a living mammal without erasing cell identity, and
it is a foundational reference for the reprogramming companies, including [[altos-labs]] and
[[newlimit]]. It is also a single tissue in a mouse, using an inducible transgenic system rather
than a deliverable therapy, and the extrapolation to systemic human rejuvenation is not supported
by it.

## Reception

Sinclair's publication record is substantial and his laboratory's core findings on NAD+ biology and
optic nerve reprogramming are taken seriously. The criticism concerns claims made outside the
papers. *Lifespan* (2019) argues that aging is a treatable disease and that readers alive today may
benefit; it also describes his personal [[dietary-supplements|supplement regimen]], which the book's
own evidence does not support for humans. Colleagues including the biogerontologist Matt Kaeberlein
and the NAD biochemist Charles Brenner have publicly disputed both specific mechanistic claims and
the general practice of presenting mouse data in language that implies human results.

The sharpest institutional rupture came in 2024. As president of the Academy for Health and
Lifespan Research, Sinclair promoted a canine supplement in terms that other members read as
claiming demonstrated age reversal in dogs. A supplement carries no regulatory review of such a
claim, unlike the canine lifespan candidates [[loyal]] is taking through the veterinary regulator.
Much of the academy's board resigned in protest and Sinclair stepped down as president. The episode
is now the standard citation in arguments that the longevity field's credibility problem is
self-inflicted.

> [!debate] Where the disagreement actually sits
> Almost nobody disputes that NAD+ falls with age or that OSK expression alters epigenetic marks.
> The dispute is whether these are causes or correlates, and whether a mouse eye result licenses
> statements about human aging. Sinclair's answer has generally been more confident than his
> colleagues'.

## Legacy

Sinclair did more than any other academic to make aging biology fundable and discussable outside
the field, and the reprogramming programme his laboratory helped seed is now among the most heavily
capitalized directions in biotechnology. He has also been an effective advocate for the
[[geroscience-hypothesis]] in policy settings, arguing that aging should be a regulatory indication
in its own right. The same visibility supplies the field's critics with their best material, and it
has complicated the parallel effort to get [[aging-biomarkers|surrogate endpoints]] taken seriously
by regulators who now associate the field with overstatement.

Whether the information theory of aging survives as a causal account or is absorbed as one more
entry in the hallmarks list depends on experiments — reprogramming delivered systemically to
normally aged animals, with lifespan rather than clock readings as the endpoint — that have not yet
been reported.

## See also

- [[epigenetic-reprogramming]]
- [[nad-precursors]]
- [[partial-reprogramming]]
- [[epigenetic-clock]]
- [[aubrey-de-grey]]
- [[shinya-yamanaka]]
- [[geroscience-hypothesis]]
- [[altos-labs]]

## References

[^sg1997]: `paper` Sinclair, D. A. and Guarente, L. "Extrachromosomal rDNA circles — a cause of aging in yeast." *Cell*, 1997.
[^howitz2003]: `paper` Howitz, K. T. et al. "Small molecule activators of sirtuins extend Saccharomyces cerevisiae lifespan." *Nature*, 2003. {The lifespan measurements are in yeast, and the SIRT1 activation was read out with a fluorophore-tagged substrate later argued to produce the effect.}
[^baur2006]: `paper` Baur, J. A. et al. "Resveratrol improves health and survival of mice on a high-calorie diet." *Nature*, 2006. {The survival benefit was in mice fed a high-calorie diet; later work found no lifespan extension in mice on a standard diet.}
[^pacholec2010]: `paper` Pacholec, M. et al. "SRT1720, SRT2183, SRT1460, and resveratrol are not direct activators of SIRT1." *Journal of Biological Chemistry*, 2010.
[^gomes2013]: `paper` Gomes, A. P. et al. "Declining NAD+ induces a pseudohypoxic state disrupting nuclear-mitochondrial communication during aging." *Cell*, 2013.
[^lu2020]: `paper` Lu, Y. et al. "Reprogramming to recover youthful epigenetic information and restore vision." *Nature*, 2020.
[^yang2023]: `paper` Yang, J. H. et al. "Loss of epigenetic information as a cause of mammalian aging." *Cell*, 2023. {The mice were engineered to sustain repeated induced double-strand breaks, a severe insult that ordinary aging does not deliver.}

==============================================================================
ARTICLE: de-extinction
TITLE:   De-extinction
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/de-extinction
SOURCE:  https://futurehumanwiki.com/raw/de-extinction
==============================================================================

---
title: "De-extinction"
slug: "de-extinction"
type: "technology"
status: "contested"
horizon: "2030s"
trl: 4
categories: ["genetics"]
tags: ["conservation", "genome editing", "cloning", "ancient dna", "biodiversity", "proxy species"]
summary: "The attempt to recreate extinct organisms, in practice by editing a close living relative to carry some of the extinct animal's traits rather than restoring the species itself."
updated: "2026-07-27"
humanEvidence: "Nothing here has been applied to people; the only extinct animal ever brought to term was a cloned Pyrenean ibex that died within minutes of birth in 2003."
access: "Nothing to obtain: no restored species exists, and the work runs inside a small number of venture-funded and academic programmes."
reversibility: "irreversible"
issues: ["The 2025 dire wolf and edited-mice announcements are described without a citation.", "The Arctic albedo and permafrost hypothesis is stated without a source."]
---

```infobox
{
  "caption": "Applied genomics programme",
  "rows": [
    { "label": "Aim", "value": "Recreating extinct phenotypes" },
    { "label": "Dominant method", "value": "Editing a living relative" },
    { "label": "Alternative methods", "value": "Cloning, back-breeding" },
    { "label": "Best-known developer", "value": "Colossal Biosciences", "link": "/wiki/colossal-biosciences" },
    { "label": "Only extinct animal cloned", "value": "Bucardo, 2003" },
    { "label": "Species restored to date", "value": "None" }
  ]
}
```

**De-extinction** is the attempt to bring back an extinct organism. In practice, no project restores an extinct species. What the working programmes produce is a proxy: a living species edited to carry some of the extinct animal's genetic variants and, if the edits work, some of its traits. Whether that constitutes de-extinction is a question of definition, and the definition has become the main point of public dispute.

## Three routes, only one of which is genomic

**Back-breeding** selectively breeds living descendants for traits resembling the ancestral form. The Heck cattle of the 1920s and the later Tauros programme aimed at something aurochs-like; the Quagga Project in South Africa breeds plains zebras for reduced striping. These produce animals that look the part and are genomically ordinary.

**Cloning by somatic cell nuclear transfer** requires intact cells, which extinction usually precludes. The one success remains the bucardo, a Pyrenean ibex whose last individual was tissue-sampled before her death in 2000; a cloned kid was born in 2003 and died within minutes from a lung malformation.[^folch2009] The same technique is used productively for species that are endangered rather than extinct, as in the cloning of a black-footed ferret and a Przewalski's horse from cell lines cryopreserved decades earlier. That is genetic rescue, and it works. The mechanics are covered in [[human-cloning]].

**Genome editing of a close relative** is the route every current headline project takes. It exists because ancient DNA is fragmentary. Even exceptionally preserved permafrost specimens yield short, chemically damaged fragments rather than chromosomes, and there are no living cells to clone from.[^vandervalk2021] Sequencing a mammoth is therefore possible; growing one from mammoth cells is not. The workaround is to sequence the extinct genome, compare it with a living relative, identify variants underlying traits of interest, and install those variants in the relative's cells with [[crispr-cas9]] or its derivatives.

## What has actually been done

```timeline
[
  { "year": "2003", "title": "Bucardo cloned", "text": "A cloned Pyrenean ibex is born and dies within minutes, the only extinct animal ever brought to term and still the only one." },
  { "year": "2013", "title": "The idea goes public", "text": "A TEDx conference organised by Revive & Restore and National Geographic gives de-extinction its name and its research agenda, including passenger pigeon and heath hen projects." },
  { "year": "2016", "title": "Conservation guidance", "text": "The IUCN Species Survival Commission issues guiding principles treating de-extinction products as proxy species, to be judged by conservation benefit rather than by resemblance." },
  { "year": "2021", "title": "Colossal founded", "text": "Ben Lamm and George Church start a company to pursue mammoth, and later thylacine and dodo, with venture funding on a scale the field had not previously seen." },
  { "year": "2025", "title": "Edited proxies announced", "text": "Colossal publicises mice edited for cold-adapted coat and metabolic traits and three grey wolves carrying, on the company's account, about twenty edits at fourteen genes and described as dire wolves; taxonomists and conservation biologists reject the second characterisation." }
]
```

The 2025 canid announcement is the clearest case study. Three grey wolf pups were produced carrying, on the company's account, about twenty edits across fourteen genes, selected from a dire wolf genome assembled from ancient specimens, and presented as the return of a species that has been extinct for roughly ten thousand years. Specialists pointed out that dire wolves are not closely related enough to grey wolves for a small edit set to reconstitute them, that the edited traits were largely morphological, and that the resulting animals are grey wolves with alterations. The company's own scientific framing has been more careful than its announcements, using a functional definition under which an animal counts if it performs the extinct species' ecological role; the exchange is set out under [[colossal-biosciences]].

Mammoth work is further from an animal and technically more interesting. It requires elephant pluripotent stem cells, which were reported only recently and remain hard to work with; multiplex editing of many loci; and gestation. The last is the binding problem. An elephant pregnancy runs close to two years, the only available surrogate species is itself endangered, and the alternative is an [[artificial-womb]] for a several-hundred-kilogram fetus, which is a much harder problem than the neonatal work described in [[ectogenesis]].

Marsupials invert the difficulty. The thylacine programme uses the fat-tailed dunnart, a mouse-sized relative, because marsupial gestation is very short and development continues in the pouch. The size difference between host and target is extreme, and the germline route requires deriving and editing marsupial stem cells with tools that were developed for mice. Birds are harder still: avian de-extinction depends on culturing and editing primordial germ cells, which works reliably in chickens and poorly in pigeons, the group containing the dodo's relatives. All three programmes are gated on reproductive biology, not on editing, and the field's dependence on [[induced-pluripotent-stem-cells]] and on [[in-vitro-gametogenesis]]-adjacent techniques is more decisive than its dependence on nucleases.

## What "de-extinct" would have to mean

A species is not a list of variants. It is a genome, a developmental environment, a gut microbiome, a set of learned behaviours transmitted between generations, and an ecological context. An edited elephant would have elephant mitochondria unless those were replaced, elephant epigenetic inheritance, an elephant microbiome and no mammoths to learn from. Beth Shapiro, an ancient-DNA specialist who later became chief science officer at [[colossal-biosciences]], argued before joining the company that the honest target is a functional proxy rather than a resurrection.[^shapiro2015]

The IUCN's guidance takes the same line, treating any such organism as a proxy to be evaluated on whether it delivers conservation benefit and on what risks its release carries.[^iucn2016] Under that standard the interesting question about a cold-adapted elephant is not whether it is a mammoth but whether releasing it into Arctic tundra does what proponents claim, which is to convert moss and shrub tundra to grassland, raise surface albedo and slow permafrost thaw. That hypothesis is contested among ecologists and has never been tested at scale.

> [!debate] Restoration or distraction
> Proponents argue that de-extinction develops genetic-rescue tools with immediate use for endangered species, and that the publicity funds conservation science that would otherwise go unfunded. Critics reply that resources spent on proxies could conserve more biodiversity if spent on habitat, and that the promise of reversal weakens the case for prevention.[^bennett2017]

## Risks and governance

Releasing an edited organism into the wild is a deliberate ecological intervention, and it inherits the governance problems of [[gene-drive]] without the self-propagating mechanism. Existing frameworks fit awkwardly. A proxy is a genetically modified organism under most national biosafety law, a candidate for reintroduction under conservation law, and in some jurisdictions not obviously a protected species at all, since the extinct taxon it names is not on any list. Whether a proxy could be granted the legal protections of the species it imitates is unsettled and matters commercially.

Animal welfare is the more immediate concern. Cloning and interspecies gestation have high failure rates and produce sick neonates, and the surrogate species used are often themselves threatened. The bucardo died of a developmental defect within minutes of birth, which is the modal outcome of nuclear transfer in a novel species rather than an unlucky one.

The strongest technical caution is unrelated to ecology. Multiplex editing of many loci compounds the chance of the chromosomal losses and rearrangements described in [[crispr-off-target-effects]], and a programme that plans to make dozens or hundreds of changes in one genome is operating well beyond the edit counts validated in therapeutic work such as [[germline-editing]] or in the engineered pigs of [[xenotransplantation]].

## Outlook

The near-term output of these programmes is likely to be tools rather than animals: better stem-cell derivation in non-model species, higher-multiplex editing, cryobanking of cell lines from living endangered populations, and improved ancient-genome assembly. Those have conservation value independent of whether a mammoth is ever born, and the same infrastructure supports work in [[synthetic-genomes]] and [[recoded-organisms]].

The unresolved question is what standard should govern the claim. If a proxy counts as de-extinction when it fills an ecological role, the burden is ecological evidence that it does, which no project has yet supplied. If it counts when it resembles the extinct animal, the field is doing selective breeding with better instruments, and the [[precautionary-principle]] arguments raised against it have little to attach to. The companies have so far been allowed to choose the standard and announce the result against it.

## See also

- [[colossal-biosciences]]
- [[human-cloning]]
- [[crispr-cas9]]
- [[gene-drive]]
- [[artificial-womb]]
- [[xenotransplantation]]
- [[induced-pluripotent-stem-cells]]
- [[precautionary-principle]]

## References

[^folch2009]: `paper` Folch, J. et al. "First birth of an animal from an extinct subspecies (Capra pyrenaica pyrenaica) by cloning." *Theriogenology*, 2009. {The donor cells were banked from the last living individual before her death; the single live-born kid died of a lung defect within minutes.}
[^vandervalk2021]: `paper` van der Valk, T. et al. "Million-year-old DNA sheds light on the genomic history of mammoths." *Nature*, 2021.
[^shapiro2015]: `paper` Shapiro, B. "Mammoth 2.0: will genome engineering resurrect extinct species?" *Genome Biology*, 2015. {Written years before the author became chief science officer at Colossal Biosciences, the company whose claims the argument now bears on.}
[^iucn2016]: `report` IUCN Species Survival Commission. *IUCN SSC Guiding Principles on Creating Proxy Species for Conservation Benefit*. International Union for Conservation of Nature, 2016. {Guidance from a conservation body with no regulatory authority; it binds no project and sets a conservation-benefit test rather than a taxonomic one.}
[^bennett2017]: `paper` Bennett, J. R. et al. "Spending limited resources on de-extinction could lead to net biodiversity loss." *Nature Ecology & Evolution*, 2017.

==============================================================================
ARTICLE: decellularized-scaffolds
TITLE:   Decellularized scaffolds
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/decellularized-scaffolds
SOURCE:  https://futurehumanwiki.com/raw/decellularized-scaffolds
==============================================================================

---
title: "Decellularized scaffolds"
slug: "decellularized-scaffolds"
type: "technology"
status: "emerging"
horizon: "2030s"
trl: 6
categories: ["bodies"]
tags: ["extracellular matrix", "regenerative medicine", "transplantation", "biomaterials", "research misconduct", "tissue engineering"]
summary: "Donor tissues stripped of their cells to leave an extracellular matrix scaffold, used as surgical implants and as a route to rebuilding whole organs from a preserved vascular tree."
updated: "2026-07-27"
humanEvidence: "Acellular matrix implants are in routine surgical use in people, and a small uncontrolled study reported new muscle formation at sites of traumatic muscle loss; recellularized whole organs remain a rodent result."
access: "Acellular dermis, submucosa, heart valves and nerve grafts are approved surgical products in wide use; no recellularized whole organ is available at any price."
reversibility: "difficult"
issues: ["The routine-products section lists product classes without a citation.", "The volumetric muscle loss study is cited without its size or design."]
---

```infobox
{
  "caption": "Biomaterial and organ-engineering method",
  "rows": [
    { "label": "Principle", "value": "Remove cells, keep matrix" },
    { "label": "Common agents", "value": "SDS, Triton X-100, deoxycholate" },
    { "label": "Residual DNA target", "value": "Under 50 ng per mg dry weight" },
    { "label": "Routine products", "value": "Dermis, submucosa, valves, nerve" },
    { "label": "Whole-organ status", "value": "Rodent proof of concept" },
    { "label": "Key failure mode", "value": "Incomplete re-endothelialization" },
    { "label": "Cautionary case", "value": "Macchiarini trachea implants" }
  ]
}
```

**Decellularized scaffolds** are tissues or whole organs from which the cells have been removed, leaving the extracellular matrix intact. The matrix retains the tissue's three-dimensional architecture, its mechanical properties, its basement membranes and, critically for organ work, its vascular tree. Acellular matrix products of this kind are routine in surgery. Whole decellularized organs, reseeded with a recipient's cells, are the most structurally complete approach anyone has to growing a replacement organ, and remain a rodent-scale result.

## How it works

Decellularization removes cellular material without destroying the matrix that surrounds it — a balance, since every agent aggressive enough to lyse cells also degrades matrix components.

Ionic detergents such as sodium dodecyl sulfate strip cells efficiently but denature proteins and remove glycosaminoglycans and matrix-bound growth factors. Non-ionic detergents such as Triton X-100 are gentler and less complete. Most protocols combine them with enzymatic steps, usually a nuclease to digest residual DNA, and with physical treatments such as freeze-thaw cycling.

For thin tissues the agents are applied by immersion and agitation. For whole organs the decisive technique is perfusion decellularization: cannulating the organ's artery and pushing detergent through the native vasculature, so that every cell is reached through the same route that once supplied it with blood. This preserves the vascular architecture down to the capillary bed, which is the property no fabrication method can currently reproduce and the reason the approach persists despite its difficulties.

Adequacy is judged against quantitative criteria rather than appearance: no visible nuclear material on staining, residual double-stranded DNA below roughly fifty nanograms per milligram of dry tissue, and remaining fragments shorter than about two hundred base pairs.[^crapo2011] Residual DNA and residual detergent are both associated with inflammatory responses in the recipient.

## What the matrix does

An acellular matrix is not inert packing. It carries laminin and fibronectin binding sites that direct cell attachment, collagen architecture that sets stiffness and alignment, and sequestered growth factors released as the matrix is degraded. Stephen Badylak's work established that implanted matrix does not simply persist as a scaffold but is remodelled: it recruits host cells, degrades over weeks, and biases infiltrating macrophages toward a reparative rather than inflammatory phenotype, producing what he termed constructive remodelling. In a small clinical study, matrix implants placed into sites of traumatic volumetric muscle loss were associated with new muscle formation and functional improvement.[^sicari2014] The result is modest against the standard set by animals capable of true [[limb-regeneration]], and it is the closest thing to induced structural regeneration achieved in human patients.

Digested matrix is also used as a culture substrate, replacing tumour-derived basement membrane gel in [[organoids]] work and serving as a tissue-specific component of printable bioinks.

This immunological steering is now understood as central rather than incidental, and connects to the finding, established in mice rather than in patients, that scaffold-driven regeneration depends on type 2 immune signalling, discussed in [[tissue-engineering]].

## Products in routine use

The commercially successful applications are all thin, load-bearing or barrier tissues where host cells can migrate in and vascularize the graft from the edges.

Acellular human dermal matrix is used in breast reconstruction, abdominal wall repair and burn care. Porcine small intestinal submucosa is used for hernia repair, wound care and soft tissue reinforcement. Decellularized human and porcine heart valves are implanted, particularly in children, where a valve that the recipient's cells can populate and that grows with the patient is worth a great deal. Decellularized nerve allografts are a standard option for bridging peripheral nerve gaps. Decellularized bone and tendon are widespread in orthopaedics.

None of these is a lab-grown organ. They are matrices that the patient's body converts into tissue, which is a lower bar and a more reliable one.

## Whole-organ decellularization

```timeline
[
  { "year": "2008", "title": "Bioartificial rat heart", "text": "Harald Ott and Doris Taylor perfusion-decellularize a rat heart, reseed it with neonatal cardiac cells, and record weak but coordinated contraction under electrical stimulation." },
  { "year": "2010", "title": "Lung and liver", "text": "Groups at Yale, Massachusetts General and Harvard report decellularized rat lungs supporting gas exchange for hours after implantation, and recellularized rat livers metabolizing drugs in perfusion." },
  { "year": "2013", "title": "Kidney", "text": "A regenerated rat kidney, reseeded with endothelial and epithelial cells, produces dilute urine when perfused and after orthotopic transplantation." },
  { "year": "2010s", "title": "Scale-up attempts", "text": "Human and porcine hearts, lungs, livers and kidneys are decellularized successfully; recellularizing them to functional cell numbers is not achieved." },
  { "year": "2020s", "title": "Assist devices rather than implants", "text": "Companies pivot toward externally perfused humanized pig livers as temporary support, a use that requires far less than a transplantable organ." }
]
```

Each of the rodent results represented a genuine first and a small fraction of native function. The decellularized rat heart generated a fraction of the pressure of a working heart; the reseeded kidney produced urine far more dilute than a native one. Scaling to human organs multiplies the cell requirement roughly a thousandfold and lengthens every diffusion path. Supplying that many cells of the right types is itself a manufacturing problem, addressed in most current work by differentiating [[induced-pluripotent-stem-cells]], and for the heart the practical alternative has remained the mechanical pump described in [[artificial-heart]].

The specific barrier is the endothelium. A decellularized vascular tree presents bare collagen to blood, which is thrombogenic; unless the entire luminal surface is recoated with functional endothelial cells, the graft clots within hours of perfusion. The interaction of blood with an imperfect artificial surface is the same constraint that governs mechanical circulatory support and the oxygen carriers described in [[artificial-blood]]. Achieving complete re-endothelialization across the whole tree, including capillaries, has not been demonstrated at human scale. The practical response has been to lower the goal: rather than an implantable organ, an externally perfused humanized pig liver used as temporary support for liver failure, where blood contacts the device outside the body and anticoagulation is manageable.

## The Macchiarini case

The field's most damaging episode began with a decellularized scaffold and continued with synthetic ones. In 2008 Paolo Macchiarini's group reported implanting a decellularized donor trachea seeded with the recipient's own cells into a woman with end-stage airway disease, and described the result as a success.[^macchiarini2008] From 2011, at the Karolinska Institute, he implanted synthetic polymer tracheas seeded with bone marrow cells into patients including a young child. Most recipients died. There had been no adequate large-animal preclinical work, the interventions were justified as compassionate use in patients who were not all critically ill, and published follow-up overstated the outcomes.

Four Karolinska clinicians who raised concerns were investigated and reprimanded before the allegations were substantiated. An external inquiry found misconduct, Macchiarini was dismissed in 2016, and a 2011 paper describing the synthetic trachea was retracted in 2018. Swedish courts pursued criminal charges; an appeals court convicted him of gross assault in 2023 and imposed a prison sentence.

> [!caution] What the case actually shows
> The failure was not that the technology was implausible in principle. It was procedural: no controls, no preclinical foundation, patient selection driven by the surgeon, institutional incentives favouring a celebrated recruit, and a compassionate-use pathway that permitted repeated first-in-human procedures without accumulating evidence. Any regenerative therapy can be run this way, and the safeguards that failed were not specific to tracheas. The structural resemblance to the death of Jesse Gelsinger, which set back [[somatic-gene-therapy]] by years, and to the [[he-jiankui-affair]] is close: in each case a researcher moved to humans ahead of the evidence, and institutional review either failed or was bypassed.

## Limitations

Beyond re-endothelialization, four constraints recur. Residual antigens in xenogeneic matrix, notably the alpha-gal epitope that also drives rejection in [[xenotransplantation]], provoke immune responses even after cells are removed. Detergent residues are cytotoxic and hard to wash out of dense matrix. Decellularization degrades mechanical strength, which matters for valves and vessels expected to last decades. And obtaining human donor organs to decellularize competes with using them directly, so the scalable version of the approach depends on animal organs, with the antigenicity that implies.

Regulators have responded to the trachea cases by tightening the compassionate-use route for engineered tissues, an instance of the [[precautionary-principle]] operating retrospectively: the restrictions arrived after the deaths rather than before them.

## Outlook

Decellularized matrix has an assured place as a surgical biomaterial and as a bioink component for [[organ-bioprinting]]. Its role in whole-organ replacement is less certain. The strongest argument for it is that the vascular tree is the hardest part of an organ to build and the easiest part to inherit; the strongest argument against is that nearly two decades after the rat heart, no group has recellularized a human-scale organ to a functional cell density.

The near-term test is whether externally perfused bioengineered organs work as bridging therapy in liver failure, which would validate the recellularization chemistry without requiring implantation. If that fails, the approach's contribution to the [[organ-shortage]] would be as a scaffold material rather than as an organ, leaving the supply problem to [[lab-grown-organs]] by other routes and to edited animals.

## See also

- [[tissue-engineering]]
- [[organ-bioprinting]]
- [[lab-grown-organs]]
- [[xenotransplantation]]
- [[organ-shortage]]
- [[organoids]]
- [[artificial-heart]]
- [[limb-regeneration]]

## References

[^crapo2011]: `paper` Crapo, P. M., Gilbert, T. W. and Badylak, S. F. "An overview of tissue and whole organ decellularization processes." *Biomaterials*, 2011. {A review; the widely quoted residual-DNA thresholds originate here as proposed criteria rather than as values derived from clinical outcomes.}
[^sicari2014]: `paper` Sicari, B. M. et al. "An acellular biologic scaffold promotes skeletal muscle formation in mice and humans." *Science Translational Medicine*, 2014. {The human arm is a small open-label case series with no control group; the controlled comparisons in the paper are in mice.}
[^macchiarini2008]: `paper` Macchiarini, P. et al. "Clinical transplantation of a tissue-engineered airway." *The Lancet*, 2008. {The paper that reported the first case as a success; its lead author was later found to have committed misconduct in the trachea programme that followed.}

==============================================================================
ARTICLE: deep-brain-stimulation
TITLE:   Deep brain stimulation
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/deep-brain-stimulation
SOURCE:  https://futurehumanwiki.com/raw/deep-brain-stimulation
==============================================================================

---
title: "Deep brain stimulation"
slug: "deep-brain-stimulation"
type: "intervention"
status: "established"
horizon: "present"
trl: 9
categories: ["cybernetics"]
tags: ["neuromodulation", "parkinsons", "implants", "psychiatry", "closed-loop", "neuroethics"]
summary: "A surgical therapy in which implanted electrodes deliver continuous electrical pulses to specific deep brain structures to suppress the symptoms of movement and psychiatric disorders."
updated: "2026-07-27"
humanEvidence: "Randomized trials in adults with Parkinson's disease report large gains in mobility and quality of life against best medical therapy, and the device is in routine clinical use; two randomized trials in depression were halted for futility."
access: "Approved for Parkinson's disease, essential tremor, dystonia and drug-resistant epilepsy, reimbursed by many high-income systems, and delivered only at specialist centres; psychiatric use is investigational beyond a US exemption for OCD."
reversibility: "difficult"
issues: ["The 150,000-patient figure in the infobox has no source."]
---

```infobox
{
  "caption": "Implanted neuromodulation therapy",
  "rows": [
    { "label": "Type", "value": "Therapeutic neuromodulation" },
    { "label": "Modern technique from", "value": "1987, Grenoble" },
    { "label": "Key figures", "value": "Alim-Louis Benabid, Pierre Pollak" },
    { "label": "Common targets", "value": "STN, GPi, VIM thalamus" },
    { "label": "Typical parameters", "value": "130–185 Hz, 60–90 µs" },
    { "label": "First US approval", "value": "1997 (tremor)" },
    { "label": "Patients treated", "value": "150,000+ worldwide" },
    { "label": "Readiness", "value": "Routine clinical care" }
  ]
}
```

**Deep brain stimulation** (DBS) is the delivery of continuous electrical pulses through electrodes
implanted in specific subcortical structures, controlled by a pulse generator placed under the skin
of the chest. It is the most widely used implanted neurotechnology after the
[[cochlear-implant]], and it is unusual among the technologies covered on this wiki in being both
routine and poorly understood: DBS reliably suppresses the motor symptoms of Parkinson's disease
without a settled account of why.

```figure
{"key": "dbs-electrode-placement", "caption": "Lead placement reconstructed against the surrounding structures. A few millimetres decide which circuit is stimulated."}
```

## How it works

A quadripolar or segmented lead, roughly 1.3 mm in diameter, is placed stereotactically with
millimetre accuracy, often with the patient awake so that clinical response and side effects can be
tested intraoperatively. The lead is tunnelled under the scalp and connected to an implanted pulse
generator. Stimulation is typically delivered at 130–185 Hz with pulse widths of tens of
microseconds, and each contact can be turned on independently, allowing a clinician to steer the
field away from structures that produce side effects.

The mechanism is contested. High-frequency stimulation was first assumed to inhibit the target,
because its effects resemble those of a lesion. The evidence now points elsewhere. Stimulation
excites axons more readily than cell bodies, so a high-frequency train imposes a regular,
information-free firing pattern on the output pathway — an "informational lesion" that overwrites
pathological bursting rather than silencing it. Antidromic activation of cortical afferents,
modulation of the wider cortico-basal-ganglia loop, and effects on astrocytes and local blood flow
all contribute. What DBS does not do is restore normal circuit function, and this matters
clinically: it treats symptoms driven by abnormal network dynamics and does nothing for the
underlying degeneration.

## Development history

```timeline
[
  { "year": "1950s–1960s", "title": "Stereotactic stimulation explored", "text": "Neurosurgeons use intraoperative stimulation to localize targets before making destructive lesions, and a handful of chronic stimulation experiments are attempted for pain and psychiatric illness." },
  { "year": "1987", "title": "Chronic thalamic stimulation for tremor", "text": "Alim-Louis Benabid and Pierre Pollak in Grenoble report that chronic high-frequency stimulation of the ventral intermediate thalamus suppresses tremor as effectively as thalamotomy, and is reversible and adjustable." },
  { "year": "1997", "title": "First US approval", "text": "The FDA approves thalamic stimulation for essential tremor and parkinsonian tremor." },
  { "year": "2002", "title": "Subthalamic and pallidal targets approved", "text": "Approval extends to the subthalamic nucleus and globus pallidus interna for advanced Parkinson's disease, the indication that now accounts for most implants." },
  { "year": "2003–2009", "title": "Dystonia and OCD", "text": "Humanitarian device exemptions cover dystonia and, later, treatment-resistant obsessive-compulsive disorder." },
  { "year": "2013–2017", "title": "Depression trials fail", "text": "Two industry-sponsored randomised trials of DBS for treatment-resistant depression are halted for futility, chilling psychiatric applications for several years." },
  { "year": "2018", "title": "Epilepsy indication", "text": "Stimulation of the anterior nucleus of the thalamus is approved in the United States for drug-resistant focal epilepsy." },
  { "year": "2020s", "title": "Sensing and adaptive systems", "text": "Pulse generators capable of recording local field potentials reach the market, and the first commercial closed-loop systems that adjust stimulation to beta-band activity are cleared in Europe and the United States in 2025." }
]
```

## Established indications

Parkinson's disease is the dominant use. Randomised trials comparing stimulation of the subthalamic
nucleus with best medical therapy in patients with motor fluctuations found substantial improvement
in quality of life and in time spent with good mobility.[^deuschl2006] A later trial extended the
benefit to patients earlier in the disease course, with fluctuations but less disability than the
classical surgical candidate.[^schuepbach2013] Stimulation of the globus pallidus interna produces
comparable motor benefit; subthalamic stimulation permits larger reductions in dopaminergic
medication, while pallidal stimulation is associated with fewer neuropsychiatric effects in some
comparisons.[^follett2010]

Essential tremor and dystonia are the other well-established indications. Tremor suppression by
thalamic stimulation is often dramatic, though habituation over years is common. Dystonia responds
more slowly, over weeks to months, which complicates programming.

Drug-resistant epilepsy is treated with thalamic stimulation, and a separate class of responsive
neurostimulation devices detects seizure onset from cortical electrodes and delivers stimulation
only when a pattern is recognized — closer in architecture to a
[[brain-computer-interface]] than to conventional DBS.

## Psychiatric applications

Psychiatric DBS is the field's cautionary tale and its most interesting open question. Two
industry-sponsored randomised trials of stimulation for treatment-resistant depression, one
targeting the subcallosal cingulate and one the ventral capsule/ventral striatum, were stopped
after interim analyses indicated they would not separate from sham.[^holtzheimer2017] Open-label
follow-up of subcallosal cingulate patients nonetheless reported durable response over years in a
majority of a small cohort, suggesting that the failures reflected trial design, target
localization, and the slow time course of antidepressant response rather than a null effect.

Two developments revived the area. Tractography-guided targeting individualizes the electrode
position to a patient's own white-matter bundles rather than to atlas coordinates. And a
single-patient demonstration used a chronically implanted sensing-and-stimulation device to identify
a personal biomarker of depressed state in the amygdala and trigger ventral striatum stimulation
only when it appeared, an approach that borrows directly from [[neural-decoding]].[^scangos2021]
That result involves one participant. A pivotal trial in treatment-resistant depression is under
way as of 2026; obsessive-compulsive disorder remains approved under a humanitarian exemption with a
modest evidence base. Treatment-resistant depression draws every available modality, down to the
drug-plus-psychotherapy protocols described under [[psychedelic-therapy]], and the recurring
difficulty is common to all of them: separating what a treatment does from what a patient who knows
they are being treated reports.

## Adaptive stimulation

Conventional DBS runs open loop: fixed parameters, day and night, regardless of what the brain is
doing. Adaptive DBS uses the implanted lead as a sensor, extracting a control signal — most often
power in the beta band, 13–30 Hz, which correlates with rigidity and bradykinesia in Parkinson's
disease — and modulating stimulation amplitude in response. Early crossover studies showed that
adaptive stimulation could match or exceed continuous stimulation while delivering substantially
less total current, which reduces stimulation-induced dyskinesia and speech disturbance and extends
battery life.[^little2013]

Commercial sensing generators arrived in 2020, and closed-loop control reached the market in 2025.
The engineering constraints are the interesting part: the stimulation artefact swamps the signal
being sensed, biomarkers drift, and a controller that misreads state can oscillate. These are the
same problems that limit any chronic implanted decoder, and progress here transfers directly to
[[neuroprosthetics]] more broadly.

## Effects on personality and agency

Stimulation of the subthalamic nucleus can produce impulsivity, hypomania, pathological gambling,
hypersexuality, and apathy, and these effects are dose- and contact-dependent. They arise partly
from the target's proximity to limbic and associative territory and partly from the rapid reduction
in dopaminergic medication that surgery permits. Most resolve with reprogramming or medication
adjustment. A distinct and less tractable phenomenon is the mismatch that sometimes appears between
restored motor function and a patient's psychosocial adjustment.

Qualitative studies of DBS recipients record a spectrum of reported self-change, from patients who
say the device restored them to themselves to a minority who describe estrangement from their own
behaviour.[^gilbert2017] The philosophical literature has used these reports to press questions
about authenticity and authorship of action that also arise, in weaker form, for psychopharmacology,
and that bear directly on [[personal-identity-and-continuity]] and on proposals for
[[moral-enhancement]]. Empirical work suggests the frequency of severe personality change has been
overstated in the [[bioethics-of-enhancement]] literature relative to what clinical series show. The
more common complaint is mundane and important: dependence on a device, a battery, and a clinic.

> [!note] Consent under a working device
> A patient whose symptoms are controlled by stimulation may be unable to assess the treatment
> without it, since turning the device off restores the disorder. This makes withdrawal of consent
> practically asymmetric in a way that ordinary drug trials avoid, and it is a recurring theme in
> [[neurorights]] discussions of implanted therapy.

## Limitations and risks

Intracranial haemorrhage occurs in a small percentage of implantations and is the most serious
surgical complication. Hardware problems — infection requiring explantation, lead fracture or
migration, skin erosion over the connector — are more common than brain injury and account for much
of the reoperation burden. Stimulation-induced side effects include dysarthria, paraesthesia, gait
freezing, and eyelid apraxia; most are reversible by reprogramming, which is itself a labour-intensive
process requiring an expert clinician and, historically, many hours of trial and error.

DBS does not slow disease progression. Axial symptoms — postural instability, freezing of gait,
speech and swallowing difficulty — respond poorly and continue to worsen, and only a
disease-modifying approach such as dopaminergic cell replacement from
[[induced-pluripotent-stem-cells]] would address the underlying loss. Cost and the requirement
for a specialist centre restrict access sharply, and the geographic distribution of implantation
rates tracks health-system wealth rather than disease burden, an instance of the pattern examined in
[[access-and-inequality]].

## Outlook

Three trajectories are visible. Sensing-enabled hardware turns every implanted patient into a source
of chronic intracranial recordings, a data resource with no precedent in human neuroscience and
obvious implications for [[mental-privacy]]. Directional leads and patient-specific field models
built from a recipient's own imaging, a clinical instance of the approach described under
[[human-digital-twins]], are making programming semi-automatic. And non-invasive alternatives are
eroding the market at the margins: magnetic-resonance-guided focused ultrasound already ablates the thalamic tremor target
without an incision, and low-intensity variants under [[non-invasive-neuromodulation]] aim to
modulate deep structures reversibly.

The unresolved scientific question is whether DBS can be made specific. Present electrodes excite
every axon within a radius, regardless of type or projection target. Techniques such as
[[optogenetics]] achieve cell-type specificity in animal brains, and the only published human use of
that method is in the eye; reaching a deep brain structure the same way would require gene delivery
to a defined population and a light source implanted beside it. Whether an electrical device can
approach that selectivity, or whether specificity requires a biological intervention, determines how
far DBS can extend beyond the movement disorders it currently treats — and whether the technique
ever has anything to offer the elective modification of healthy brains discussed under
[[human-enhancement]], which on present evidence it does not.

## See also

- [[neuroprosthetics]]
- [[brain-computer-interface]]
- [[non-invasive-neuromodulation]]
- [[memory-prosthesis]]
- [[optogenetics]]
- [[neurorights]]
- [[personal-identity-and-continuity]]
- [[cochlear-implant]]

## References

[^deuschl2006]: `paper` Deuschl, G. et al. "A randomized trial of deep-brain stimulation for Parkinson's disease." *New England Journal of Medicine*, 2006.

[^schuepbach2013]: `paper` Schuepbach, W. M. M. et al. "Neurostimulation for Parkinson's disease with early motor complications." *New England Journal of Medicine*, 2013.

[^follett2010]: `paper` Follett, K. A. et al. "Pallidal versus subthalamic deep-brain stimulation for Parkinson's disease." *New England Journal of Medicine*, 2010.

[^holtzheimer2017]: `paper` Holtzheimer, P. E. et al. "Subcallosal cingulate deep brain stimulation for treatment-resistant depression: a multisite, randomised, sham-controlled trial." *The Lancet Psychiatry*, 2017. {The trial was stopped at a futility analysis rather than completed, so it establishes that the endpoint would not be met, not that stimulation has no effect.}

[^scangos2021]: `paper` Scangos, K. W. et al. "Closed-loop neuromodulation in an individual with treatment-resistant depression." *Nature Medicine*, 2021. {A single participant whose biomarker and stimulation site were mapped individually beforehand; a feasibility demonstration, not an efficacy estimate.}

[^little2013]: `paper` Little, S. et al. "Adaptive deep brain stimulation in advanced Parkinson disease." *Annals of Neurology*, 2013. {A short crossover in a handful of patients with temporarily externalised leads, not a chronic comparison in fully implanted devices.}

[^gilbert2017]: `paper` Gilbert, F., Goddard, E., Viaña, J. N. M., Carter, A., Horne, M. "I miss being me: phenomenological effects of deep brain stimulation." *AJOB Neuroscience*, 2017. {A qualitative study; it characterises the kinds of self-change recipients describe and cannot say how common any of them are.}

==============================================================================
ARTICLE: designer-babies
TITLE:   Designer babies
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/designer-babies
SOURCE:  https://futurehumanwiki.com/raw/designer-babies
==============================================================================

---
title: "Designer babies"
slug: "designer-babies"
type: "concept"
status: "contested"
horizon: "present"
categories: ["reproduction", "society"]
tags: ["reproduction", "embryo selection", "germline", "bioethics", "genomics", "enhancement"]
summary: "A popular term for children whose traits are chosen before birth, covering a real but narrow practice of embryo selection and a largely hypothetical practice of embryo editing."
updated: "2026-07-27"
issues: ["The claim about the size of achievable polygenic selection gains carries no source."]
---

```infobox
{
  "caption": "Popular term in reproductive genetics",
  "rows": [
    { "label": "Type", "value": "Journalistic term" },
    { "label": "In circulation since", "value": "1990s" },
    { "label": "Realised form", "value": "Embryo selection", "link": "/wiki/embryo-selection" },
    { "label": "Hypothetical form", "value": "Heritable editing", "link": "/wiki/germline-editing" },
    { "label": "Known edited births", "value": "Three, 2018–2019", "link": "/wiki/he-jiankui-affair" },
    { "label": "Main technical limit", "value": "Polygenicity" }
  ]
}
```

**Designer babies** is a popular term for children whose genetic characteristics have been deliberately chosen by their parents. It covers two very different things. One is [[embryo-selection]], which is routine in fertility clinics and can reliably deliver only a short list of outcomes. The other is heritable [[germline-editing]], which has been attempted once, illegally, and would face a different and harder set of technical obstacles. Most public argument treats the two as a single prospect, which is where the confusion begins.

## Origins of the term

The phrase is journalistic rather than technical, built on the same pattern as "designer drugs" and in wide circulation by the 1990s. It attached itself to a sequence of real clinical developments. In vitro fertilisation produced its first birth in 1978; preimplantation genetic diagnosis followed in 1990, when a British team biopsied embryos and selected female ones for couples at risk of X-linked disease.[^handyside1990] The film *Gattaca* fixed the popular image in 1997, some years before any of the genomic tools that would be needed to realise it existed.

A defining case arrived in 2000, when a couple used preimplantation diagnosis both to avoid Fanconi anaemia and to select an embryo whose tissue type matched their affected daughter's, so that cord blood from the resulting child could treat her.[^verlinsky2001] The "saviour sibling" framing dominated coverage for a decade and set the pattern for subsequent debate: a specific, medically motivated use, discussed as though it were the first step of a general capability.

## What selection can actually deliver

Selection is constrained by what the parents already carry. An IVF cycle produces a set of embryos that are recombinations of two existing genomes; no embryo can have an allele neither parent possesses. Within that set, selection is highly reliable for categorical outcomes determined by one locus or by chromosome count.

- **Single-gene disorders.** Where both parents carry a recessive allele, roughly a quarter of embryos are affected and can be identified with high accuracy. This is PGT-M, and it works.
- **Chromosomal aneuploidy.** PGT-A screens for the wrong number of chromosomes. Its value in improving live-birth rates remains debated, but the measurement itself is dependable.
- **Sex.** Sex is a categorical genetic fact, which is why it is the one non-medical trait selection delivers with near certainty, and why most jurisdictions that regulate embryo testing address it explicitly.
- **Tissue type.** HLA matching, as in the saviour-sibling case, is likewise a discrete genotype.

Everything else is probabilistic. [[polygenic-embryo-screening]] extends selection to traits shaped by thousands of variants, and the expected shift it can produce is a fraction of the variation within a single family. The constraint is not the accuracy of the laboratory work but the small genetic distance between siblings and the small number of embryos a cycle yields.

> [!key] The arithmetic that popular coverage skips
> A couple choosing among a handful of their own embryos is sampling from a narrow distribution. Even a perfect predictor of a trait could only pick the best of what those particular gametes happened to produce, which for most traits is a modest range.

## What editing would add, and what it would not

Editing changes the constraint. [[crispr-cas9]] and its derivatives can in principle install an allele neither parent carries, which selection can never do. That matters in a small number of situations where selection genuinely fails: when one parent is homozygous for a dominant disease allele, when both are homozygous for a recessive one, or when a couple produces too few embryos for selection to have anything to work with. These cases are real but rare, and they are the strongest medical argument for heritable editing.

For trait enhancement, editing runs into the same wall as selection, from a different direction. Traits such as height or cognitive test performance are influenced by thousands of variants, each shifting the outcome by a tiny amount, so a noticeable change means editing at a scale nobody has attempted in an embryo of any species. The number of changes required and the tolerance of a one-cell embryo for damage point in opposite directions. [[base-editing]] and [[prime-editing]], which rewrite a base without severing both strands of the helix, are the chemistries that would have to carry such an attempt; both were developed for single, well-characterised disease alleles rather than for thousands of small-effect ones. Nuclease editing at several sites at once carries a documented risk of the [[crispr-off-target-effects|large deletions and rearrangements]] that a biopsy of a few cells would not reliably detect.

Embryos add a further problem. Editing a one-cell zygote often produces a mosaic, an individual whose cells do not all carry the intended change, and mosaicism cannot be reliably detected by biopsying a few trophectoderm cells. Research on non-viable and research-consented human embryos has reported edits at target loci, though the mechanism claimed in the most publicised case was disputed by other groups who argued that apparent correction reflected allele dropout rather than repair.[^ma2017] The [[he-jiankui-affair]] of 2018 illustrated the whole set of failures at once: the intended CCR5 deletion was not faithfully reproduced, at least one child was mosaic, and the medical rationale did not hold up.

## Polygenicity as the real constraint

The gap between the popular image and the genetics is almost entirely explained by polygenicity. Public discussion imagines a menu of traits with switches attached, because Mendelian disease provides the mental model and Mendelian disease is genuinely switch-like. Nearly everything people would want to choose is not.

This is the central point of [[genetic-enhancement-of-intelligence]] and applies equally to athletic ability, temperament and appearance. The variants involved are numerous, individually negligible, context-dependent, and pleiotropic, meaning the same variant influences several traits at once. A technology that can flip one switch cleanly is not on a continuum with a technology that can retune ten thousand dials.

> [!caution] What has actually been done
> As of 2026 the only publicly documented births following deliberate genome editing remain the three children of the He Jiankui experiment. Everything else described as a designer baby has involved selecting among embryos, choosing a gamete donor, or, in the distinct case of [[mitochondrial-replacement-therapy]], substituting mitochondria rather than editing nuclear DNA.

## Ethics and law

The ethical literature separates arguments that apply to selection from those that apply to editing. Selection debates centre on the expressivist objection developed within [[disability-rights-and-enhancement]], which holds that screening out a trait expresses a judgement about people who have it, and on [[procreative-beneficence]], Julian Savulescu's principle that parents have reason to choose the child expected to have the best life.[^savulescu2001] Editing debates add the questions of consent from a person who does not yet exist, of irreversibility across generations, and of the therapy-enhancement boundary examined in [[human-enhancement]].

Bioconservative critics, surveyed in [[bioconservatism]], argue that treating a child's characteristics as a parental design choice alters the relationship between generations regardless of whether the technique works. Distributional critics make a different argument, set out in [[access-and-inequality]]: a capability priced at the level of an IVF cycle plus genomic analysis is not evenly available, and one available only to the wealthy could compound advantage rather than health.

Law is fragmented. Heritable editing is prohibited outright in many countries and unfunded rather than banned in others; embryo testing is licensed condition-by-condition in some jurisdictions and unregulated in others. An international commission convened by national academies concluded that no clinical use of heritable editing should proceed until safety and efficacy could be established for a narrow set of indications, and that no country then had a pathway for authorising it.[^nasem2020] The details, and the weakness of enforcement, are covered in [[governance-of-genome-editing]].

## Outlook

Two technical developments would move the term closer to its popular meaning. [[in-vitro-gametogenesis]], if it works in humans, would turn selection from a choice among a handful of embryos into a choice among hundreds, which is the only route by which polygenic selection could produce changes large enough to notice. Reliable multiplex editing of embryos, with mosaicism solved, would remove the parental-genome constraint entirely.

Neither is close, and the second may never be permitted. The open question is whether a capability that arrives gradually, one condition at a time and framed each time as disease prevention, ever confronts the public with the choice that "designer babies" names, or whether it simply passes without anyone having decided anything.

## See also

- [[embryo-selection]]
- [[polygenic-embryo-screening]]
- [[germline-editing]]
- [[he-jiankui-affair]]
- [[in-vitro-gametogenesis]]
- [[procreative-beneficence]]
- [[genetic-enhancement-of-intelligence]]
- [[bioethics-of-enhancement]]

## References

[^handyside1990]: `paper` Handyside, A. H. et al. "Pregnancies from biopsied human preimplantation embryos sexed by Y-specific DNA amplification." *Nature*, 1990. {The embryos were sexed rather than tested for the disease itself, which is why the first clinical uses were all X-linked conditions.}
[^verlinsky2001]: `paper` Verlinsky, Y. et al. "Preimplantation diagnosis for Fanconi anemia combined with HLA matching." *JAMA*, 2001.
[^ma2017]: `paper` Ma, H. et al. "Correction of a pathogenic gene mutation in human embryos." *Nature*, 2017. {Other groups argued the apparent correction reflected allele dropout rather than interhomolog repair; the interpretation is still disputed.}
[^savulescu2001]: `paper` Savulescu, J. "Procreative Beneficence: Why We Should Select the Best Children." *Bioethics*, 2001.
[^nasem2020]: `report` International Commission on the Clinical Use of Human Germline Genome Editing. *Heritable Human Genome Editing*. National Academies Press, 2020. {A consensus report convened by national academies; it proposes criteria for a first clinical use and has no legal force in any country.}

==============================================================================
ARTICLE: dietary-supplements
TITLE:   Dietary supplements
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/dietary-supplements
SOURCE:  https://futurehumanwiki.com/raw/dietary-supplements
==============================================================================

---
title: "Dietary supplements"
slug: "dietary-supplements"
type: "intervention"
status: "contested"
horizon: "present"
trl: 9
categories: ["longevity", "society"]
tags: ["supplements", "regulation", "vitamins", "clinical trials", "consumer health", "antioxidants"]
summary: "Vitamins, minerals, botanicals, and other compounds sold without pre-market approval, whose large randomized trials have repeatedly failed to show benefit in well-nourished adults."
updated: "2026-07-27"
humanEvidence: "Correcting a documented deficiency is established medicine, but large randomized trials in well-nourished adults have been null for cancer and cardiovascular endpoints, and beta-carotene and vitamin E raised cancer incidence in some groups."
access: "Sold without prescription in supermarkets, pharmacies, and online almost everywhere; in the United States no product needs approval before sale, and monthly cost ranges from a few dollars to several hundred."
reversibility: "reversible"
issues: ["Regulation section covers the US, EU, and Australia only.", "No market-size figure; usage and harm data cited are US-only."]
---

```infobox
{
  "caption": "Consumer product category",
  "rows": [
    { "label": "US legal basis", "value": "DSHEA, 1994" },
    { "label": "Pre-market approval", "value": "Not required" },
    { "label": "Efficacy review before sale", "value": "None" },
    { "label": "Claims permitted", "value": "Structure/function, not disease" },
    { "label": "US adult use", "value": "More than half" },
    { "label": "Shown to extend human lifespan", "value": "None" },
    { "label": "Readiness", "value": "TRL 9" }
  ]
}
```

**Dietary supplements** are vitamins, minerals, botanical extracts, amino acids, and other
compounds sold for ingestion outside the drug-approval system. In the United States the category
was created by statute in 1994 and treated as a kind of food: no agency reviews a product for
safety or efficacy before it reaches a shelf. The category contains established medicine, chiefly
the correction of documented deficiencies, wrapped in a much larger commercial layer of compounds
sold for healthy aging, and the large randomized trials of that second layer have mostly returned
null or harmful results.

```keyfacts
[
  { "value": "1994", "label": "Year US supplements were exempted from pre-market review", "note": "DSHEA; the manufacturer, not a regulator, judges safety before sale" },
  { "value": ">50%", "label": "US adults reporting supplement use", "note": "above half in every national survey since the late 1990s" },
  { "value": "0", "label": "Supplements shown to extend human lifespan", "note": "as of 2026" }
]
```

## The regulatory bargain

The Dietary Supplement Health and Education Act of 1994 (DSHEA) defines a dietary supplement as a
food rather than a drug or a food additive.[^dshea] The consequences are structural. A
manufacturer is responsible for concluding that its product is safe, but need not demonstrate that
conclusion to anyone before selling it, and need not show any benefit at all. To remove a product,
the Food and Drug Administration must establish that it presents a significant or unreasonable
risk — after it is on the market and often after people have been hurt. Ingredients not sold in the
United States before October 1994 require a pre-market notification rather than an approval, and
compliance with even that is widely acknowledged to be partial.

Labels may carry structure/function claims ("supports immune health") but not disease claims
("prevents colds"), and must state that the product is not intended to diagnose, treat, cure, or
prevent any disease. Good manufacturing practice rules issued in 2007 and phased in over the
following years bind manufacturers on identity, purity, and record-keeping. Those rules govern
process, not effect: a product can be made exactly as specified and still do nothing.

Other jurisdictions police claims more tightly without changing the underlying deal. The European
Union harmonized permitted vitamin and mineral sources in 2002 and requires health claims to be
authorized on a European Food Safety Authority assessment, which has left most botanical claims
unresolved for over a decade. Australia lists low-risk products against a permitted-ingredients
register without assessing whether they work. In none of these systems must a manufacturer produce
clinical evidence of benefit before selling.

> [!key] The asymmetry
> A drug may not be sold until its maker has convinced a regulator that it works. A supplement may
> be sold until a regulator can prove that it is dangerous. That inversion of the
> [[precautionary-principle|precautionary posture]] applied to medicines explains most of what
> follows: no company needs a trial in order to sell, so almost all of the large trials have been
> paid for by public funders asking whether products already in millions of cabinets do anything.

## Development history

```timeline
[
  { "year": "1970", "title": "Pauling popularizes megadoses", "text": "Linus Pauling's book on vitamin C and the common cold establishes the idea that intakes far above nutritional requirement are therapeutic, a claim later trials did not support." },
  { "year": "1976", "title": "Proxmire Amendment", "text": "US law bars the FDA from limiting the potency of vitamin and mineral supplements or regulating them as drugs on the basis of dose alone." },
  { "year": "1994", "title": "Beta-carotene increases lung cancer", "text": "The Finnish ATBC trial reports higher lung cancer incidence in male smokers given beta-carotene than in those given placebo." },
  { "year": "1994", "title": "DSHEA becomes law", "text": "A bill sponsored by Orrin Hatch and Tom Harkin, passed after a large industry-organized public campaign, exempts supplements from pre-market approval and leaves the FDA to prove a marketed product unsafe." },
  { "year": "1996", "title": "CARET halted early", "text": "A second beta-carotene trial in smokers and asbestos-exposed workers is stopped ahead of schedule for excess lung cancer and higher total mortality in the supplemented arm." },
  { "year": "2004", "title": "Ephedra banned", "text": "After roughly a decade of adverse-event reports and several deaths, the FDA declares ephedrine-alkaloid supplements adulterated; a district court sets the rule aside before an appeals court upholds it in 2006." },
  { "year": "2007", "title": "Manufacturing rules issued", "text": "The FDA finalizes good manufacturing practice requirements for supplements, phased in over three years, covering identity and purity but not efficacy." },
  { "year": "2011", "title": "SELECT reports harm", "text": "Extended follow-up of a prostate cancer prevention trial finds significantly more prostate cancer among men who had taken vitamin E." },
  { "year": "2019", "title": "VITAL reports null primaries", "text": "In more than 25,000 older US adults, neither vitamin D nor marine omega-3 reduces invasive cancer or major cardiovascular events." },
  { "year": "2022", "title": "USPSTF recommends against two", "text": "The US Preventive Services Task Force finds the evidence insufficient for multivitamins and recommends against beta-carotene and vitamin E for preventing cancer or cardiovascular disease." }
]
```

## The trial record

The antioxidant hypothesis gave the field its central prediction. If aging reflects accumulated
oxidative damage, supplementing antioxidants should slow it — an idea still embedded in consumer
marketing long after it stopped being how biogerontologists describe the [[hallmarks-of-aging]].
Tested at scale, it did worse than null. In the Finnish ATBC trial, male smokers given
beta-carotene developed more lung cancer than those given placebo.[^atbc1994] CARET, testing
beta-carotene with vitamin A in smokers and asbestos-exposed workers, was stopped early for the
same reason and for higher overall mortality.[^omenn1996] Extended follow-up of the SELECT prostate
cancer prevention trial found significantly more prostate cancer in the men who had received
vitamin E.[^klein2011] Three large trials, three findings of harm from compounds sold as
protective.

The nutrients with the strongest observational associations have fared no better under
randomization. VITAL enrolled more than 25,000 older US adults and tested vitamin D and marine
omega-3 against placebo; neither reduced the primary cancer or cardiovascular
endpoints.[^manson2019] An ancillary study in the same cohort found no reduction in fractures
either.[^leboff2022] A separate European trial of vitamin D, omega-3, and a home exercise
programme in older adults found no benefit on its
co-primary clinical endpoints, though a later analysis of stored samples reported that vitamin D
slightly slowed several DNA-methylation age estimates. That last result is worth naming precisely:
a shift in an [[epigenetic-clock]] reading is a change in a prediction, not in an outcome, and no
regulator accepts one as a surrogate.

Reviewing this literature in 2022, the US Preventive Services Task Force found the evidence
insufficient to recommend multivitamins or most single nutrients to healthy adults for preventing
cardiovascular disease or cancer, and recommended against beta-carotene and vitamin E
outright.[^uspstf2022]

> [!debate] The multivitamin exception
> A minority position holds that broad-spectrum multivitamins are the one defensible product. It
> rests on a US trial in male physicians reporting a small reduction in total cancer incidence, and
> on cognitive substudies of a later cocoa and multivitamin trial reporting modestly slower decline
> on test batteries in older adults. Critics answer that the effects are small and measured on
> continuous test scores rather than diagnoses. Defenders answer that the products are cheap and
> safe and that real-world nutrient intakes are not uniformly adequate. Unresolved as of 2026.

## Where supplementation is established medicine

Against a deficiency, supplementation is not contested. Iron corrects iron-deficiency anaemia;
vitamin B12 corrects deficiency from malabsorption or from diets containing no animal products;
vitamin D treats rickets and osteomalacia; iodine prevents goitre and developmental injury; vitamin
A reduces child mortality in populations where deficiency is endemic, though the size of that
effect has been disputed since a very large Indian trial reported a smaller one. Periconceptional
folic acid prevents most neural tube defects, a result established by a randomized trial stopped
early once the effect was clear, and one that led many countries to fortify flour.[^mrc1991]

The distinction that organizes the whole subject is between restoring a deficient person to
sufficiency and pushing a sufficient person higher. The first has an evidence base measured in
decades. The second is what the large prevention trials tested, and it is where the record is
null.

## The longevity shelf

A distinct commercial layer sells compounds for healthy aging rather than for deficiency. It
includes [[nad-precursors]]; resveratrol, promoted through the sirtuin research programme
associated with [[david-sinclair]] and since substantially deflated; flavonoids such as fisetin and
quercetin, marketed on their [[senolytics|senolytic]] activity against
[[cellular-senescence|senescent cells]] in mice; spermidine, sold as an inducer of [[autophagy]]
and a mimetic of [[caloric-restriction]], which produced no cognitive benefit in a randomized trial
in older adults with subjective memory complaints; urolithin A, sold on mitophagy and
[[mitochondrial-dysfunction|mitochondrial decline]], with small randomized trials in adults
reporting changes in muscle endurance measures; and taurine, after a 2023 report in *Science* that
supplementation extended median lifespan in mice, with human evidence limited to observational
associations.

The pattern repeats: a mechanism in cells, a result in mice, a small human trial reporting a
biomarker, and a product. The missing step in every case is a randomized trial with a functional
endpoint in people. Marketing fills the gap with movement in a
[[biological-age|composite age estimate]], which is why the absence of validated
[[aging-biomarkers]] matters commercially and not only scientifically. Consumers arriving through
[[biohacking]] communities, [[quantified-self|self-tracking]], and
[[consumer-blood-testing|direct-to-consumer blood panels]] often meet a panel result and a
recommended product in the same interface.

The boundary with drugs is not stable either. [[metformin]] and [[rapamycin]] are prescription
medicines used off-label toward the same goal; [[nootropics]] straddle the line; and US law
excludes from the supplement definition a compound authorized for investigation as a new drug
before it was sold as a supplement, the clause at the centre of the dispute over NMN. The contrast
worth holding is [[glp-1-receptor-agonists|the GLP-1 drugs]] — an approved class with
cardiovascular outcome trials, which is what the supplement market has never had to produce.

## Content, contamination, and harm

Independent testing repeatedly finds products that do not match their labels. Botanical products
have been found substituted or diluted, most publicly in a 2015 New York attorney general action
against major retailers, though that investigation's DNA-barcoding method drew criticism from
analytical chemists because extraction destroys much of the DNA it looks for. Quantities deviate:
an analysis of melatonin gummies sold in the United States, a product taken for
[[sleep-and-longevity|sleep]], found large discrepancies between labelled and measured content. The most serious failure is undeclared pharmaceuticals — sildenafil analogues in
sexual-performance products, sibutramine in weight-loss products, unapproved stimulants in sports
products. One study found banned drugs still present in supplements bought months after FDA
recalls.[^cohen2014]

Harm is not hypothetical. Supplement-related adverse events are a documented cause of emergency
department visits in the United States, concentrated in cardiac symptoms among young adults taking
weight-loss and energy products.[^geller2015] Herbal and dietary supplements account for a growing
share of cases in US drug-induced liver injury registries. Fat-soluble vitamins and selenium
accumulate to toxicity. St John's wort induces drug-metabolizing enzymes and can weaken
prescription medicines that people are relying on.

> [!caution] Strict liability in sport
> Anti-doping rules hold athletes responsible for whatever is found in their bodies, regardless of
> intent. Contaminated or spiked supplements are a recurring source of sanctions, which is why
> sports bodies point athletes to third-party certification schemes. Those schemes verify identity
> and screen for contaminants; none of them assesses whether a product works, a distinction routinely
> blurred in advertising. See [[enhancement-in-sport]].

## Outlook

Nothing in the current structure creates a reason to run the missing trials. Proposals to require
manufacturers to register products with the FDA have been debated for years without being enacted,
and a mandatory listing would improve visibility rather than evidence. Retailer-demanded
third-party verification is spreading and would address identity and contamination, which are real
problems, but not efficacy.

The scientifically useful trials would look different from the ones already run: functional
endpoints such as mobility, infection, or cognition rather than cancer incidence; populations
selected for plausible insufficiency rather than well-nourished volunteers; and durations long
enough to matter. Whether any compound on the longevity shelf survives that test is open. The
record for the compounds already examined at scale is that they did not, and that two of them,
beta-carotene and vitamin E, harmed some of the people who took them. The interventions with the
best functional evidence for [[healthspan|staying healthy longer]],
[[exercise-and-aging|structured exercise]] above all, remain the ones nobody can bottle.
Supplements are the most widely consumed application of the [[geroscience-hypothesis]] and the
least tested.

## See also

- [[nad-precursors]]
- [[senolytics]]
- [[metformin]]
- [[aging-biomarkers]]
- [[exercise-and-aging]]
- [[consumer-blood-testing]]
- [[healthspan]]
- [[geroscience-hypothesis]]

## References

[^dshea]: `law` United States Congress. *Dietary Supplement Health and Education Act of 1994*, Public Law 103-417. {Defines supplements as a food category and leaves the FDA to prove a marketed product unsafe rather than requiring a maker to prove it safe.}
[^atbc1994]: `paper` The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group. "The Effect of Vitamin E and Beta Carotene on the Incidence of Lung Cancer and Other Cancers in Male Smokers." *New England Journal of Medicine*, 1994.
[^omenn1996]: `paper` Omenn, G.S. et al. "Effects of a Combination of Beta Carotene and Vitamin A on Lung Cancer and Cardiovascular Disease." *New England Journal of Medicine*, 1996. {CARET was terminated ahead of schedule because the supplemented arm had more lung cancer and higher total mortality.}
[^klein2011]: `paper` Klein, E.A. et al. "Vitamin E and the Risk of Prostate Cancer: The Selenium and Vitamin E Cancer Prevention Trial (SELECT)." *JAMA*, 2011. {The excess of prostate cancer emerged in extended follow-up, after study supplements had been stopped.}
[^manson2019]: `paper` Manson, J.E. et al. "Vitamin D Supplements and Prevention of Cancer and Cardiovascular Disease." *New England Journal of Medicine*, 2019. {A companion paper reports the same null result for marine omega-3 in the identical cohort.}
[^leboff2022]: `paper` LeBoff, M.S. et al. "Supplemental Vitamin D and Incident Fractures in Midlife and Older Adults." *New England Journal of Medicine*, 2022. {The participants were not selected for low vitamin D status or for osteoporosis, so the result does not speak to treating either.}
[^uspstf2022]: `report` US Preventive Services Task Force. "Vitamin, Mineral, and Multivitamin Supplementation to Prevent Cardiovascular Disease and Cancer: Recommendation Statement." *JAMA*, 2022.
[^mrc1991]: `paper` MRC Vitamin Study Research Group. "Prevention of neural tube defects: results of the Medical Research Council Vitamin Study." *The Lancet*, 1991. {Tested recurrence in women who had already had an affected pregnancy; a separate Hungarian trial addressed first occurrence.}
[^cohen2014]: `paper` Cohen, P.A. et al. "Presence of banned drugs in dietary supplements following FDA recalls." *JAMA*, 2014.
[^geller2015]: `paper` Geller, A.I. et al. "Emergency Department Visits for Adverse Events Related to Dietary Supplements." *New England Journal of Medicine*, 2015.

==============================================================================
ARTICLE: differential-technological-development
TITLE:   Differential technological development
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/differential-technological-development
SOURCE:  https://futurehumanwiki.com/raw/differential-technological-development
==============================================================================

---
title: "Differential technological development"
slug: "differential-technological-development"
type: "concept"
status: "contested"
horizon: "present"
categories: ["society", "foundations"]
tags: ["governance", "risk", "biosecurity", "existential risk", "policy", "coordination"]
summary: "Nick Bostrom's proposal to influence the order in which technologies arrive, hastening protective ones and delaying dangerous ones, rather than trying to halt progress."
updated: "2026-07-27"
issues: ["The US synthetic nucleic acid screening policy is described without a source."]
---

```infobox
{
  "caption": "Proposal in technology policy",
  "rows": [
    { "label": "Proposed by", "value": "Nick Bostrom", "link": "/wiki/nick-bostrom" },
    { "label": "First stated", "value": "2002" },
    { "label": "Core move", "value": "Change ordering, not totals" },
    { "label": "Related framing", "value": "Defensive acceleration (d/acc)" },
    { "label": "Main obstacle", "value": "Coordination and dual use" },
    { "label": "Adoption", "value": "Partial, in biosecurity and AI policy" }
  ]
}
```

**Differential technological development** is the proposal that, since stopping technological progress is neither feasible nor desirable, effort should go into changing the *order* in which capabilities arrive: accelerating technologies that reduce risk and delaying those that create it. The claim is not that dangerous technologies can be prevented indefinitely, but that arrival order determines whether a society meets a hazard with defences already in place or without them.

```timeline
[
  { "year": "2002", "title": "The principle stated", "text": "Nick Bostrom's survey of existential risks proposes retarding dangerous technologies and accelerating protective ones, naming the strategy differential technological development." },
  { "year": "2014", "title": "Restated for machine intelligence", "text": "Superintelligence applies the principle to AI, arguing that safety and control research should be advanced relative to capability research." },
  { "year": "2019", "title": "The vulnerable world hypothesis", "text": "Bostrom formalises the case that some possible inventions are 'black balls' civilisation has no default defence against, and examines what stabilisation would require." },
  { "year": "2023", "title": "Defensive acceleration", "text": "Vitalik Buterin proposes d/acc, framing the goal as accelerating defensive and decentralising technologies specifically rather than technology in general." },
  { "year": "2023–2024", "title": "Screening becomes policy", "text": "US federal policy directs a framework for screening synthetic nucleic acid orders, an example of a defensive capability being deliberately advanced ahead of misuse." }
]
```

## The proposal

[[nick-bostrom]]'s formulation asks policymakers to retard the development of dangerous and harmful technologies, particularly those that raise [[existential-risk]], and to accelerate beneficial ones, particularly those that reduce risks posed by nature or by other technologies.[^bostrom2002] Three features distinguish it from ordinary risk regulation.

It is *comparative*. The target is not any technology's absolute pace but the gap between paired capabilities — offence and defence, capability and interpretability, synthesis and detection.

It is *strategic rather than prohibitive*. It concedes that a determined actor will eventually obtain a capability, and asks only what else exists by then.

It is *neutral about totals*. In principle it is compatible with faster overall progress, provided the sequencing improves, which is why it is invoked both by people who want less technology and by people who want more.

## Origins

The principle appeared in Bostrom's 2002 taxonomy of existential risks and was restated in *Superintelligence*, where the paired capabilities are machine intelligence and the ability to control it.[^bostrom2014] The underlying observation is older: the argument that nuclear weapons arrived before any adequate international arrangement for managing them is the twentieth century's canonical case, and the 1975 [[asilomar-conference]] its counterexample, where a research community paused, built containment standards, and resumed.

The framing has a longer intellectual lineage in the idea of *differential intellectual progress* — the suggestion that risk-reducing understanding should be advanced faster than risk-increasing capability — and it converges with older arms-control reasoning about the offence–defence balance.

## The vulnerable world argument

Bostrom's 2019 paper supplies the strongest motivation.[^bostrom2019] He asks the reader to imagine technological discovery as drawing balls from an urn: most are white and beneficial, some grey and mixed, and the question is whether the urn contains a black ball — a technology that by default destroys the civilisation that discovers it, because it is cheap, widely accessible, and destructive without requiring an organised state to deploy.

Nuclear weapons turned out grey rather than black, partly because enrichment is difficult and capital-intensive. Bostrom's point is that this was not guaranteed by physics, and that we have no reason to expect the property to hold for future draws. He then examines what would be required to survive a black ball and reaches an uncomfortable conclusion: the two mechanisms that would reliably work — preventive policing with near-universal surveillance, and effective global governance — carry serious risks of their own, including permanent authoritarian entrenchment. Differential development is attractive largely because it is the least invasive of the available strategies.

> [!debate] The core disagreement
> Proponents treat arrival order as a policy variable that funding, regulation and norms can influence. Critics hold that technological development is too coupled for the ordering to be steered: the tools that enable defence usually enable offence, and delaying a capability in one jurisdiction relocates it rather than postponing it.

## What it looks like in practice

Biosecurity is where the principle has come closest to implementation, because the paired capabilities are unusually separable. Detection, sequencing-based surveillance, broad-spectrum medical countermeasures, improved personal protective equipment and germicidal ultraviolet light are defensive and do not directly enable harm. The clearest instance is screening of synthetic nucleic acid orders: an industry consortium built voluntary screening protocols, and US federal policy from 2023 onward directed a formal screening framework tied to federal funding. The screening capability is being built ahead of, rather than in response to, widespread misuse.

The same logic governs the debate over enhanced-potential-pathogen research covered in [[dual-use-research]], and the 2024 argument for halting work toward [[mirror-life]] before the synthetic capability matures — a case where proponents of a halt argue explicitly that no defensive technology exists or is in prospect.

In artificial intelligence, the corresponding proposals are to advance evaluation, interpretability and control research relative to capability research; to use compute as a governance lever because it is measurable, excludable and concentrated in a way that algorithms are not; and to build monitoring infrastructure before rather than after autonomous deployment. Whether any of this has changed relative rates is unclear, and the honest assessment as of 2026 is that capability research has been far better funded than safety research throughout the period in which the argument has been made.

In genome editing, the deliberate sequencing is visible in the treatment of somatic versus heritable applications: [[somatic-gene-therapy]] proceeds through ordinary regulatory channels while [[germline-editing]] is held under a broad moratorium, a division examined in [[governance-of-genome-editing]]. [[gene-drive]] governance follows the same pattern, with phased-testing frameworks and reversal or immunising drives developed alongside the drives themselves.

## Objections

**Inseparability.** The principle requires that offensive and defensive capabilities be distinguishable in advance. Frequently they are not. Sequencing enables both surveillance and design. The models covered under [[ai-protein-design]] assist both vaccine development and toxin engineering. The same argument applies to [[synthetic-genomes]] and [[recoded-organisms]], where biocontainment and novel-organism construction are the same research programme.

**The offence–defence balance may not favour defence.** Analytical work on how the balance scales suggests that in some domains defensive advantage erodes as destructive capability per unit cost rises, which would mean accelerating defence is not sufficient even where it is possible.[^garfinkel2019]

**Coordination.** Delaying a capability requires that all serious developers delay. The *unilateralist's curse* describes the structure precisely: when many actors can each independently take an action, and each estimates its value with error, the action occurs whenever any one of them overestimates — so the threshold for release is set by the most optimistic actor rather than the median.[^bostrom2016] This is a strong argument that delay is unstable absent enforcement.

**Prediction.** Sequencing requires knowing which technologies are dangerous. The record of anticipating this is poor, and the forecasting problems described in [[accelerating-change]] apply directly. Resources spent hardening against the wrong hazard are not neutral; they are spent.

**Enforcement.** Every mechanism strong enough to slow a capability against commercial and military incentive requires state power, and Bostrom's own analysis concedes that the required apparatus is itself a risk.

## Relation to precaution and acceleration

Differential development sits between the two positions it is often confused with. Unlike the [[precautionary-principle]], it does not presume against action under uncertainty; it accepts that the technology will arrive and asks what should exist first. Unlike [[effective-accelerationism]], it denies that faster is better in general, holding that the composition of progress matters more than its rate.

Vitalik Buterin's *d/acc* framing is the most-adopted contemporary restatement, defining the goal as accelerating defensive, decentralising and democratic technologies specifically. Its practical contribution is a heuristic for classification: ask whether a capability advantages the party trying to prevent harm or the party trying to cause it, and fund accordingly. Critics note that many technologies, including most of those covered in [[artificial-general-intelligence]] research, do not sort cleanly on that test.

## Outlook

The principle is the most widely endorsed and least implemented idea in technology governance. It has no institutional home: no agency has a mandate to accelerate one field relative to another, funding decisions are made domain by domain, and the international mechanisms that could coordinate ordering are the same ones that have failed to coordinate on simpler matters. The unresolved question is whether ordering can be influenced at all by anything short of the enforcement capacity that the argument was designed to avoid needing — and if it cannot, whether the principle is a policy or a description of what luck would have to look like.

## See also

- [[existential-risk]]
- [[precautionary-principle]]
- [[nick-bostrom]]
- [[dual-use-research]]
- [[mirror-life]]
- [[effective-accelerationism]]
- [[technological-singularity]]
- [[future-of-humanity]]

## References

[^bostrom2002]: `paper` Bostrom, N. "Existential Risks: Analyzing Human Extinction Scenarios and Related Hazards." *Journal of Evolution and Technology*, 2002. {The paper that names the strategy states it in a short passage within a risk taxonomy; the elaboration all came later.}
[^bostrom2014]: `book` Bostrom, N. *Superintelligence: Paths, Dangers, Strategies*. Oxford University Press, 2014.
[^bostrom2019]: `paper` Bostrom, N. "The Vulnerable World Hypothesis." *Global Policy*, 2019.
[^garfinkel2019]: `paper` Garfinkel, B. and Dafoe, A. "How Does the Offense-Defense Balance Scale?" *Journal of Strategic Studies*, 2019. {An analytical treatment of how the balance behaves as investment scales; it measures no particular technology.}
[^bostrom2016]: `paper` Bostrom, N., Douglas, T. and Sandberg, A. "The Unilateralist's Curse and the Case for a Principle of Conformity." *Social Epistemology*, 2016. {A formal argument about decision structure; it shows that the release threshold is set by the most optimistic actor, not that this has occurred in any recorded case.}

==============================================================================
ARTICLE: digital-immortality
TITLE:   Digital immortality
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/digital-immortality
SOURCE:  https://futurehumanwiki.com/raw/digital-immortality
==============================================================================

---
title: "Digital immortality"
slug: "digital-immortality"
type: "concept"
status: "emerging"
horizon: "present"
categories: ["minds", "society"]
tags: ["griefbots", "digital afterlife", "uploading", "identity", "data ethics", "postmortem privacy"]
summary: "The persistence of a person in digital form after death, ranging from chatbot reconstructions built from their data to the hypothetical emulation of their brain."
updated: "2026-07-27"
issues: ["The GDPR and Tennessee likeness-law claims carry no source."]
---

```infobox
{
  "caption": "Concept and commercial sector",
  "rows": [
    { "label": "Weak form", "value": "Data-trained reconstructions" },
    { "label": "Strong form", "value": "Whole brain emulation" },
    { "label": "First widely reported griefbot", "value": "2016" },
    { "label": "Sector name", "value": "Digital afterlife industry" },
    { "label": "Regulation", "value": "Minimal; postmortem data largely unprotected" },
    { "label": "Continuity of experience", "value": "None in the weak form" }
  ]
}
```

**Digital immortality** is the persistence of a person in computational form after biological death.
The term covers two things that share almost nothing except a marketing vocabulary. The weak form
builds a conversational model from what a person left behind — messages, recordings, video, social
media — and produces something that talks like them. The strong form is [[mind-uploading]]: running
an emulation derived from the structure of the person's brain, which would on some accounts be the
person. The weak form exists and is sold commercially. The strong form has no working demonstration
at any scale.

## The weak form

The idea entered wide circulation in 2016, when a chatbot built from a dead friend's text messages
was reported at length and read by many as a glimpse of something larger.[^newton2016] A
reconstruction is a stylistic model. Given enough text and audio, a language model can be
conditioned or fine-tuned to reproduce a person's turns of phrase, opinions, and voice; add
synthesized speech and a generated face and the output is a video call with a dead relative. Nothing
about the underlying brain is involved, and the model has access only to what was recorded. Where
the source data is thin, the system fills gaps by generalizing from its training corpus — which
means it will confidently produce statements the person never made and could not have made.

The lineage runs back through lifelogging, the archival cousin of the self-tracking practices
described under [[quantified-self]]. Gordon Bell's MyLifeBits project at Microsoft Research
digitized a life's worth of documents and recordings in the 2000s on the premise that comprehensive
capture would preserve something.[^bell2009] Martine Rothblatt's Terasem Movement, an offshoot of
organized [[transhumanism]], promoted "mindfiles" — curated personal archives — as a substrate for
later revival, and commissioned the Bina48 robot as a demonstration. [[ray-kurzweil]] has described
building a conversational replica of his late father from letters and documents, and treats it as a
partial reconstruction rather than a restoration. None of this touches the biological storage of
memory that a [[memory-prosthesis]] attempts to interface with; the archive is external throughout.

```timeline
[
  { "year": "2016", "title": "The Roman bot", "text": "Eugenia Kuyda trains a chatbot on text messages from her friend Roman Mazurenko after his death. The Verge reports it at length, and her company later launches the companion app Replika." },
  { "year": "2017", "title": "Dadbot", "text": "Journalist James Vlahos builds a conversational agent from an oral history recorded with his dying father, later founding a company offering the service commercially." },
  { "year": "2021", "title": "Project December", "text": "A man's use of a GPT-3-based service to simulate his deceased fiancée is reported in detail, prompting debate about consent and the emotional effects of such systems." },
  { "year": "2021", "title": "A patent draws attention", "text": "Reporting on a Microsoft patent covering conversational agents modelled on a specific person from their social data prompts public debate; the company says it has no plans to build the product." },
  { "year": "2024", "title": "Ethics catches up", "text": "Researchers publish design frameworks for 'deadbots', proposing consent requirements, dignity constraints, and exit procedures for the digital afterlife industry." }
]
```

## The digital afterlife industry

Products fall into three groups. *Pre-recorded interactive testimony* records answers to a large
question set while a person is alive and retrieves the closest match on demand; the USC Shoah
Foundation's Dimensions in Testimony project, which lets visitors question Holocaust survivors, works
this way and generates nothing — every answer is something the person actually said. *Curated
generative* systems combine an interview archive with a language model, which is what most
commercial services offer. *Uncurated generative* systems train on whatever data relatives supply,
with no involvement from the person being modelled and no consent at all.

The economics matter. A reconstruction requires ongoing hosting, so the service is a subscription
whose lapse ends a relationship; and because the models sit on the most intimate archive a family
possesses, the sector inherits every unresolved problem in consumer data governance. Carl Öhman and
Luciano Floridi have argued that the industry's structure, in which the dead become an asset class
managed for profit, should be evaluated the way archaeology and funerary practice are — with duties
to the deceased that are not reducible to the interests of the living.[^ohman2017]

> [!caution] Not a continuation
> A reconstruction has no memories, only records of memories; no experience, only generated text.
> Whatever one thinks of [[substrate-independence]], nothing in these systems has the causal
> relationship to the original brain that even the most permissive theory of
> [[personal-identity-and-continuity]] requires. Calling them immortality is a category error, not
> an overstatement of degree.

## Consent, dignity, and postmortem data

The people being modelled are mostly not in a position to object. Most jurisdictions give the dead
weak or no data-protection rights: the European GDPR does not apply to deceased persons, leaving the
matter to member states, and in the United States protection depends on state right-of-publicity
law, which was written for celebrity endorsement rather than for grief. Some recent statutes aimed
at synthetic media, including Tennessee's 2024 law on voice and likeness, extend protection in ways
that may reach griefbots incidentally.

Ethicists working on the problem have proposed concrete constraints rather than prohibition: obtain
consent from the person during life for the specific use; require that the system disclose its
nature; provide a dignified way to retire it; and prohibit its use to advertise to bereaved
users.[^hollanek2024] The last is not hypothetical — a system with a grieving user's attention is
commercially valuable in a way that invites exactly the wrong incentives, a concern that overlaps
with [[mental-privacy]] and with the broader question of who controls intimate inference.

There is also a scale argument. Modelling of social platform data has projected that profiles of
deceased users will number in the billions within this century, making the accumulated digital
remains of the dead a substantial and largely ungoverned archive.[^ohman2019]

## What the evidence says about grief

Little, so far. Grief research has moved away from the older assumption that healthy mourning
requires detachment; continuing-bonds theory holds that maintaining a relationship with the deceased
is normal and often adaptive.[^klass1996] That gives a prima facie case that interactive
reconstructions could help some people. Against it sits the clinical category of prolonged grief
disorder, formally added to psychiatric diagnostic manuals in the 2020s, and the concern that a
system which never refuses contact could sustain avoidance of the loss.

As of 2026 there is no substantial trial evidence either way — no randomized comparison of
griefbot use against standard bereavement support, and no longitudinal data on outcomes. Published
work is dominated by case reports, qualitative interviews, and ethical analysis. Claims in either
direction outrun what has been measured.

## The strong form

Actual digital immortality in the sense advocates intend requires [[whole-brain-emulation]]: a scan
of the specific brain at a resolution sufficient to reconstruct its causal organization. Nothing on
that scale exists. As of 2026 the destructive methods of [[connectomics]] have reconstructed human
cortex in volumes of about a cubic millimetre, roughly a millionth of a brain, and no nervous system
of any size has been emulated. Because the technology does not exist, the practical form of the
strong claim is preservation — [[brain-preservation]] and [[cryonics]] — on the argument that
keeping the information available is a precondition for any later attempt.

Even granting the engineering, the strong form inherits the unresolved questions: whether an
emulation would be conscious, addressed under [[machine-consciousness]] and
[[neural-correlates-of-consciousness]]; and whether it would be the original person, addressed by
the [[teleportation-problem]]. A reconstruction avoids all of these by not making the claim.

## Why the conflation matters

Marketing a reconstruction with the vocabulary of continuation misleads in a specific way: it
suggests that recording more data now brings a person closer to surviving later. It does not. No
quantity of text captures synaptic structure, and the two forms are not points on a single scale but
different kinds of thing. The confusion also runs the other way, letting critics dismiss serious
work on emulation by pointing at chatbots.

The open question in the near term is not technical but institutional. Nobody has established who
owns a dead person's likeness once it is a running model, who may switch it off, what a person can
bind their heirs to in advance, or whether a right to be forgotten survives death — and services are
being sold while the answers are absent.

## See also

- [[mind-uploading]]
- [[whole-brain-emulation]]
- [[personal-identity-and-continuity]]
- [[brain-preservation]]
- [[cryonics]]
- [[machine-consciousness]]
- [[mental-privacy]]
- [[substrate-independence]]

## References

[^ohman2017]: `paper` Öhman, C. and Floridi, L. "The Political Economy of Death in the Age of Information: A Critical Approach to the Digital Afterlife Industry." *Minds and Machines*, 2017.

[^ohman2019]: `paper` Öhman, C. J. and Watson, D. "Are the dead taking over Facebook? A Big Data approach to the future of death online." *Big Data & Society*, 2019. {A projection from one platform's user data under stated growth assumptions, not a measurement of any future number.}

[^hollanek2024]: `paper` Hollanek, T. and Nowaczyk-Basińska, K. "Griefbots, Deadbots, Postmortem Avatars: On Responsible Applications of Generative AI in the Digital Afterlife Industry." *Philosophy & Technology*, 2024. {A normative design framework built on hypothetical scenarios; it reports no data on people who have used these systems.}

[^klass1996]: `book` Klass, D., Silverman, P. R. and Nickman, S. L. (eds.) *Continuing Bonds: New Understandings of Grief*. Taylor & Francis, 1996.

[^newton2016]: `news` Newton, C. "Speak, Memory." *The Verge*, 2016. {A long-form report on one bot, built by a founder who went on to sell a companion app; it describes the system rather than evaluating it.}

[^bell2009]: `book` Bell, G. and Gemmell, J. *Total Recall: How the E-Memory Revolution Will Change Everything*. Dutton, 2009. {A first-person account by the subject of the lifelogging experiment, written as advocacy for total capture.}

==============================================================================
ARTICLE: disability-rights-and-enhancement
TITLE:   Disability rights and enhancement
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/disability-rights-and-enhancement
SOURCE:  https://futurehumanwiki.com/raw/disability-rights-and-enhancement
==============================================================================

---
title: "Disability rights and enhancement"
slug: "disability-rights-and-enhancement"
type: "concept"
status: "established"
horizon: "present"
categories: ["society", "enhancement"]
tags: ["disability", "bioethics", "neurodiversity", "selection", "accessibility", "deafness"]
summary: "The disability-rights critique of selection and enhancement, centred on the expressivist objection, the social model, and the difference between curing and accommodating."
updated: "2026-07-27"
issues: ["The paused 2021 UK autism genetics study is described without a source.", "The claim about prosthetic-hand abandonment rates carries no citation."]
---

```infobox
{
  "caption": "Debate in bioethics and disability studies",
  "rows": [
    { "label": "Central objection", "value": "Expressivist" },
    { "label": "Framing concept", "value": "Social model of disability" },
    { "label": "Named", "value": "United Kingdom, 1976–1983" },
    { "label": "Key figure", "value": "Adrienne Asch" },
    { "label": "Treaty basis", "value": "UN CRPD, 2006" },
    { "label": "Point of conflict", "value": "Prenatal and embryo selection" }
  ]
}
```

**Disability rights and enhancement** covers the arguments made by disability scholars and activists against biomedical selection and enhancement, and the responses to them. The dispute is not between people who want disabled people to suffer and people who do not. It is about whether the appropriate target of intervention is the body or the environment, about what a decision to select against a trait says to people who have it, and about who is entitled to make that judgment.

## The social model

British disability activists in the 1970s drew a distinction that reorganised the field: between *impairment*, a feature of a body, and *disability*, the disadvantage produced when an environment is built for bodies of a different kind. On this account a wheelchair user is not disabled by paralysis but by stairs. The formulation appeared in a 1976 statement by a UK activist organisation and was named the social model by Michael Oliver in the early 1980s.[^oliver1983]

The model's practical achievement is large. It moved policy from charity and rehabilitation toward access and civil rights, and it underwrites the UN Convention on the Rights of Persons with Disabilities, adopted in 2006, which frames disability as arising from interaction between impairment and barriers rather than as a property of individuals.

Its strong form — that impairment contributes nothing to disadvantage — is now defended by few scholars, including within disability studies. Chronic pain and progressive conditions are not straightforwardly environmental. The useful version is comparative: for any given impairment, ask how much of the disadvantage is fixed by biology and how much by design, and note that the second share is usually larger than clinicians assume.

## The expressivist objection

The objection most directly relevant to this wiki concerns selection. Adrienne Asch argued that prenatal testing followed by termination, and by extension [[embryo-selection]], sends a message: that a life with this trait is not worth living, and by implication that existing people with the trait should not exist.[^parens1999]

Her sharper version is about information rather than message. A prenatal test reports one fact — the presence of a variant — and a decision is made on that fact alone about a whole future life. Asch called this a part-whole error: a single trait is treated as determining the value of a life that will contain schooling, work, relationships, and everything else. Parents deciding on this basis are not weighing a life; they are weighing a diagnosis.

Three replies are standard.

The first distinguishes trait from person. Preferring that a child not have a painful condition is compatible with full respect for people who have it, exactly as preferring that a child not be injured is compatible with respecting injured people.

The second appeals to the non-identity structure examined in [[procreative-beneficence]]: selection does not remove anyone, since the alternative to the selected embryo is not a worse life but a different person.

The third, developed at length in *From Chance to Choice*, concedes the sociological point while denying the moral one — that selection may in fact reduce social support for disabled people, and that this is a reason to build the support rather than to prohibit the choice.[^buchanan2000]

> [!key] What the objection is not
> It is not a claim that impairments are good, and its leading exponents supported abortion rights. The claim is about the informational and expressive structure of a routinised screening programme, not about individual reproductive liberty.

## Deaf and autistic positions

Two communities have developed positions that treat their condition as an identity rather than a deficit, and both have become reference cases.

Deaf opposition to paediatric cochlear implantation, articulated in a position paper by the US National Association of the Deaf in 1991 and revised subsequently, rests on the claim that Deaf people constitute a linguistic and cultural minority with a full natural language, and that implanting a child amounts to a decision to raise them outside that community. The technical picture has shifted since: outcomes improve markedly with earlier implantation, which forces the decision at an age when the child cannot participate in it, and current position statements are considerably less oppositional than those of the 1990s while still emphasising access to sign language. [[cochlear-implant]] covers the technology and the outcome data.

The inverse case attracted more attention than its size warranted: in 2002 a Deaf American couple deliberately sought a deaf sperm donor, an episode used in the enhancement literature ever since to test whether selecting *for* a trait is different in kind from selecting against it. Most commentators concluded it is not, which is uncomfortable for both camps.

Autistic advocacy, organised around the concept of neurodiversity that emerged in the late 1990s, has opposed cure-oriented research framing more successfully than any other disability movement. A large autism genetics study launched in the UK in 2021 was paused within weeks after sustained objections from autistic people about consent, data use, and the possibility that the findings would support prenatal screening. The episode demonstrated that community objection can halt a well-funded study, which nothing in the ethics literature had predicted.

## Cure and accommodation

The two responses to impairment are not exclusive, and the argument between them is mostly about allocation and default.

Accommodation has a strong record. Curb cuts, captioning, and screen readers deliver benefits at low cost, immediately, to everyone with the relevant need, and often to people without it. Cure has produced genuine results for some conditions and none for most; the technologies in [[retinal-implant]] restore very limited function after decades of work, while [[limb-regeneration]] remains far from human application.

The allocation objection is that enhancement and cure research absorbs resources and attention that accessibility would use better, and that it is chosen by funders partly because it promises a technical fix rather than a social one. Ashley Shew's term for the resulting pattern — technoableism, the presentation of technology as making disabled people acceptable rather than making society accessible — has been widely adopted in disability studies.[^shew2023]

There is a countervailing observation. Disabled people are the first and often the only users of most of the technologies this wiki covers. [[brain-computer-interface]] research is conducted almost entirely with participants who have paralysis or ALS; [[myoelectric-prosthetics]], [[exoskeleton]] systems, and [[deep-brain-stimulation]] were all developed for clinical populations. Treating enhancement research as opposed to disabled people's interests misdescribes a field in which they are the primary beneficiaries and, in the case of implanted-device trials, bear most of the risk. High abandonment rates for advanced prosthetic hands are the field's standing evidence that devices designed without user priorities fail regardless of their capability.

## The moving baseline

Enhancement raises a distinct problem that selection does not. Disability is defined relative to a norm, and enhancement moves the norm.

If pharmacological or genetic augmentation of some capacity became widespread, people at today's median would be below the new median, and environments would be built for the enhanced. The mechanism is familiar from literacy: in a society where nearly everyone reads, being unable to read is disabling in a way it was not when nobody could. Whether this counts as an argument against enhancement or merely as a prediction about what accommodation will be needed depends on whether one thinks the enhanced society is also richer in the resources to accommodate — which is the same disagreement, unresolved, that runs through [[enhancement-arms-race]] and [[access-and-inequality]].

> [!debate] Whose judgment counts
> Non-disabled people consistently rate life with a given impairment as worse than disabled people rate their own lives, a discrepancy documented in quality-of-life research and known as the disability paradox.[^albrecht1999] Since selection decisions are made by prospective parents who are usually not disabled, and health-economic weightings are derived largely from general-population surveys, the ratings driving both are the ones that adaptation shows to be mistaken.

## Open problems

The severity gradient is the question neither side has answered. Almost nobody defends selection against traits with minimal effect on wellbeing, and almost nobody objects to preventing conditions causing early death with constant pain. Between those poles there is no principle, only case-by-case judgment made by parents with a few days to decide and a clinician's summary to decide from.

Nor is it settled whether the expressivist objection survives the shift from single-gene testing to [[polygenic-embryo-screening]], which produces probabilities across many traits at once rather than a diagnosis. A parent selecting on an aggregate score is not making a judgment about any particular condition, which arguably defuses the message — or generalises it, since the profile being selected against is now a whole distribution of human variation rather than one identified disease.

## See also

- [[cochlear-implant]]
- [[procreative-beneficence]]
- [[bioethics-of-enhancement]]
- [[embryo-selection]]
- [[polygenic-embryo-screening]]
- [[bioconservatism]]
- [[enhancement-in-sport]]
- [[morphological-freedom]]

## References

[^oliver1983]: `book` Oliver, M. *Social Work with Disabled People*. Macmillan, 1983. {The book that named the social model; the impairment-disability distinction itself comes from the 1976 activist statement it builds on.}
[^parens1999]: `book` Parens, E. and Asch, A. (eds.) *Prenatal Testing and Disability Rights*. Georgetown University Press, 1999. {An edited collection: Asch develops the expressivist argument, and other contributors to the same volume argue against it.}
[^buchanan2000]: `book` Buchanan, A., Brock, D.W., Daniels, N., Wikler, D. *From Chance to Choice: Genetics and Justice*. Cambridge University Press, 2000.
[^shew2023]: `book` Shew, A. *Against Technoableism: Rethinking Who Needs Improvement*. W. W. Norton, 2023.
[^albrecht1999]: `paper` Albrecht, G.L. and Devlieger, P.J. "The disability paradox: high quality of life against all odds." *Social Science & Medicine*, 1999. {An interview study of people with serious persistent disabilities; the comparison with how non-disabled raters score the same states comes from other work.}

==============================================================================
ARTICLE: dna-nanotechnology
TITLE:   DNA nanotechnology
PORTAL:  Nanomedicine
URL:     https://futurehumanwiki.com/wiki/dna-nanotechnology
SOURCE:  https://futurehumanwiki.com/raw/dna-nanotechnology
==============================================================================

---
title: "DNA nanotechnology"
slug: "dna-nanotechnology"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 4
categories: ["nanomedicine", "genetics"]
tags: ["nanomedicine", "dna origami", "self-assembly", "molecular machines", "drug delivery", "biosensing"]
summary: "The use of DNA as a construction material rather than a carrier of genes, folding programmed sequences into nanoscale shapes, devices, and logic-gated containers."
updated: "2026-07-27"
humanEvidence: "No DNA nanostructure has entered a human trial; the strongest results are logic-gated payload release on cultured human cells and thrombin-loaded tubes that slowed tumour growth in mice."
access: "Research use only: scaffolds and staple strands are ordinary laboratory reagents, and no DNA nanostructure is available as a medical product anywhere."
reversibility: "reversible"
issues: ["The gram-scale biological production of scaffold and staples needs a source."]
---

```infobox
{
  "caption": "Molecular fabrication technique",
  "rows": [
    { "label": "Founded by", "value": "Nadrian Seeman, 1982" },
    { "label": "Key method", "value": "DNA origami (Rothemund, 2006)" },
    { "label": "Feature resolution", "value": "About 3–6 nanometres" },
    { "label": "Typical scaffold", "value": "M13 phage genome, ~7,200 bases" },
    { "label": "Assembly", "value": "Thermal annealing in one pot" },
    { "label": "Medical use", "value": "Animal studies only" },
    { "label": "Main limitation", "value": "Stability and immunogenicity in blood" },
    { "label": "Readiness", "value": "TRL 4" }
  ]
}
```

**DNA nanotechnology** treats DNA as a building material rather than as genetic information. Because base pairing is predictable, a set of synthetic strands can be designed so that the only low-energy configuration they can adopt is a particular shape, and that shape then assembles itself when the strands are mixed and cooled. The field has produced nanoscale objects of arbitrary geometry, molecular machines that walk and compute, and containers that open in response to a chosen biological signal.

## How it works

The founding insight is that the double helix is not the only structure DNA can form. Branched junctions occur naturally during recombination, but they migrate. Nadrian Seeman proposed in 1982 that by designing sequences with no symmetry, a junction could be made immobile, and immobile junctions could be linked into a rigid lattice.[^seeman1982] That converts DNA from a linear molecule into a construction kit with a programmable connectivity.

DNA origami, introduced by Paul Rothemund in 2006, made the approach practical.[^rothemund2006] A single long scaffold strand — usually the genome of the M13 bacteriophage, around 7,200 bases — is folded into a target shape by roughly two hundred short synthetic "staple" strands, each designed to bind two or three distant regions of the scaffold and hold them together. Mixing scaffold and staples and annealing over a few hours yields the designed object in enormous copy number, with folding yields high enough that misassembled structures are usually the minor product; the excess staples are then removed by filtration or gel extraction. Extending the method into three dimensions produced honeycomb and square lattices, then DNA "bricks" that build shapes without a scaffold at all, then assemblies in the gigadalton range.

The resolution is set by the helix: features are addressable at roughly 3 to 6 nanometres, and every position on the structure has a known sequence, so a chemical group, a fluorophore, a protein, or a nanoparticle can be attached at a chosen coordinate. That addressability is the field's real advantage over the self-assembling but unaddressable vesicles of [[lipid-nanoparticles]], and it is what distinguishes it from the mechanical construction proposed in [[molecular-assembler]].

## Development history

The field moved in two phases. Two decades of structural work established that designed sequences could be made to hold a shape at all, mostly on lattices and polyhedra assembled from strands of similar length. The second phase began when folding a long natural scaffold with cheap synthetic staples turned that into a one-pot procedure any molecular biology laboratory could run, after which the interesting question stopped being whether a shape could be built and became what to attach to it.

```timeline
[
  { "year": "1982", "title": "Immobile junctions proposed", "text": "Nadrian Seeman argues that sequence-asymmetric branched DNA could form rigid lattices, founding the field on paper." },
  { "year": "1991", "title": "First DNA polyhedron", "text": "Chen and Seeman assemble a molecule with the connectivity of a cube from synthetic strands." },
  { "year": "1994", "title": "DNA computing", "text": "Leonard Adleman solves a small Hamiltonian path problem using DNA hybridisation, showing that sequence design can encode computation." },
  { "year": "1998", "title": "Algorithmic self-assembly", "text": "Winfree and colleagues build two-dimensional crystals from DNA tiles, linking self-assembly to computation." },
  { "year": "2006", "title": "DNA origami", "text": "Rothemund folds a viral scaffold into arbitrary two-dimensional shapes with short staple strands, making complex structures routine." },
  { "year": "2009", "title": "Three-dimensional origami", "text": "Douglas and colleagues extend the method to solid 3D shapes built on a honeycomb lattice." },
  { "year": "2012", "title": "Logic-gated container", "text": "A barrel-shaped origami device latched by aptamers opens only on cells displaying specified antigens, delivering payloads in cell culture." },
  { "year": "2018", "title": "Therapeutic effect in mice", "text": "Thrombin-loaded origami tubes triggered by tumour vasculature markers induce clotting in tumour vessels and slow growth in mouse models." }
]
```

## Devices and logic

Static shapes were the first product; dynamic ones followed. None of them swims: motion in DNA devices means a conformational change or a walker stepping along a track, not locomotion through tissue, which remains the province of the externally steered machines in [[microrobots-in-medicine]]. Toehold-mediated strand displacement, in which an incoming strand invades a partially paired duplex from a single-stranded overhang, gives a reliable primitive for state change. Chained together, displacement reactions implement logic gates and multi-layer circuits operating entirely in solution, with no enzymes and no external control. Bipedal DNA walkers move along tracks by the same chemistry.

The medical application of this machinery is conditional release. In 2012 a Harvard group built an origami barrel held shut by two aptamer locks, each of which opens only on binding a specified cell-surface protein; the device carried antibody fragments and released them selectively onto target leukaemia cells in culture.[^douglas2012] The logic was genuinely AND-gated: both antigens had to be present.

The strongest in vivo result to date used a simpler trigger. Origami tubes were loaded with thrombin and closed with an aptamer that binds nucleolin, a protein displayed on tumour endothelium but not on normal vessels. Injected into tumour-bearing mice, the devices opened at the tumour vasculature, released their cargo, and induced local thrombosis, slowing tumour growth in several models.[^li2018] It remains the most cited demonstration that a programmable DNA device can produce a therapeutic effect in an animal.

> [!note] What "nanorobot" means here
> These devices contain no power source, no processor, and no means of locomotion. They circulate passively and change state when a specific molecule binds. Calling them robots is a convention of the field rather than a description of their capabilities, a distinction that matters when comparing them with the designs in [[medical-nanorobots]] — [[eric-drexler]]'s cell-repair machines and Freitas's [[respirocytes|respirocyte]] are specified with onboard sensing, computation, and actuation, none of which an origami device has.

## Limitations in the body

DNA is a poor material for blood. Three problems dominate.

Structures assembled at 10 to 20 millimolar magnesium partially unfold at the roughly 1 millimolar free magnesium of extracellular fluid. Nucleases in serum degrade unprotected origami within hours. And unmethylated CpG motifs in bacterial-derived scaffold sequence are recognised by Toll-like receptor 9, provoking an innate immune response that is useful when the structure is meant to be a vaccine scaffold and unwelcome when it is meant to be a carrier.

Coatings mitigate the first two. Oligolysine conjugated to polyethylene glycol electrostatically wraps the structure, stabilising it at low salt and slowing nuclease attack by an order of magnitude or more.[^ponnuswamy2017] Lipid bilayer coatings and protein shells achieve similar effects. None of these has yet carried a DNA device through a human trial, whereas the two crude carriers it competes with — [[aav-vectors]] and lipid vesicles — have each delivered approved products.

Payload capacity is a further constraint. An origami barrel holds a handful of protein molecules; a liposome of comparable size holds thousands of small-molecule drugs. For most therapeutic purposes the cargo-to-carrier ratio is unfavourable, which is one reason the clinical successes of nanomedicine have gone to formulations rather than to structures, as described in [[targeted-drug-delivery]].

Cost was a fourth constraint until recently. Synthetic staple strands are expensive at scale, and a route that produces both scaffold and staples biologically in bacteriophage-infected bacteria has reduced the price of bulk origami by orders of magnitude, moving gram-scale production from implausible to demonstrated. The same falling cost of writing DNA underpins [[synthetic-genomes]] and has made sequence screening at synthesis providers a standing item in [[dual-use-research]] policy.

## Uses beyond therapy

The field's most consistent value has been as a laboratory tool. Origami acts as a molecular breadboard: a rigid, addressable surface for placing molecules at defined distances. DNA-PAINT exploits transient strand binding to achieve super-resolution imaging below the diffraction limit. Origami rulers calibrate microscopes. Nanopore experiments use DNA structures as gates and adapters, connecting to the sensing methods described in [[nanoparticle-diagnostics]].

Immunology has produced a distinctive application. Because antigen spacing on a surface strongly affects B-cell activation, origami scaffolds displaying antigens at controlled separations have been used to dissect that dependence and to design immunogens — a use of the technology as a precision instrument rather than as a delivery vehicle.[^veneziano2020]

DNA is also a storage medium, a line of work associated with [[george-church]] among others, and one that shares synthesis and sequencing infrastructure with the fabrication side of the field but little else.

## Outlook

The obstacle to clinical translation is not design but pharmacology: an object made of nucleic acid must survive serum, avoid the liver, escape the endosome, and release cargo where intended, and the field has partial answers to each and a complete answer to none. The most likely first clinical entries are those where DNA's drawbacks matter least — vaccine and immunotherapy scaffolds, where immune recognition is the point, and topical or local applications, where circulation is not involved.

Whether the addressability advantage ever justifies the fragility is the open question. A structure that positions two proteins exactly seven nanometres apart can do things no liposome can. As of 2026 nobody has shown that this buys enough in a patient to outweigh a material that dissolves in blood, and the [[crispr-cas9]] era has meanwhile demonstrated how much can be achieved with delivery vehicles that have no internal architecture at all.

## See also

- [[medical-nanorobots]]
- [[targeted-drug-delivery]]
- [[molecular-assembler]]
- [[nanoparticle-diagnostics]]
- [[lipid-nanoparticles]]
- [[microrobots-in-medicine]]
- [[george-church]]
- [[synthetic-genomes]]

## References

[^seeman1982]: `paper` Seeman, N.C. "Nucleic acid junctions and lattices." *Journal of Theoretical Biology*, 1982.
[^rothemund2006]: `paper` Rothemund, P.W.K. "Folding DNA to create nanoscale shapes and patterns." *Nature*, 2006. {The shapes demonstrated here are two-dimensional; folding into solid three-dimensional objects came from other groups later.}
[^douglas2012]: `paper` Douglas, S.M., Bachelet, I., Church, G.M. "A logic-gated nanorobot for targeted transport of molecular payloads." *Science*, 2012. {The AND-gated release was demonstrated on cultured cells in a dish; the mouse work that followed used a simpler single-aptamer trigger.}
[^li2018]: `paper` Li, S. et al. "A DNA nanorobot functions as a cancer therapeutic in response to a molecular trigger in vivo." *Nature Biotechnology*, 2018. {The therapeutic effect is in tumour-bearing mice; nothing comparable has been reported in people.}
[^ponnuswamy2017]: `paper` Ponnuswamy, N. et al. "Oligolysine-based coating protects DNA nanostructures from low-salt denaturation and nuclease degradation." *Nature Communications*, 2017.
[^veneziano2020]: `paper` Veneziano, R. et al. "Role of nanoscale antigen organization on B-cell activation probed using DNA origami." *Nature Nanotechnology*, 2020.

==============================================================================
ARTICLE: dual-use-research
TITLE:   Dual-use research of concern
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/dual-use-research
SOURCE:  https://futurehumanwiki.com/raw/dual-use-research
==============================================================================

---
title: "Dual-use research of concern"
slug: "dual-use-research"
type: "risk"
status: "established"
horizon: "present"
categories: ["society", "genetics"]
tags: ["biosecurity", "governance", "pandemics", "gain of function", "policy", "risk"]
summary: "Life-science research that produces knowledge or tools with legitimate purposes but foreseeable potential for catastrophic misuse, and the oversight regimes built around it."
updated: "2026-07-27"
humanEvidence: "No documented case exists of published dual-use research being used to cause mass harm; the recorded human toll is from laboratory-acquired infections and containment failures."
access: "The enabling capability is widely available: gene synthesis is a commercial service and pathogen sequences are public, with order screening voluntary and incomplete."
reversibility: "irreversible"
issues: ["The 2025 report of AI-designed toxin variants evading synthesis screening has no citation."]
---

```infobox
{
  "caption": "Category of biosecurity risk",
  "rows": [
    { "label": "Term formalised", "value": "United States, mid-2000s" },
    { "label": "Origin document", "value": "Fink report, 2004" },
    { "label": "Defining episode", "value": "H5N1 transmissibility papers, 2011" },
    { "label": "Current US policy", "value": "Unified DURC/PEPP policy, 2024" },
    { "label": "Treaty basis", "value": "Biological Weapons Convention, 1972" },
    { "label": "Verification mechanism", "value": "None" }
  ]
}
```

**Dual-use research of concern**, or DURC, is life-science research conducted for legitimate purposes that could reasonably be anticipated to yield knowledge, methods, or materials directly applicable to causing large-scale harm. The category was formalised in US policy in the 2000s, following a National Research Council report that identified a short list of experiments of concern,[^fink2004] and covers a narrow slice of biology — work that increases a pathogen's transmissibility, virulence, host range, or resistance to countermeasures, or that renders a dangerous agent easier to produce. It is the clearest case in which the same result is both a public good and a hazard, and it has no solution, only trade-offs.

## The mechanism

Biological risk differs from other technological risk in three ways that shape the governance problem.

**Information is the weapon.** A method published in a journal cannot be recalled, is available identically to everyone who reads it, and does not degrade. The barrier to misuse is tacit skill and materials rather than knowledge, and both barriers have fallen.

**Capability is dual by construction.** Understanding what makes an influenza virus transmissible between mammals is the same knowledge that would allow one to be made transmissible. Surveillance, vaccine design, and countermeasure development all require it. There is no version of the research that produces only the defensive half.

**Self-replication.** Unlike chemical or radiological agents, a pathogen released once can propagate beyond the point of release. This is why biological risk features in the taxonomies discussed under [[existential-risk]] and why engineered pandemics are usually named as the most plausible near-term route to a global catastrophe. The same property drives concern about the self-spreading constructs described in [[gene-drive]] and about the chirally inverted organisms in [[mirror-life]], and it is what distinguished real biological hazard from the self-replicating machines imagined in [[grey-goo]], which never existed.

[[nick-bostrom|Nick Bostrom]]'s typology of information hazards supplies the vocabulary the field now uses: the harm can come from the specific data, from the demonstration that something is possible, or merely from directing attention to a vulnerability.[^bostrom2011]

## The cases that defined the category

```timeline
[
  { "year": "2001", "title": "Mousepox", "text": "Australian researchers inserting an immune-signalling gene into mousepox found it killed vaccinated mice. The result was incidental to the project's contraceptive aim and demonstrated that virulence enhancement could be stumbled into." },
  { "year": "2002", "title": "Poliovirus synthesised", "text": "A team assembled infectious poliovirus from mail-order DNA fragments using the published sequence, showing that a virus could be reconstructed without access to a natural sample." },
  { "year": "2005", "title": "1918 influenza reconstructed", "text": "The virus responsible for the deadliest pandemic on record was rebuilt from sequence recovered from preserved tissue and characterised in animals." },
  { "year": "2011–2012", "title": "H5N1 transmissibility", "text": "Two laboratories produced avian influenza variants transmissible between ferrets by respiratory droplet. A US advisory board first recommended redacting the methods, then reversed; the papers were published in full after a voluntary research moratorium." },
  { "year": "2018", "title": "Horsepox", "text": "A Canadian group synthesised horsepox virus, a relative of smallpox, from commercially ordered DNA at modest cost, reopening the question of whether smallpox could be reconstructed." }
]
```

The 2011 episode is the field's reference case because it exposed every unresolved question at once: whether a journal should publish methods, who decides, whether redaction is feasible when dozens of laboratories can reproduce the work, and whether a voluntary moratorium accomplishes anything beyond delay. The papers were published, no misuse followed, and both camps regard this as confirming their position.

## Plausibility

Assessing how likely deliberate misuse is requires separating three questions that are routinely merged.

**Capability** has clearly increased. Gene synthesis is a commercial service; sequences for most known pathogens are in public databases; the reverse-genetics methods for reconstructing RNA viruses are taught in graduate programmes, and the editing tools described in [[crispr-cas9]] made targeted modification of a genome a routine laboratory operation. The whole-genome construction methods surveyed in [[synthetic-genomes]] were developed for microbial engineering and apply without modification to viral genomes. The 2018 horsepox work was performed by a small academic team at a reported cost of about a hundred thousand dollars.

**Intent** is the term with the least evidence behind it. Historical bioweapons programmes were state-run, expensive, and largely unsuccessful at producing usable weapons; the small number of non-state attempts caused limited casualties. There is no observed case of a modern actor using published dual-use research to cause mass harm. Analysts disagree sharply about whether this reflects a genuine difficulty or a period of good fortune.

**Accident** is the neglected term and the one with a documented record. Laboratory-acquired infections and containment failures have occurred repeatedly in high-containment facilities across multiple countries, including with smallpox and with influenza strains. Any assessment of research risk that considers only malice omits the more frequent pathway.

> [!caution] What the debate cannot resolve
> The costs of restricting research are diffuse and counterfactual — vaccines not developed, variants not anticipated — while the benefits are also counterfactual. Neither side can produce the comparison that would settle the question, which is why two decades of argument have not converged.

## AI-assisted biodesign

The newest input is machine learning, and it has changed the analysis in a specific way rather than a general one.

A 2022 paper reported that a [[ai-drug-discovery|drug-discovery model]] built to avoid toxicity could be inverted to seek it, generating tens of thousands of candidate toxic molecules in under six hours.[^urbina2022] The molecules were never synthesised, and synthesis remains the hard step, but the demonstration reframed the bottleneck: design is no longer scarce.

[[ai-protein-design|Protein design models]] raise a sharper version. In 2025 researchers reported that AI-generated variants of known toxin sequences could evade the sequence-screening software used by commercial DNA synthesis providers, with patches developed and distributed before publication. The episode is notable less for the vulnerability than for the response, which followed a coordinated-disclosure model borrowed from computer security and is the first such process in biosecurity.

Frontier AI developers began applying elevated deployment safeguards to their models on biological-weapons-relevant capability grounds from 2025, and biosecurity evaluations are now a standard component of frontier model assessment. Whether the safeguards constrain a determined user or merely raise the effort required is not publicly established, and the same models are available in open-weight form from other developers without comparable restrictions. [[artificial-general-intelligence]] treats the broader capability question.

## Mitigation

Controls operate at four points, and their effectiveness declines as one moves down the list.

**Physical containment** — biosafety levels, negative pressure, personnel training — is mature, well specified, and addresses accidents rather than misuse. Genetic containment is the newer variant: the synthetic auxotrophy built into the strains described in [[recoded-organisms]] makes an escaped organism unable to survive outside a supplemented medium, which is a stronger guarantee than a physical barrier and applies only to engineered organisms rather than to natural pathogens.

**Synthesis screening** is the most-discussed intervention. Members of the International Gene Synthesis Consortium, formed in 2009, screen order sequences against lists of agents of concern and verify customer legitimacy. Coverage is voluntary and incomplete, the screening is defeatable by ordering fragments from multiple providers, and benchtop synthesisers distribute the capability to sites that no consortium reaches. The US required screening for federally funded work under a 2023 executive order that was revoked in January 2025, leaving the status of the associated framework unclear as of 2026.

**Funding-stage review** allows a funder to decline work before it is done, which is the only intervention that acts before the information exists. The US paused federal funding for certain gain-of-function influenza and coronavirus work from 2014 to 2017, replaced the pause with a review framework, and in 2024 issued a unified policy covering both DURC and research on pathogens with enhanced pandemic potential. An executive order in May 2025 further restricted federal support for dangerous gain-of-function research. All of this binds federally funded work in one country.

**Publication review** is the last line and the weakest. Journals have no investigative capacity, editors are not biosecurity analysts, and by the time a manuscript is submitted the work has been done and discussed at conferences.

## Governance

The Biological Weapons Convention, in force since 1975, prohibits development, production, and stockpiling of biological weapons and has near-universal membership. It has no verification mechanism: negotiations on a compliance protocol collapsed in 2001 and have not been revived. Its implementation support unit is very small. There is no international body that inspects laboratories, no reporting obligation for dual-use experiments, and no forum with authority to stop a specific project.

What exists instead is national policy of varying rigour, institutional review committees, and the professional norms inherited from the [[asilomar-conference]] — a model whose limits were demonstrated in the [[he-jiankui-affair]] and which assumes the relevant actors care about standing among peers. The same structural weakness runs through [[governance-of-genome-editing]]: the enforceable layer is domestic law, and it reaches only those who submit to it.

The field's characteristic policy proposal is not prohibition but sequencing. [[differential-technological-development]] argues for advancing detection, countermeasures, and protective capability ahead of the research that creates the hazard, which avoids the objection that a blanket [[precautionary-principle]] would also block the work needed to respond to natural outbreaks.

The open problem is what happens as the capability distributes. The controls described above are designed for a world in which dangerous work requires institutional resources: a laboratory, a funder, a supplier. Benchtop synthesis, contract research organisations, and design tools that run on a laptop each erode that assumption. The proposals for what replaces it — from licensing of synthesis equipment to the surveillance architectures contemplated in the vulnerable-world literature — are either insufficient or, as their own authors concede, worse than the problem.

## See also

- [[existential-risk]]
- [[mirror-life]]
- [[gene-drive]]
- [[asilomar-conference]]
- [[synthetic-genomes]]
- [[differential-technological-development]]
- [[precautionary-principle]]
- [[governance-of-genome-editing]]

## References

[^fink2004]: `report` National Research Council. *Biotechnology Research in an Age of Terrorism*. National Academies Press, 2004. {The source of the original list of experiments of concern; it is advice to US policymakers and carried no legal force of its own.}
[^bostrom2011]: `paper` Bostrom, N. "Information Hazards: A Typology of Potential Harms from Knowledge." *Review of Contemporary Philosophy*, 2011.
[^urbina2022]: `paper` Urbina, F., Lentzos, F., Invernizzi, C., Ekins, S. "Dual use of artificial-intelligence-powered drug discovery." *Nature Machine Intelligence*, 2022. {None of the generated molecules were synthesised or assayed, so the paper reports a design capability rather than demonstrated toxicity.}

==============================================================================
ARTICLE: ectogenesis
TITLE:   Ectogenesis
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/ectogenesis
SOURCE:  https://futurehumanwiki.com/raw/ectogenesis
==============================================================================

---
title: "Ectogenesis"
slug: "ectogenesis"
type: "concept"
status: "speculative"
horizon: "2040s"
categories: ["reproduction", "society"]
tags: ["reproduction", "ectogenesis", "bioethics", "feminism", "abortion", "personhood"]
summary: "The development of a mammalian embryo or fetus outside a body, either for part of gestation or for all of it, and the ethical arguments the prospect generates."
updated: "2026-07-27"
humanEvidence: "No part of human gestation has been sustained outside a body; the closest results are weeks of survival in preterm lambs and mouse embryos cultured to roughly half of gestation."
issues: ["The 1960s human fetal perfusion experiments are covered only in artificial-womb and are not summarised here."]
---

```infobox
{
  "caption": "Concept in reproductive biology and bioethics",
  "rows": [
    { "label": "Coined by", "value": "J. B. S. Haldane" },
    { "label": "Year", "value": "1923" },
    { "label": "Partial form", "value": "Experimental in animals", "link": "/wiki/artificial-womb" },
    { "label": "Complete form", "value": "No demonstration" },
    { "label": "Longest mammalian ex utero culture", "value": "About half of mouse gestation" },
    { "label": "Main legal issue", "value": "Status of the gestating entity" }
  ]
}
```

**Ectogenesis** is the development of an embryo or fetus outside a body. The literature distinguishes *partial* ectogenesis, in which a fetus is transferred to a device part-way through gestation, from *complete* ectogenesis, in which fertilisation, implantation, and the whole of gestation occur outside any organism. Partial ectogenesis has been achieved in sheep and is the goal of every device programme now running. Complete ectogenesis in a mammal has never been demonstrated and the missing step is not the equipment but placentation.

## Origins of the term

J. B. S. Haldane coined the word in a lecture given at Cambridge in 1923 and published as *Daedalus; or, Science and the Future*.[^haldane1924] Writing in the voice of a future undergraduate looking back, he imagined that gestation outside the body would become normal within a century and that the change would be greeted first with disgust and then with indifference. His point was less about wombs than about the pattern: every biological invention, he argued, appears blasphemous before it becomes ordinary.

Aldous Huxley, who knew Haldane, took the same image and inverted its valence. The hatcheries of *Brave New World* in 1932 attached ectogenesis to a state that had abolished families, and fixed the association between artificial gestation and totalitarian control that most subsequent fiction has reproduced.[^huxley1932] The two readings have travelled together ever since, and much of the modern argument is a continuation of the disagreement between them.

Haldane's essay is also part of the prehistory of [[transhumanism]]; his contemporary J. D. Bernal and later Julian Huxley worked the same ground, and the mid-century optimism that biology could be deliberately redesigned runs directly into the movement described in [[posthuman]].

## Partial and complete

The two versions are technically unrelated, which is the most common source of confusion.

```compare
{
  "columns": ["Partial ectogenesis", "Complete ectogenesis"],
  "rows": [
    { "label": "Starting point", "values": ["A fetus already gestated in a body", "A fertilised egg or an embryo"] },
    { "label": "Core problem", "values": ["Gas exchange and circulation without a placenta's help", "Building implantation and placentation from nothing"] },
    { "label": "Status", "values": ["Weeks of survival in preterm lambs", "No mammalian demonstration"] },
    { "label": "Nearest clinical relative", "values": ["Neonatal intensive care", "None"] },
    { "label": "Regulatory identity", "values": ["Medical device", "Undefined"] }
  ]
}
```

The engineering of the partial case is covered in [[artificial-womb]]; the longest reported support is
about four weeks, in preterm lambs.[^partridge2017] The obstacle in the complete case is that a placenta is not a filter but an organ built jointly by two genomes, in which fetal trophoblast invades maternal tissue, remodels spiral arteries, and negotiates an immune truce that is still poorly understood. Human embryos are grown routinely to the blastocyst stage in IVF laboratories and, under research licence, to around the fourteen-day limit discussed in [[synthetic-embryos]]. Mouse embryos have been cultured ex utero from before gastrulation into late organogenesis, roughly half of gestation, in a perfused rotating-bottle system.[^hanna2021] Nothing comparable exists for a species with a deeply invasive haemochorial placenta and a nine-month term.

> [!note] Terminology
> "Artificial womb" names a device; "ectogenesis" names the process. A third term, *gestateling*, was proposed by the legal scholar Elizabeth Chloe Romanis for an entity being gestated in a device, on the grounds that it is neither a fetus nor a neonate and that existing categories carry legal consequences it should not automatically inherit.[^romanis2018]

## The feminist arguments, in both directions

Ectogenesis is unusual among reproductive technologies in that feminist theory has produced sustained arguments both for and against it, from the same premises.

Shulamith Firestone made the case for in 1970, arguing that pregnancy is the material basis of women's subordination and that only its abolition, not its reform, would end the asymmetry.[^firestone1970] Anna Smajdor later restated the claim in medical-ethics terms: gestation imposes serious risks and burdens on one class of people and no other, and a technology that removes them is not optional but obligatory once safe.[^smajdor2007] On this reading ectogenesis belongs with contraception rather than with eugenics.

The case against came fastest from Gena Corea, who argued that reproductive technology historically expanded medical authority over women's bodies rather than women's authority over their own.[^corea1985] A device that gestates places fetal welfare under the control of clinicians, institutions, and courts, and creates a benchmark against which ordinary pregnancy can be judged deficient. A related worry concerns coercion: if gestation outside the body is available, refusing it, or continuing a pregnancy while smoking or working a hazardous job, becomes a choice for which a person can be blamed.

Both arguments predate any working device. What has been built is a neonatal life-support system, and its immediate ethical questions are about consent and equipoise rather than about the future of the family.

## Abortion and the shape of the argument

Judith Jarvis Thomson's 1971 defence of abortion grounds the right in bodily autonomy: no one may be compelled to lend their body to sustain another, whatever the other's moral status.[^thomson1971] The argument is deliberately silent on whether a fetus is a person, and this is what makes it vulnerable to ectogenesis. A right not to gestate is not obviously a right to a dead fetus. If a fetus can be removed and sustained elsewhere, the autonomy argument appears to license extraction rather than termination.

Responses split. Some hold that a genetic-parenthood interest is separate from a bodily-autonomy interest, and that no one may be made a parent against their will even if their body is no longer involved. Others argue that the extraction procedure is itself more invasive than an abortion at the same gestational age, so the autonomy argument survives on the facts. A third position accepts the conclusion and treats it as a reason to expect the legal category of viability, which several jurisdictions use as a statutory threshold, to become unstable.

> [!debate] Which way the technology cuts
> Opponents of abortion have described ectogenesis as a resolution that satisfies both sides; abortion-rights scholars generally read the same scenario as a mechanism for compelling transfers. Nothing about the technology settles which reading prevails, because the disagreement was never about the availability of an alternative.

## Personhood, parenthood, and law

A device raises questions the law has not needed to answer. Statutes distinguish embryos from fetuses from born persons, and often use birth or viability as the switch. An entity in a device has been extracted from a body but is not breathing, is not independent, and may be at a gestational age at which termination would be lawful. Rights of inheritance, of citizenship by place of birth, and of protection from harm all depend on categories that assume gestation happens inside someone.

Parenthood is equally unsettled. Where a gestational carrier is involved, most jurisdictions assign maternity by gestation or by court order under surrogacy law; a device is not a legal person and cannot be a parent. Combined with [[in-vitro-gametogenesis]], which would allow gametes to be made from somatic cells, complete ectogenesis would sever every physical link between having a child and any bodily process, leaving parenthood to be determined entirely by intention and contract. Whether that is liberation or commodification is the live disagreement, and it maps closely onto the argument over [[morphological-freedom]].

## Criticism and open problems

The strongest scientific criticism of ectogenesis as a research goal is that it is not one project. Partial support is an achievable extension of neonatal medicine with a clear patient population; complete ectogenesis is a problem in developmental biology with no obvious clinical driver, since [[embryo-selection]] and IVF already address most infertility, [[uterus-transplantation]] has produced live births for women without a functioning uterus, and gestational surrogacy addresses much of the rest. Framing the two together attracts opposition to the clinical programme without advancing the biological one, in much the way that speculation about [[human-cloning]] shaped the reception of therapeutic stem-cell research it had little to do with.

A distributional criticism applies to either version. Gestation in a device would be priced somewhere above an IVF cycle and would require sustained specialist supervision, which places it with the technologies analysed in [[access-and-inequality]] rather than with public-health measures. It would also do nothing for the majority of the world's pregnancies, whose principal risks are lack of skilled attendance and haemorrhage.

The strongest ethical criticism from the bioconservative side, surveyed in [[bioconservatism]], is not that ectogenesis is unsafe but that it treats gestation as a service that can be outsourced without remainder. Leon Kass's argument that some repugnance tracks a real moral insight has been applied here more often than to almost any other technology. The counter, from the tradition running through [[procreative-beneficence]] and [[bioethics-of-enhancement]], is that the same argument was made about anaesthesia in childbirth, IVF, and [[mitochondrial-replacement-therapy]], and has a poor record.

What would change the argument is a demonstration. Complete ectogenesis in a mouse, using an artificial implantation substrate rather than a transferred fetus, would move the question out of philosophy and into a regulatory queue. No group has publicly claimed to be close, and the placental biology needed for it is not obviously easier than the problems of [[lab-grown-organs]] that have resisted solution for three decades.

## See also

- [[artificial-womb]]
- [[in-vitro-gametogenesis]]
- [[synthetic-embryos]]
- [[embryo-selection]]
- [[reproductive-longevity]]
- [[bioconservatism]]
- [[procreative-beneficence]]
- [[morphological-freedom]]

## References

[^partridge2017]: `paper` Partridge, E. A. et al. "An extra-uterine system to physiologically support the extreme premature lamb." *Nature Communications*, 2017. {Eight preterm lambs supported for up to four weeks; sheep only, and the animals were euthanised at the end of support.}
[^haldane1924]: `book` Haldane, J. B. S. *Daedalus; or, Science and the Future*. Kegan Paul, Trench, Trubner, 1924. Delivered as a paper to the Heretics Society, Cambridge, in 1923.
[^huxley1932]: `book` Huxley, A. *Brave New World*. Chatto & Windus, 1932.
[^hanna2021]: `paper` Aguilera-Castrejon, A. et al. "Ex utero mouse embryogenesis from pre-gastrulation to late organogenesis." *Nature*, 2021. {The embryos were recovered from a uterus after implantation had already happened, so the system reproduces neither implantation nor placentation.}
[^romanis2018]: `paper` Romanis, E. C. "Artificial womb technology and the frontiers of human reproduction: conceptual differences and potential implications." *Journal of Medical Ethics*, 2018. {A legal scholar's proposal for a new category; gestateling has no statutory standing in any jurisdiction.}
[^firestone1970]: `book` Firestone, S. *The Dialectic of Sex: The Case for Feminist Revolution*. Morrow, 1970.
[^smajdor2007]: `paper` Smajdor, A. "The Moral Imperative for Ectogenesis." *Cambridge Quarterly of Healthcare Ethics*, 2007. {An argument that the technology ought to be developed, written a decade before any device sustained a mammalian fetus.}
[^corea1985]: `book` Corea, G. *The Mother Machine: Reproductive Technologies from Artificial Insemination to Artificial Wombs*. Harper & Row, 1985.
[^thomson1971]: `paper` Thomson, J. J. "A Defense of Abortion." *Philosophy & Public Affairs*, 1971.

==============================================================================
ARTICLE: effective-accelerationism
TITLE:   Effective accelerationism
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/effective-accelerationism
SOURCE:  https://futurehumanwiki.com/raw/effective-accelerationism
==============================================================================

---
title: "Effective accelerationism"
slug: "effective-accelerationism"
type: "concept"
status: "contested"
horizon: "present"
categories: ["people", "society"]
tags: ["movements", "artificial intelligence", "technology policy", "optimism", "ai safety", "internet culture"]
summary: "An online techno-optimist movement holding that technological acceleration is a moral imperative and that precautionary restraint on AI causes more harm than it prevents."
updated: "2026-07-28"
issues: ["The criticism of the Marinetti reference is reported without a citation."]
---

```infobox
{
  "caption": "Online intellectual movement",
  "rows": [
    { "label": "Emerged", "value": "2022" },
    { "label": "Abbreviation", "value": "e/acc" },
    { "label": "Origin medium", "value": "Anonymous accounts on X" },
    { "label": "Best-known figure", "value": "Guillaume Verdon" },
    { "label": "Core claim", "value": "Acceleration is a moral duty" },
    { "label": "Principal opponent", "value": "AI safety and effective altruism" },
    { "label": "Antecedent", "value": "Extropianism", "link": "/wiki/extropianism" }
  ]
}
```

**Effective accelerationism**, usually written e/acc, is a techno-optimist movement that emerged online in 2022 holding that the development and deployment of powerful technology — [[artificial-general-intelligence]] above all — should be accelerated rather than restrained, and that attempts to slow it are themselves the greater harm. Its name is a deliberate inversion of effective altruism, the movement it defines itself against. It is better understood as a rhetorical and political formation than as a research programme: it has produced manifestos, memes and companies, but no body of theory.

## Overview

The movement's positive claim is that technological progress is the mechanism by which human welfare has historically improved, that this mechanism is now being obstructed by a coalition of regulators and safety researchers, and that the expected cost of obstruction — lives not saved, poverty not ended, capabilities not gained — exceeds the expected cost of moving fast. Its negative claim, pressed harder and more often, is that AI-risk arguments are unfalsifiable, self-serving to the institutions that make them, and a route to regulatory capture by incumbents.

Neither claim is new. Both were made in near-identical form by [[extropianism|extropians]] in the 1990s, and the risk-of-inaction argument is a restatement of the proactionary principle that [[max-more]] drafted in 2004 as a counter to the [[precautionary-principle]]. The same structure appears in arguments for unrestricted [[human-enhancement]] and against moratoria on [[germline-editing]]: delay is not neutral, and the harms of waiting are real but diffuse and therefore politically invisible. What is new is the medium and the political moment.

## Origins

E/acc coalesced on X, formerly Twitter, during 2022, propagated by a small number of pseudonymous accounts, chiefly one writing as "Beff Jezos" and another as "Bayeslord". Its early output consisted of Substack posts and a stream of posts in a deliberately hyperbolic register, mixing thermodynamics, startup vocabulary and irony in a way that made the boundary between argument and performance difficult to locate — a feature its participants regarded as a defence against outside criticism.

In December 2023 *Forbes* identified Beff Jezos as Guillaume Verdon, a physicist who had worked on quantum machine learning at Google and had founded a hardware startup, Extropic, whose name is an explicit nod to the earlier movement.[^forbes2023] The unmasking gave the movement a named figure and moved it from anonymous internet culture into technology-industry politics.

Its most consequential adjacent text is Marc Andreessen's "The Techno-Optimist Manifesto", published in October 2023, which argues that technology is the source of all growth and that "our enemy" includes sustainability, trust and safety, tech ethics, and risk management as named categories.[^andreessen2023] Andreessen listed a set of intellectual "patron saints" that included the Italian Futurist Filippo Marinetti, whose later fascist affiliations drew immediate criticism and which Andreessen did not address.

```timeline
[
  { "year": "2022", "title": "The term appears", "text": "Anonymous accounts on X begin using 'e/acc' as a self-description, framed against effective altruism and AI-safety advocacy." },
  { "year": "2023", "title": "Manifesto season", "text": "Marc Andreessen publishes 'The Techno-Optimist Manifesto', which names risk management and tech ethics among its enemies and is widely read as an e/acc-adjacent document." },
  { "year": "2023", "title": "Beff Jezos identified", "text": "Forbes reports that the movement's most visible account belongs to Guillaume Verdon, a former Google physicist and founder of the hardware startup Extropic." },
  { "year": "2023–2024", "title": "Entry into industry politics", "text": "The vocabulary spreads into venture capital and AI-company hiring, and 'e/acc' appears in professional profiles as a political signal." }
]
```

## The thermodynamic argument

The movement's characteristic move is to ground its politics in physics. The claim, in its usual form, is that life and intelligence are dissipative structures that persist by increasing the rate at which free energy is degraded, that the universe therefore selects for systems that consume more energy and process more information, and that accelerating technology aligns with this tendency. Resisting it is described as fighting thermodynamics.

The scientific substrate for this is Jeremy England's work on dissipation-driven adaptation, which derives statistical-mechanical constraints on the fluctuations of driven systems and argues that self-replication is thermodynamically favoured under certain conditions.[^england2013] England's results are legitimate non-equilibrium statistical mechanics. They do not establish a direction for history, do not imply that any particular technology should be built, and do not license normative conclusions of any kind. The inference from "driven systems tend to organize" to "therefore build AGI faster" has no support in the underlying physics, and nothing in England's papers asserts it.

> [!caution] The physics does not do the work
> Non-equilibrium thermodynamics places constraints on what physical systems can do. It contains no premise about what anyone ought to do. Every version of the e/acc argument that appears to derive an imperative from entropy production has smuggled in an unstated value claim between the two.

## Positions and opponents

E/acc's practical positions cluster around a small number of issues:

- **Open weights.** Model weights should be released publicly; concentration of capability in a few laboratories is treated as the primary risk, ahead of misuse.
- **Against licensing regimes.** Proposals to require government approval for training runs above a compute threshold are read as incumbent protection.
- **Against pause proposals.** Calls for a moratorium on frontier training are rejected outright, with the [[dual-use-research]] debates in biology sometimes cited as an example of restriction failing to prevent capability spread while suppressing beneficial work.
- **Abundance framing.** Energy, compute and housing constraints are treated as artificial and politically imposed rather than physical.

Its opponents are AI-safety researchers, the effective-altruism funding network, and the broader precautionary tradition in technology policy. The disagreement about [[existential-risk]] is genuine and unresolved, but much of the public conflict has been about status and institutional power rather than about probabilities. Both sides accuse the other of motivated reasoning, and both are partly right: safety institutions are funded to find risk, and acceleration advocates are frequently invested in the companies whose regulation is at issue. The dispute over whether a [[technological-singularity]] would be survivable is the same one that split [[singularitarianism]] into safety and acceleration wings twenty years earlier, conducted at higher volume and with real money at stake.

## Relation to earlier accelerationisms

The word "accelerationism" has an older and largely unrelated history in critical theory. Nick Land's writing at the Cybernetic Culture Research Unit in the 1990s treated capitalism and technology as an impersonal process that would dissolve the human, and did so approvingly; the position is nihilist rather than optimist, and its later political trajectory took it to the far right.[^land2011] Left accelerationism, set out by Nick Srnicek and Alex Williams in 2013, argued instead for seizing technological capacity from capital and directing it toward post-scarcity ends.[^srnicek2013]

E/acc borrows the family name and almost nothing else. Its content — individual optimism, entrepreneurship, hostility to regulation, faith in technological solutions to social problems — is closer to Californian libertarianism, to [[morphological-freedom]] as a political principle, and to the [[transhumanism|transhumanist]] mainstream than to either. Participants have at times acknowledged the Land lineage and at other times disclaimed it, a flexibility consistent with the movement's ironic register.

## Criticism

The most common substantive criticism is that e/acc has no theory of failure. It argues that acceleration is good on net without specifying what evidence would change that assessment, and it treats the historical record of technology improving welfare as if it entailed that all future technologies will do so. That inference does not hold for technologies whose failure modes are irreversible, which is precisely the category the argument is meant to address.

A second criticism is that "acceleration" has no referent. Technological progress is not a scalar with a throttle; a policy that accelerates semiconductor fabrication may slow biomedical research by absorbing capital and talent. [[differential-technological-development]], the proposal that [[nick-bostrom]] set out to address exactly this, is rejected by e/acc on the grounds that no one can sequence technologies reliably — an objection that applies with equal force to undifferentiated acceleration. Absent a mechanism, the position reduces to opposition to whatever regulation is currently proposed.

A third concerns access. The movement's welfare argument depends on benefits diffusing broadly, but it has little to say about the [[access-and-inequality]] mechanisms by which that would happen, and its preferred policy of minimal state involvement is the one least likely to produce it. This is the same gap that [[bioconservatism|bioconservative]] critics identified in transhumanism decades earlier, arriving from a different direction.

A fourth is stylistic but not trivial. Argument conducted primarily through irony is difficult to hold to account: any position can be disowned as a joke, and the movement's most extreme statements — that human extinction by superior intelligence would be an acceptable outcome of cosmic development — are made and retracted in the same register. Whether anyone holds that view sincerely is genuinely unclear.

## Influence

The movement's measurable effects are political rather than technical. Its vocabulary is now standard in parts of the venture-capital and AI-startup world, it supplied a ready-made counter-position during the regulatory debates of 2023 through 2025, and it made open-weight release a partisan identity rather than a technical choice. It has not produced research, a policy programme with detail, or an institution comparable to those it opposes.

Its longer significance may be as evidence about how technological ideologies now form. Extropianism took a mailing list and a decade to assemble a few hundred people; e/acc assembled a comparable intellectual position, and considerably more political weight, in under two years and without a single peer-reviewed page. Whether a movement built that way can sustain a position once its opponents' predictions become testable is the question its next few years will answer.

## See also

- [[extropianism]]
- [[precautionary-principle]]
- [[technological-singularity]]
- [[differential-technological-development]]
- [[existential-risk]]
- [[bioconservatism]]
- [[transhumanism]]
- [[artificial-general-intelligence]]

## References

[^forbes2023]: `news` Baker-White, E. "Who Is Beff Jezos, The Leader Of The Tech Elite's 'E/Acc' Movement?" *Forbes*, December 2023. {Identifies who was behind a pseudonymous account; it establishes that, not how large the movement is or what it holds.}
[^andreessen2023]: `statement` Andreessen, M. "The Techno-Optimist Manifesto." Andreessen Horowitz, October 2023. {A venture firm's own manifesto, adjacent to e/acc rather than of it; it records one investor's position, not the movement's.}
[^england2013]: `paper` England, J.L. "Statistical physics of self-replication." *The Journal of Chemical Physics*, 2013.
[^land2011]: `book` Land, N. *Fanged Noumena: Collected Writings 1987–2007*. Urbanomic/Sequence Press, 2011.
[^srnicek2013]: `statement` Srnicek, N. and Williams, A. "#Accelerate: Manifesto for an Accelerationist Politics." *Critical Legal Thinking*, 2013. {A manifesto the authors published for themselves; it states the left-accelerationist position rather than reporting on it.}

==============================================================================
ARTICLE: ecog-interfaces
TITLE:   Electrocorticography interfaces
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/ecog-interfaces
SOURCE:  https://futurehumanwiki.com/raw/ecog-interfaces
==============================================================================

---
title: "Electrocorticography interfaces"
slug: "ecog-interfaces"
type: "technology"
status: "emerging"
horizon: "late 2020s"
trl: 6
categories: ["cybernetics"]
tags: ["bci", "ecog", "speech decoding", "neural recording", "implants", "epilepsy"]
summary: "Brain–computer interfaces that record from electrode arrays resting on the cortical surface rather than penetrating it, trading single-neuron resolution for stability and coverage."
updated: "2026-07-27"
humanEvidence: "Almost the entire evidence base is human: epilepsy patients temporarily implanted for seizure mapping, plus single participants with paralysis or ALS who have used chronic arrays for speech, cursor and walking control."
access: "Clinical grids are cleared only for temporary use around epilepsy surgery; every chronic interface use is a small investigational study, and no chronic product is approved."
reversibility: "difficult"
issues: ["The 2025 US clearance for a high-density cortical film is neither named nor sourced."]
---

```infobox
{
  "caption": "Class of neural interface",
  "rows": [
    { "label": "Signal", "value": "Cortical surface potentials" },
    { "label": "Key frequency band", "value": "High gamma, ~70–200 Hz" },
    { "label": "Spatial resolution", "value": "Millimetres" },
    { "label": "Clinical origin", "value": "Epilepsy seizure mapping" },
    { "label": "Typical contacts", "value": "64–1,024" },
    { "label": "Placement", "value": "Subdural or epidural" },
    { "label": "Regulatory status", "value": "Cleared for temporary use; chronic BCI investigational" }
  ]
}
```

**Electrocorticography interfaces** are brain–computer interfaces that record from electrode arrays
laid on the surface of the cortex, either beneath the dura (subdural) or on top of it (epidural),
without penetrating brain tissue. They occupy the middle of the invasiveness spectrum: better
spatial and spectral resolution than scalp electroencephalography, worse than a penetrating
[[utah-array]], with mechanical stability that neither of the alternatives matches.

The technique exists because neurosurgery needed it first. Grids of platinum contacts have been
placed on the cortex for decades to localize seizure onset and map eloquent cortex before epilepsy
surgery, which means the field inherited a large clinical base of patients temporarily instrumented
with electrodes and willing to participate in research. Most of what is known about human cortical
representations of speech comes from that population.

```figure
{"key": "bci-sensor-types", "caption": "Three placements and three trade-offs: signal quality rises with invasiveness, and so does surgical risk."}
```

## How it works

A standard clinical grid is a silicone sheet holding platinum discs a few millimetres across on a
one-centimetre pitch. Research arrays go finer: high-density grids at two to four millimetres, and
thin-film micro-electrocorticography arrays with hundreds or thousands of contacts at
sub-millimetre spacing on a substrate thin enough to conform to the cortical surface, including
into sulci.

The most useful feature for a [[brain-computer-interface]] is high-gamma power, roughly 70 to 200
hertz. Broadband high-frequency activity recorded at the surface correlates well with the firing
rate of the population of neurons directly underneath, so it functions as a proxy for local spiking
that can be measured without pushing anything into the tissue. Lower-frequency rhythms carry
complementary information about state and movement preparation. Together these features feed
standard [[neural-decoding]] pipelines.

Because the array does not penetrate, it does not sever capillaries or provoke the encapsulation
response that displaces neurons from a penetrating tip. Fibrous tissue does form around a subdural
grid, but the recorded signal degrades far more slowly than intracortical spikes do. This is the
central engineering argument for the approach.

> [!note] The resolution ceiling
> Surface recording cannot resolve individual neurons, and no improvement in electrode density
> changes that. Cerebrospinal fluid and the dura conduct and blur; the signal at any contact is a
> weighted sum over a cortical patch. Denser arrays sample that blurred field more finely, which
> helps, but the information about individual cells is gone before it reaches the electrode.

## Development history

The clinical practice long predates the interface application. Patients admitted for invasive
seizure monitoring spend one to two weeks with grids or depth electrodes in place, awake and
available to perform tasks, and that window has supplied most of the human cortical physiology
behind modern decoders. The same population, instrumented with hippocampal depth electrodes rather
than surface grids, supplies the data behind [[memory-prosthesis]] research. The arrangement has
its own ethical literature: participants are patients first, recruited during a stressful clinical
episode, and the research question is never the reason the electrodes are there.

```timeline
[
  { "year": "1930s–1950s", "title": "Clinical electrocorticography", "text": "Wilder Penfield and Herbert Jasper establish intraoperative cortical recording and stimulation mapping at the Montreal Neurological Institute." },
  { "year": "2004–2010", "title": "ECoG as a BCI signal", "text": "Groups working with epilepsy patients show that high-gamma activity from surface electrodes supports rapid, accurate cursor control learned within minutes." },
  { "year": "2013", "title": "Multi-dimensional arm control", "text": "A participant with tetraplegia controls a robotic arm in three dimensions using a subdural array, demonstrating that surface signals support continuous control." },
  { "year": "2016", "title": "Home use in ALS", "text": "A fully implanted electrocorticography system in Utrecht lets a woman with late-stage ALS select letters at home, remaining functional over years." },
  { "year": "2019", "title": "Exoskeleton control", "text": "A French team reports a tetraplegic participant operating a whole-body exoskeleton using bilateral wireless epidural implants." },
  { "year": "2019–2025", "title": "Speech decoding", "text": "Successive reports from UCSF decode sentences, then conversational-rate text, synthesized voice, and near-real-time streaming speech from surface arrays over speech motor cortex." },
  { "year": "2023", "title": "Brain–spine interface", "text": "Epidural cortical implants drive a spinal stimulator to restore volitional walking after spinal cord injury." }
]
```

## Speech and communication

Surface arrays have dominated speech decoding for an anatomical reason. The representations of the
lips, jaw, tongue, and larynx are laid out across several centimetres of ventral sensorimotor
cortex on the lateral surface of the brain — a large, shallow, two-dimensional map, which is
exactly what a conformal grid is good at sampling. Penetrating arrays sample it densely in a few
square millimetres; a grid samples it sparsely across the whole map.

The line of work from UCSF began with synthesizing intelligible speech from cortical activity
recorded during spoken sentences,[^anumanchipalli2019] then decoded attempted words in a man with
anarthria after brainstem stroke,[^moses2021] then reached conversational rates with text,
synthesized voice, and an animated avatar,[^metzger2023] and more recently reduced the latency
enough for near-continuous streaming output. These results, discussed in detail under
[[speech-neuroprosthesis]], come from single participants using research hardware.

Communication systems built on smaller chronic implants have a longer service record. A fully
implanted system using a small number of surface electrode strips was used at home by a participant
with ALS to select letters for years, without recalibration drift severe enough to stop it
working.[^vansteensel2016] The bit rate was low. The reliability was the point.

## Movement restoration

Surface arrays also drive motor systems. A participant with tetraplegia achieved three-dimensional
robotic arm control from a subdural grid,[^wang2013] and a wireless epidural implant developed in
Grenoble allowed a tetraplegic participant to operate a powered [[exoskeleton]], though slowly and
under laboratory supervision.[^benabid2019]

The most consequential application connects cortex to spinal cord rather than to a machine. In a
2023 report, epidural cortical implants decoded walking intention and drove an implanted epidural
spinal stimulator, restoring volitional walking to a man with an incomplete spinal cord injury, who
also showed some neurological recovery that persisted with the system off.[^lorach2023] This "digital bridge" framing — using a BCI to reconnect the nervous system to
itself rather than to a robot — is now a major direction in [[neuroprosthetics]].

## Advantages and limits

The advantages are stability, coverage, and channel scaling. Thin-film arrays can carry a thousand
or more contacts without a thousand penetrations, and one company's high-density cortical film
received United States clearance in 2025 for temporary use during surgery, a step toward chronic
approval. The durability argument has an existing precedent: [[deep-brain-stimulation]] leads sit
in human brains for a decade or more without failing, which suggests that a chronically implanted
neural device is not intrinsically impossible, only that penetrating recording tips are a
particularly demanding case. Flexible penetrating threads of the kind used by [[neuralink]] attempt
to split the difference, accepting penetration but minimizing stiffness.

The limits are structural. Access still requires opening the skull, though narrow-slot and
burr-hole insertion techniques reduce the exposure compared with a full craniotomy. Subdural
placement carries risks of haemorrhage, infection, and cerebrospinal fluid leak. Grids can shift.
And the resolution ceiling caps the number of independent control dimensions available, which is
why the highest-dimensional results still come from penetrating electrodes, and why the endovascular
[[stentrode]] — with far fewer contacts and greater distance from cortex — sits lower still.

Any chronic recording of cortical activity, whatever the electrode, generates a data stream from
which more can be inferred than the user intends to communicate. That problem is examined under
[[mental-privacy]], and it applies with particular force to speech-decoding systems.

## Outlook

The plausible near-term path runs through fully implanted, wireless, high-channel thin films
targeted at speech loss, where the clinical need is unambiguous and the decoding results are
strongest. Whether such devices sustain performance for a decade, and whether the surgery becomes
routine enough for a neurologist to recommend it to a newly diagnosed ALS patient, are open
questions that no current trial is old enough to answer.

A quieter question is what surface arrays reveal about cortical organization itself. Dense grids
over speech and motor cortex have produced maps that were unavailable from animal work, and that
information feeds back into decoder design — a loop between measurement and application that
resembles the relationship between [[connectomics]] and theories of neural computation.

## See also

- [[brain-computer-interface]]
- [[speech-neuroprosthesis]]
- [[utah-array]]
- [[stentrode]]
- [[neural-decoding]]
- [[neuroprosthetics]]
- [[neuralink]]
- [[mental-privacy]]

## References

[^anumanchipalli2019]: `paper` Anumanchipalli, G. K., Chartier, J. and Chang, E. F. "Speech synthesis from neural decoding of spoken sentences." *Nature*, 2019.

[^moses2021]: `paper` Moses, D. A. et al. "Neuroprosthesis for decoding speech in a paralyzed person with anarthria." *New England Journal of Medicine*, 2021.

[^metzger2023]: `paper` Metzger, S. L. et al. "A high-performance neuroprosthesis for speech decoding and avatar control." *Nature*, 2023. {One participant, severely paralysed after a brainstem stroke; the rates reported are for a decoder trained on her own attempted speech in laboratory sessions.}

[^vansteensel2016]: `paper` Vansteensel, M. J. et al. "Fully implanted brain–computer interface in a locked-in patient with ALS." *New England Journal of Medicine*, 2016. {A single participant; what the report establishes is sustained unattended home use at a very low bit rate, not decoding performance.}

[^wang2013]: `paper` Wang, W. et al. "An electrocorticographic brain interface in an individual with tetraplegia." *PLOS ONE*, 2013.

[^benabid2019]: `paper` Benabid, A. L. et al. "An exoskeleton controlled by an epidural wireless brain–machine interface in a tetraplegic patient: a proof-of-concept demonstration." *The Lancet Neurology*, 2019.

[^lorach2023]: `paper` Lorach, H. et al. "Walking naturally after spinal cord injury using a brain–spine interface." *Nature*, 2023. {A single participant who had already undergone spinal stimulation and rehabilitation, which complicates attributing the residual recovery to the brain-spine link.}

==============================================================================
ARTICLE: embryo-selection
TITLE:   Embryo selection
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/embryo-selection
SOURCE:  https://futurehumanwiki.com/raw/embryo-selection
==============================================================================

---
title: "Embryo selection"
slug: "embryo-selection"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["reproduction", "genetics"]
tags: ["reproduction", "ivf", "embryo selection", "genomics", "bioethics", "screening"]
summary: "Testing embryos created by IVF and choosing which to transfer, a routine clinical practice for single-gene disease whose extension to complex traits remains disputed."
updated: "2026-07-27"
humanEvidence: "In human clinical use since the first births from biopsied embryos in 1990: testing for a known single-gene variant reliably identifies affected embryos, while randomized trials of aneuploidy screening have not shown higher live-birth rates."
access: "Legal and clinically available in nearly every country with IVF, but a cycle with genetic testing costs more than most health systems will reimburse for a couple who are not infertile."
reversibility: "irreversible"
issues: ["The 2008 UK legislation on tissue-type matching is described without a citation.", "The claim that PGT-A is used on a large share of cycles in some markets carries no source."]
---

```infobox
{
  "caption": "Clinical reproductive genetics",
  "rows": [
    { "label": "First clinical use", "value": "1990" },
    { "label": "Sample", "value": "Trophectoderm biopsy, day 5–6" },
    { "label": "Established forms", "value": "PGT-M, PGT-SR" },
    { "label": "Disputed forms", "value": "PGT-A, PGT-P" },
    { "label": "Embryos per cycle", "value": "Usually a handful" },
    { "label": "Main limit", "value": "Sibling genetic similarity" }
  ]
}
```

**Embryo selection** is the practice of testing embryos created by in vitro fertilisation and choosing which ones to transfer to a uterus. It is the only form of genetic choice over offspring in routine clinical use anywhere, and it is the concrete reality behind most of what is discussed under the heading of [[designer-babies]]. What it can deliver depends almost entirely on whether the trait in question is determined by a single locus or by thousands, and public debate consistently underweights that distinction.

## The four tests

Preimplantation genetic testing is now divided by what it looks for. The names were standardised in the late 2010s, replacing the older PGD and PGS labels.

- **PGT-M**, for monogenic disease. Where the parents' variants are known, embryos can be genotyped for the disease allele alongside linked markers from family members, which guards against amplification failure. Accuracy is high and this is the technique's uncontested core.
- **PGT-SR**, for structural rearrangements. Carriers of balanced translocations produce a high proportion of unbalanced embryos; testing identifies them.
- **PGT-A**, for aneuploidy. A screen for the wrong number of chromosomes, done on the assumption that transferring euploid embryos raises live-birth rates. Widely sold, and the least well supported.
- **PGT-P**, for polygenic scores. Ranking embryos by predicted risk of common disease or by predicted value of a complex trait. Offered by a small number of commercial laboratories and endorsed for clinical use by no major professional society. It has its own article at [[polygenic-embryo-screening]].

The physical procedure is the same in each case: five to ten cells are removed from the trophectoderm of a day-5 or day-6 blastocyst, the portion that becomes placenta rather than fetus, and their DNA is amplified and sequenced. The embryo is vitrified while the result is awaited.

## Origins

The first births followed embryo biopsy in 1990, when a team led by Alan Handyside amplified a Y-specific sequence to identify female embryos for couples carrying X-linked disorders.[^handyside1990] Selecting on sex was a proxy; testing for the disease allele itself followed within a few years as PCR and later array and sequencing methods improved.

Two cases shaped public understanding more than the technique did. In 2000, preimplantation testing was used both to avoid Fanconi anaemia and to select an embryo whose tissue type matched an affected sibling, so that cord blood from the resulting child could treat her.[^verlinsky2001] In 2008 the United Kingdom legislated on the practice explicitly, permitting tissue-type matching under licence and prohibiting the deliberate preference of an embryo known to carry a serious abnormality over one that does not.

## Why PGT-A is contested

PGT-A is performed on a large share of IVF cycles in some markets, and the evidence that it improves outcomes is weak. A multicentre randomised trial published in 2019 found no improvement in ongoing pregnancy rate per patient randomised, with a possible benefit in a subgroup of older patients in per-protocol analysis.[^munne2019] A subsequent randomised trial in China, restricted to patients with several good-quality blastocysts, likewise failed to show superiority over conventional morphological selection on cumulative live-birth rate.[^yan2021]

The mechanistic explanation is that the biopsy is not the embryo. Trophectoderm cells can be aneuploid while the inner cell mass is not, and blastocysts classified as mosaic have produced healthy children after transfer.[^greco2015] Because the test causes some embryos to be discarded, a screen with imperfect predictive value can reduce the number of transfers without improving the odds of any one of them. This is the same statistical structure that makes overdiagnosis a problem in cancer screening and in [[consumer-blood-testing|direct-to-consumer testing]].

> [!caution] A test's accuracy is not its clinical value
> PGT-A measures chromosome copy number in the sampled cells reliably. That is a different claim from the claim that acting on the measurement produces more babies, which is what patients are buying and what the randomised evidence does not support.

## The arithmetic that caps the gain

Selection can only choose among embryos that exist, and each embryo is a recombination of two genomes the parents already have. Two constraints follow.

The first is supply. A stimulation cycle in a patient under 35 commonly yields a handful of usable blastocysts; the number falls with age, and many cycles in patients over 40 yield none. Selection therefore operates on a small sample.

The second is variance. Full siblings share about half their genomes, so the spread of any polygenic trait within one couple's embryos is much narrower than the spread across the population. Modelling by Ehud Karavani and colleagues estimated that choosing the top-scoring of ten embryos would shift adult height by roughly two and a half centimetres and a cognitive test score by a comparable number of points on average, with the realised gain in any individual case varying widely and sometimes being negative.[^karavani2019] Predictive scores also perform substantially worse within families than across populations, because much of their apparent accuracy comes from population structure and indirect parental effects rather than from the embryo's own genotype.[^turley2021]

These limits are not artefacts of current technology. Better prediction narrows the gap between realised and theoretical gain, but the theoretical gain itself is set by sibling similarity and the number of embryos. Only [[in-vitro-gametogenesis]], by supplying far more embryos, would loosen the second constraint, and even then the return diminishes sharply: a hundredfold increase in the number of embryos available is worth less than a doubling of the expected shift. The consequences of that ceiling for cognitive traits specifically are worked through in [[genetic-enhancement-of-intelligence]].

Embryo supply also interacts with a separate question about what an embryo is. Stem-cell-derived structures of the kind described in [[synthetic-embryos]] are not products of fertilisation and could not be transferred, but they are increasingly used to study the developmental stages that selection is implicitly betting on.

## Objections

**Disability rights.** The expressivist objection, developed most fully by Adrienne Asch, holds that selecting against an embryo on the basis of a trait expresses a judgement about existing people with that trait: not that a particular life would be bad, but that a life like theirs is not worth starting.[^asch1999] The counterargument distinguishes preventing an impairment from devaluing a person, and points out that parents routinely act to reduce their future children's risks without insulting anyone. Deaf and autistic community positions on screening are among the sharpest cases and are examined in [[disability-rights-and-enhancement]].

**Scope creep.** Jurisdictions differ on non-medical sex selection, on testing for adult-onset and incompletely penetrant conditions such as *BRCA1* variants, and on saviour-sibling matching. Each extension has been argued for as a small step from the last, which is the structure critics identify as a slope and defenders as ordinary case-by-case reasoning. The licensing model used in the United Kingdom, where a regulator approves conditions one at a time, is the main institutional alternative to leaving the question to clinics; the broader regulatory picture is set out in [[governance-of-genome-editing]].

**Whether choosing is obligatory.** Julian Savulescu's principle of [[procreative-beneficence]] holds that parents who are already selecting have reason to select the embryo expected to have the best life, which turns a permission into a weak duty. Critics within [[bioethics-of-enhancement]] object both to the ranking that requires and to the assumption that expected welfare can be read off a genotype.

**Access.** An IVF cycle with genetic testing costs more than most health systems will reimburse for a non-infertile couple, which places selection among the technologies analysed in [[access-and-inequality]]. Unlike heritable [[germline-editing]], it is legal nearly everywhere, so cost rather than law is the binding constraint.

**Information asymmetry.** Reports generated by commercial polygenic services convey risk reductions that are small in absolute terms and are difficult to interpret without training. The concern raised by professional societies is less that patients will be harmed by a transfer decision than that they will make it on a misunderstanding of what the number means, a problem that also runs through [[genetic-discrimination]] and consumer genomics generally.

## Where it sits among the alternatives

Selection is often discussed as an early form of the same capability as embryo editing. It is not. Selection cannot introduce an allele neither parent carries and cannot exceed the best combination those two genomes can produce; editing can in principle do both, at the cost of the mosaicism and [[crispr-off-target-effects|off-target damage]] demonstrated badly in the [[he-jiankui-affair]]. For the narrow set of couples where selection fails outright, such as one parent homozygous for a dominant disease allele or both homozygous for a recessive one, editing is the only genetic option, and that is the strongest medical argument for it.

[[mitochondrial-replacement-therapy]] occupies a third position, replacing an organelle rather than selecting or editing nuclear DNA, and for many mitochondrial conditions selection among embryos with variable heteroplasmy is a competing and simpler option.

A further consideration is timing. Testing an embryo competes with testing a fetus, and non-invasive prenatal screening from maternal blood already detects common aneuploidies in ordinary pregnancies at far lower cost, which is why embryo selection remains concentrated among couples who are using IVF anyway or who carry a known variant.

The technical direction most likely to change practice is not more sensitive sequencing but non-invasive testing, in which cell-free DNA is recovered from the medium the embryo was cultured in — the same cell-free-DNA chemistry that underpins the liquid biopsies described in [[nanoparticle-diagnostics]]. It would remove the biopsy, which is not risk-free, but concordance with trophectoderm biopsy has so far been inconsistent enough that it remains a research method rather than a clinical one.

## See also

- [[polygenic-embryo-screening]]
- [[designer-babies]]
- [[in-vitro-gametogenesis]]
- [[germline-editing]]
- [[mitochondrial-replacement-therapy]]
- [[procreative-beneficence]]
- [[disability-rights-and-enhancement]]
- [[reproductive-longevity]]

## References

[^handyside1990]: `paper` Handyside, A. H. et al. "Pregnancies from biopsied human preimplantation embryos sexed by Y-specific DNA amplification." *Nature*, 1990.
[^verlinsky2001]: `paper` Verlinsky, Y. et al. "Preimplantation diagnosis for Fanconi anemia combined with HLA matching." *JAMA*, 2001.
[^munne2019]: `paper` Munné, S. et al. "Preimplantation genetic testing for aneuploidy versus morphology as selection criteria for single frozen-thawed embryo transfer in good-prognosis patients: a multicenter randomized clinical trial." *Fertility and Sterility*, 2019.
[^yan2021]: `paper` Yan, J. et al. "Live Birth with or without Preimplantation Genetic Testing for Aneuploidy." *New England Journal of Medicine*, 2021.
[^greco2015]: `paper` Greco, E., Minasi, M. G. and Fiorentino, F. "Healthy Babies after Intrauterine Transfer of Mosaic Aneuploid Blastocysts." *New England Journal of Medicine*, 2015. {A short case series of transfers; it shows mosaic embryos can produce healthy children, not how often they do.}
[^karavani2019]: `paper` Karavani, E. et al. "Screening Human Embryos for Polygenic Traits Has Limited Utility." *Cell*, 2019. {A modelling study: the gains are estimated, not observed in children born after polygenic selection.}
[^turley2021]: `paper` Turley, P. et al. "Problems with Using Polygenic Scores to Select Embryos." *New England Journal of Medicine*, 2021.
[^asch1999]: `paper` Asch, A. "Prenatal Diagnosis and Selective Abortion: A Challenge to Practice and Policy." *American Journal of Public Health*, 1999.

==============================================================================
ARTICLE: emmanuelle-charpentier
TITLE:   Emmanuelle Charpentier
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/emmanuelle-charpentier
SOURCE:  https://futurehumanwiki.com/raw/emmanuelle-charpentier
==============================================================================

---
title: "Emmanuelle Charpentier"
slug: "emmanuelle-charpentier"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "genetics"]
tags: ["crispr", "gene editing", "rna", "nobel prize", "microbiology", "bacteria"]
summary: "French microbiologist who identified the second RNA that CRISPR-Cas9 requires, co-authored the 2012 paper that made the system programmable, and shared the 2020 Nobel Prize in Chemistry."
updated: "2026-07-28"
---

```infobox
{
  "caption": "Microbiologist",
  "rows": [
    { "label": "Born", "value": "11 December 1968, Juvisy-sur-Orge, France" },
    { "label": "Nationality", "value": "French" },
    { "label": "Education", "value": "PhD, Université Pierre et Marie Curie, 1995" },
    { "label": "Known for", "value": "tracrRNA; programmable CRISPR–Cas9", "link": "/wiki/crispr-cas9" },
    { "label": "Field", "value": "Bacterial pathogens; RNA biology" },
    { "label": "Affiliation", "value": "Max Planck Unit for the Science of Pathogens, Berlin" },
    { "label": "Co-founded", "value": "CRISPR Therapeutics, 2013" },
    { "label": "Award", "value": "Nobel Prize in Chemistry, 2020" }
  ]
}
```

**Emmanuelle Charpentier** is a French microbiologist who reached CRISPR from the study of bacterial
disease rather than from genome engineering. Her laboratory identified tracrRNA, the second small RNA
that the type II CRISPR system needs in order to work, and she then co-authored the 2012 paper showing
that the system could be reduced to a protein and a single engineered guide and directed at any chosen
DNA sequence. She shared the 2020 Nobel Prize in Chemistry with [[jennifer-doudna]] for that result,
and has continued to run a laboratory on bacterial pathogens rather than on editing.

## Career

Charpentier was born in Juvisy-sur-Orge, south of Paris, studied biochemistry and microbiology at the
Université Pierre et Marie Curie, and completed a doctorate in 1995 on molecular mechanisms of
antibiotic resistance, with the research carried out at the Institut Pasteur. She then spent six years
in the United States — at Rockefeller University working on *Streptococcus pneumoniae*, then at NYU
Medical Center, St. Jude Children's Research Hospital and the Skirball Institute.

She returned to Europe in 2002 and moved repeatedly: the University of Vienna and the Max F. Perutz
Laboratories from 2002, the Laboratory for Molecular Infection Medicine Sweden at Umeå University from
2008, and the Helmholtz Centre for Infection Research in Braunschweig together with Hannover Medical
School from 2013. In 2015 she became a director at the Max Planck Institute for Infection Biology in
Berlin, and in 2018 the Max Planck Society established the Max Planck Unit for the Science of Pathogens
with her as founding director. Five countries and two decades of moves separate the doctorate from the
permanent position, and both of the papers the Nobel rests on were published while she was at Umeå.

## Small RNAs in a streptococcus

The subject that runs through Charpentier's career is how *Streptococcus pyogenes* — the group A
streptococcus, cause of scarlet fever, necrotizing fasciitis and a large burden of ordinary throat
infection — regulates itself with small non-coding RNA. Her group's early work described an RNA that
controls the synthesis of virulence factors in that organism, and the CRISPR result came out of the same
programme: a survey of the bacterium's small-RNA transcriptome, run with Jörg Vogel's group.

That survey turned up an abundant small RNA transcribed from beside the CRISPR locus. The 2011 paper
reporting it showed that this trans-encoded RNA — tracrRNA — base-pairs with the repeat portion of the
CRISPR precursor transcript, and that the resulting duplex is processed into mature CRISPR RNAs by the
host enzyme RNase III in the presence of the protein then called Csn1, now known as Cas9.[^deltcheva2011]

> [!key] Why the second RNA mattered
> The type I and type III systems studied at the time process their guides without a partner RNA. That
> the type II system needed one was not predictable from them, and nothing about Cas9 could be
> reconstituted correctly without it. The single guide of 2012 is a fusion of the two RNAs this paper
> identified.

## The 2012 result

Charpentier met Doudna at a conference in Puerto Rico in 2011, and the two laboratories agreed to work
out what Cas9 actually does. The resulting paper, with Martin Jinek and Krzysztof Chylinski as its first
two authors, showed with purified components that Cas9 alone cleaves double-stranded DNA when supplied
with the CRISPR RNA and tracrRNA, that the two can be fused into one chimeric guide without loss of
activity, and that the cut site is set by base-pairing to twenty nucleotides of that guide next to a
short motif in the target.[^jinek2012] The mechanism and its consequences are set out under
[[crispr-cas9]].

Virginijus Šikšnys's group in Vilnius reached an overlapping conclusion independently, showing that a
Cas9–crRNA complex from *Streptococcus thermophilus* cuts DNA at a sequence matching its guide. That
paper was received by its journal in May 2012 and published in September, after the *Science*
paper.[^gasiunas2012] The 2018 Kavli Prize in Nanoscience was awarded to Charpentier, Doudna and Šikšnys
together; the Nobel statutes permit three laureates and the 2020 chemistry prize named two.

> [!debate] Which experiment is the invention
> The 2012 work was biochemistry in a tube. Editing inside human cells was reported months later by
> other groups, among them one led by [[george-church]]. Whether the tube experiment already contains
> the invention, or whether making it work in a eukaryotic nucleus was the inventive step, is the
> question underneath both the credit dispute and the patent litigation, and it has no answer that both
> sides accept.

## Companies and translation

Charpentier co-founded CRISPR Therapeutics in 2013 with Rodger Novak and Shaun Foy, and ERS Genomics
the same year to license her patent rights outside human therapeutics. CRISPR Therapeutics and Vertex
Pharmaceuticals went on to develop the sickle cell and beta-thalassemia treatment approved in 2023 as
[[casgevy]], the first approved medicine using the system — an ex vivo cell therapy that requires
chemotherapy conditioning and a hospital stay, and whose price places it out of reach of most of the
patients the disease affects, as discussed under [[access-and-inequality]].

The clinical record beyond that remains thin. Delivery, not editing chemistry, is the binding
constraint: in vivo work has concentrated on the liver, where [[lipid-nanoparticles]] reliably deposit
their cargo, and the [[crispr-off-target-effects|off-target and large-deletion problems]] created by
double-strand breaks have pushed much of the field towards [[base-editing]], [[prime-editing]] and
[[epigenome-editing]], which avoid cutting both strands.

## Credit and the patent dispute

Charpentier is a party to the American patent fight in her own right. The group opposing the Broad
Institute is designated CVC, for the University of California, the University of Vienna and Charpentier;
the Broad's applications, covering use in eukaryotic cells, were granted first under an accelerated
route. The Patent Trial and Appeal Board ruled for the Broad on priority in 2022. In May 2025 the
Federal Circuit vacated that determination and sent it back, holding that the board had conflated the
legal standards for conceiving an invention and for reducing it to practice.[^fedcir2025] Nothing about
who invented what was decided by that ruling, which returned the question to the board rather than
answering it. The dispute has now run more than a decade across two continents, and the American
position has been reversed on appeal before; this article does not track its current posture, and a
reader who needs the present state of the American patents should check the docket rather than rely
on an encyclopedia entry.

The credit question is separate from the legal one and has its own history. A 2016 perspective essay in
*Cell* by Eric Lander, founding director of the Broad Institute, narrated CRISPR's development in a way
that many readers judged to understate Charpentier's and Doudna's contributions and to overstate the
Broad's; the criticism was directed both at the handling of the evidence and at the author's position
at an institution party to the dispute.[^lander2016] Lander told a US Senate committee in 2021 that he
had understated the two women's role.

## Reception and legacy

The 2020 prize was awarded "for the development of a method for genome editing" and was the first
science Nobel given to two women and no one else.[^nobel2020] Charpentier is the less publicly visible
of the two laureates and has stayed with the subject she started in: her Berlin unit is devoted to how
bacterial pathogens regulate themselves, with CRISPR one system among several, rather than to genome
engineering or its applications.

That distance is itself part of the record the field argues over. CRISPR arrived from curiosity-driven
microbiology, not from a programme aimed at editing genomes, which is the case regularly made for
funding basic work whose applications cannot be named in advance. The counter-observation is that the
same tool has since made a set of governance problems urgent — [[germline-editing|heritable editing]]
and its one confirmed clinical use in the [[he-jiankui-affair]], [[gene-drive|self-propagating drives]]
in wild populations, and the [[dual-use-research|dual-use]] questions that attach to cheap, precise
sequence modification — that a laboratory studying streptococcal virulence had no reason to anticipate
and no standing to settle. The institutions that must settle them are the subject of
[[governance-of-genome-editing]].

## See also

- [[crispr-cas9]]
- [[jennifer-doudna]]
- [[casgevy]]
- [[base-editing]]
- [[germline-editing]]
- [[governance-of-genome-editing]]
- [[somatic-gene-therapy]]
- [[crispr-off-target-effects]]

## References

[^deltcheva2011]: `paper` Deltcheva, E., Chylinski, K., Sharma, C. M., Gonzales, K., Chao, Y., Pirzada, Z. A., Eckert, M. R., Vogel, J. and Charpentier, E. "CRISPR RNA maturation by trans-encoded small RNA and host factor RNase III." *Nature*, 2011. {The work is bacterial cell biology; it describes how a streptococcus processes its own guide RNAs and makes no claim about editing anything.}
[^jinek2012]: `paper` Jinek, M., Chylinski, K., Fonfara, I., Hauer, M., Doudna, J. A. and Charpentier, E. "A Programmable Dual-RNA-Guided DNA Endonuclease in Adaptive Bacterial Immunity." *Science*, 2012. {Purified protein and RNA cutting DNA in a tube; the paper reports no experiment inside a living cell of any kind.}
[^gasiunas2012]: `paper` Gasiunas, G., Barrangou, R., Horvath, P. and Šikšnys, V. "Cas9–crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria." *PNAS*, 2012. {An independent line of work on a different streptococcal species, received in May 2012 and published in September.}
[^fedcir2025]: `law` Regents of the University of California v. Broad Institute, US Court of Appeals for the Federal Circuit, decided 12 May 2025. {The court vacated and remanded on the standard for conception; it made no finding about who invented eukaryotic CRISPR editing.}
[^lander2016]: `statement` Lander, E. S. "The Heroes of CRISPR." *Cell*, 2016. {A perspective essay by the founding director of an institution party to the patent dispute, so it records a participant's account rather than an independent history.}
[^nobel2020]: `statement` The Royal Swedish Academy of Sciences. "The Nobel Prize in Chemistry 2020." Press release, 2020.

==============================================================================
ARTICLE: engineered-pandemics
TITLE:   Engineered pandemics
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/engineered-pandemics
SOURCE:  https://futurehumanwiki.com/raw/engineered-pandemics
==============================================================================

---
title: "Engineered pandemics"
slug: "engineered-pandemics"
type: "risk"
status: "speculative"
horizon: "present"
categories: ["society", "genetics"]
tags: ["biosecurity", "pandemics", "gain of function", "governance", "catastrophic risk", "synthetic biology"]
summary: "Outbreaks caused by a pathogen deliberately modified, synthesised, or released rather than one that emerged from nature, and the biosecurity programme built around preventing them."
updated: "2026-07-28"
humanEvidence: "No engineered pathogen has caused a pandemic; the documented toll from deliberate or accidental release runs to dozens of deaths, chiefly the 1979 Sverdlovsk anthrax escape and the 2001 anthrax letters."
access: "The enabling inputs are ordinary: pathogen sequences are published, synthetic DNA is a commercial service, and order screening is voluntary in most jurisdictions."
reversibility: "irreversible"
issues: ["Wastewater and travel-hub metagenomic surveillance programmes are described without citations."]
---

```infobox
{
  "caption": "Category of catastrophic biological risk",
  "rows": [
    { "label": "Events on record", "value": "None" },
    { "label": "Nearest precedents", "value": "Sverdlovsk 1979, anthrax letters 2001" },
    { "label": "Defining research episode", "value": "H5N1 ferret transmissibility, 2011–2012" },
    { "label": "Treaty basis", "value": "Biological Weapons Convention, 1972" },
    { "label": "Verification mechanism", "value": "None" },
    { "label": "Current US policy", "value": "Executive Order 14292, 2025" },
    { "label": "Standard framing", "value": "Delay, detect, defend" }
  ]
}
```

**Engineered pandemics** are outbreaks caused by a pathogen that has been deliberately modified, deliberately built from a published sequence, or deliberately released, rather than one that crossed into humans from an animal reservoir. No such event is known to have happened. The category is nonetheless treated as one of the more plausible routes to a global catastrophe, for three reasons that are structural rather than speculative: the agent replicates on its own, the enabling knowledge is published and cannot be recalled, and the cost of the underlying techniques has fallen by orders of magnitude while the number of people able to use them has risen.

## What would make a pathogen pandemic-capable

Natural selection optimises pathogens for transmission, not for lethality, and the two usually trade against each other: an agent that incapacitates its host quickly spreads less. The combinations that concern biosecurity analysts are the ones evolution rarely assembles and an engineer might. Long presymptomatic infectiousness paired with high case fatality is the standard example. Resistance to existing vaccines and antivirals is another, and it is the one that turns a manageable agent into an unmanageable one without changing its biology much.

Whether such a combination can actually be built is not established. What has been shown is narrower and still significant: that a specific barrier to mammalian airborne transmission can be crossed deliberately. Two laboratories in 2011 and 2012 produced avian H5N1 influenza variants that spread between ferrets by respiratory droplet, identifying a small number of amino acid substitutions that sufficed.[^herfst2012] The transmissible viruses were not more lethal in that model, and they remained sensitive to oseltamivir. The result nonetheless established that the property most relevant to pandemic potential is engineerable, and it did so in the open literature.

Mortality in respiratory pandemics also concentrates steeply by age, tracking the decline in immune function described under [[immunosenescence]], which means the burden of any such event would fall on the same population that most other risk analysis in this wiki treats as the beneficiary of longer life.

## The record so far

```timeline
[
  { "year": "1972", "title": "Biological Weapons Convention", "text": "The first treaty banning an entire class of weapons opens for signature. It contains no verification mechanism, and the Soviet Union ran a large offensive programme while a party to it." },
  { "year": "1979", "title": "Sverdlovsk", "text": "An aerosol of anthrax spores escapes from a Soviet military microbiology facility, killing at least 66 people. The Soviet government attributed the deaths to contaminated meat until 1992." },
  { "year": "1984", "title": "The Dalles", "text": "Members of an Oregon commune contaminate restaurant salad bars with Salmonella, sickening at least 751 people. It remains the largest bioterrorist attack in US history and used a food-borne organism, not a transmissible one." },
  { "year": "2001", "title": "Anthrax letters", "text": "Spores mailed within the United States kill five people and infect seventeen. The FBI concluded the material originated in a US biodefence laboratory." },
  { "year": "2002", "title": "Poliovirus assembled from mail-order DNA", "text": "A team reconstructs infectious poliovirus from published sequence and commercially ordered fragments, showing a virus can be rebuilt without a natural sample." },
  { "year": "2011–2012", "title": "H5N1 transmissibility", "text": "Ferret-transmissible avian influenza variants are made in two laboratories. A US advisory board recommends redacting the methods, then reverses; the papers appear in full." },
  { "year": "2018", "title": "Horsepox", "text": "A Canadian group synthesises horsepox virus, a relative of smallpox, from chemically synthesised DNA fragments at modest cost." },
  { "year": "2025", "title": "US pause on dangerous gain-of-function work", "text": "An executive order halts federally funded research that would make a pathogen more transmissible or harder to counter, superseding the previous year's oversight policy." }
]
```

Two patterns run through this record. The deadly events came from state facilities: the Sverdlovsk aerosol killed at least 66 people,[^meselson1994] and the 2001 letters used material the FBI traced to a US biodefence laboratory. The largest non-state attack on record made at least 751 people ill and killed nobody,[^torok1997] and the other serious attempt was technically limited: the Aum Shinrikyo cult sprayed an anthrax preparation over Tokyo without producing a single case, having cultured a vaccine strain. Against that, the capability those attempts lacked is exactly the capability that has since become cheap.

## The routes

**State programmes.** These have produced the only demonstrated mass-casualty capability. The Biological Weapons Convention prohibits them,[^bwc1972] and the Soviet Union ran one of the largest on record while a party to it, which is documented history rather than inference.

**Non-state actors.** The historical base rate of success is zero and the historical base rate of attempts is not. Analysts who take this route seriously argue that the relevant question is not how many groups want to cause mass casualties but how much skill the least skilled of them now needs. The amateur culture surveyed under [[biohacking]] is usually invoked here, although its documented output is closer to insulin production and self-administered [[crispr-cas9|CRISPR]] constructs than to anything pathogenic.

**Laboratory accident.** Laboratory-acquired infections are documented regularly, and the origin of the COVID-19 pandemic remains genuinely unresolved in mainstream scientific opinion, with both zoonotic and research-related hypotheses still argued by serious people. Whichever is correct, the dispute itself has done more to shape biosafety policy than any deliberate-release scenario.

**Research on enhanced pathogens.** The category the H5N1 episode created, covered in full under [[dual-use-research]], is the one place where the risk is generated by people trying to reduce it.

**Assistance from AI systems.** Students without laboratory training, working with chatbots for an hour, were able to elicit candidate pandemic pathogens, reverse-genetics routes to obtaining them, and the names of synthesis providers unlikely to screen orders.[^soice2023] A red-team study run at RAND the following year found no statistically significant difference in the viability of biological attack plans produced with and without large language models.[^mouton2024] Both results are usually cited as though they settle the question; neither does. The design capability demonstrated in protein and sequence design work such as [[ai-protein-design]] is a separate concern from the tutoring capability these two studies measured.

## Plausibility

There is no frequency to extrapolate from, so every estimate is a judgement. Toby Ord's stated credence in *The Precipice* puts engineered pandemics among the largest anthropogenic existential risks of this century, above natural pandemics by a wide margin, and he presents the figure as his own credence rather than a calculation.[^ord2020] Kevin Esvelt has argued that within about a decade tens of thousands of people will have the knowledge to cause a pandemic single-handedly, which is a claim about the diffusion of capability rather than about intent.[^esvelt2022] Neither figure is a measurement, and the wider field of [[existential-risk]] analysis contains no agreed method for producing one. What can be said without a number is that the risk sits in the same structural class as the self-spreading constructs of [[gene-drive|gene drives]] and the hypothetical organisms of [[mirror-life]], and that it is the closest real analogue to the hypothetical [[nick-bostrom|Bostrom]] uses in the vulnerable world hypothesis, a technology cheap enough to be widely available and destructive enough that a single user matters.

> [!debate] Is capability the binding constraint?
> One camp holds that tacit laboratory skill, working materials, and the difficulty of aerosolisation remain formidable, and points to the RAND null result and to the consistent failure of non-state attempts. The other holds that these barriers are exactly the ones automation, protocol publication, and contract research organisations are dismantling, and that a red-team exercise with a fixed scenario cannot measure a capability that has not been reached yet. The disagreement is about the shape of a trend, not about any observed event, and it is unlikely to be resolved by evidence before the fact.

## Mitigation

The standard framing separates three lines of defence. **Delay** covers everything that slows the spread of capability: screening synthetic DNA orders against hazardous sequences, which is voluntary in most jurisdictions and performed by a subset of providers; withholding methods that would enhance transmissibility; and export controls. **Detect** covers surveillance designed to identify an unknown agent rather than a known one, principally metagenomic sequencing of wastewater and travel hubs, on the reasoning that a stealth pathogen would be recognised too late by clinical reporting. **Defend** covers stockpiled protective equipment, air treatment, and rapid countermeasure platforms.[^esvelt2022]

The third line is where the most has changed. The vaccine timeline achieved in 2020 using [[lipid-nanoparticles|lipid nanoparticle]] delivery was unprecedented and would still have been far too slow against a pathogen with a long presymptomatic phase. That asymmetry is the argument for weighting detection and physical protection over medical countermeasures, and it is a rare case where the cheapest interventions are also the most robust.

## Governance

The Biological Weapons Convention prohibits development, production and stockpiling, has broad membership, and cannot verify anything. Six years of negotiation toward a verification protocol collapsed in 2001 when the United States rejected the draft, and no binding successor has been attempted since. What exists instead is a patchwork of national oversight policies, institutional review, journal norms, and voluntary industry screening, none of which binds a determined state and all of which depend on the compliance of the people they govern. It is a weaker regime than the one described under [[governance-of-genome-editing]], and for the same underlying reason: the equipment is ordinary and the jurisdictions are many. Engineered biocontainment of the kind proposed for [[recoded-organisms]] is one of the few technical rather than legal answers on offer, and it applies to laboratory strains rather than to a pathogen someone intends to release.

The United States has moved twice in quick succession: a unified oversight policy for dual-use research and pandemic-potential pathogens in 2024, then Executive Order 14292 in 2025, which paused federally funded work meeting a definition of dangerous gain-of-function research and directed a replacement policy.[^eo14292] The pattern of rapid reversal is itself the governance problem in miniature. A field in which the rules change with an administration cannot support the long-term coordination that the [[differential-technological-development|sequencing argument]] and the [[asilomar-conference|Asilomar]] precedent both assume, and unilateral restriction moves work rather than stopping it.

## Open problems

Attribution is the deepest one. A deliberate release that resembled a natural spillover would be difficult to distinguish from one, which weakens deterrence in a way that has no analogue in nuclear security. Openness is the second: nearly every measure that reduces misuse risk also slows the research that produces countermeasures, and the [[precautionary-principle|precautionary]] framing does not resolve the trade because both sides of it are safety arguments. The third is unglamorous. The mitigations with the best expected value are stockpiles, ventilation, and sequencing infrastructure, none of which produces a publication or a product, and all of which are funded accordingly.

## See also

- [[dual-use-research]]
- [[existential-risk]]
- [[mirror-life]]
- [[gene-drive]]
- [[differential-technological-development]]
- [[immunosenescence]]
- [[synthetic-genomes]]
- [[asilomar-conference]]

## References

[^herfst2012]: `paper` Herfst, S. et al. "Airborne Transmission of Influenza A/H5N1 Virus Between Ferrets." *Science*, 2012. {Transmissibility was measured in ferrets, the standard influenza model; none of the recipient animals died after airborne infection.}
[^meselson1994]: `paper` Meselson, M., Guillemin, J., Hugh-Jones, M. et al. "The Sverdlovsk Anthrax Outbreak of 1979." *Science*, 1994. {The reconstruction was made from case locations and wind records thirteen years after the event, once Russian authorities permitted the investigation.}
[^torok1997]: `paper` Török, T. J. et al. "A large community outbreak of salmonellosis caused by intentional contamination of restaurant salad bars." *JAMA*, 1997. {The intentional cause was established by a criminal investigation, not by the epidemiology, which is the point the paper is usually cited for.}
[^bwc1972]: `law` Convention on the Prohibition of the Development, Production and Stockpiling of Bacteriological (Biological) and Toxin Weapons and on Their Destruction. Opened for signature 1972; entered into force 1975.
[^soice2023]: `preprint` Soice, E. H. et al. "Can large language models democratize access to dual-use biotechnology?" arXiv preprint, 2023. {A classroom exercise with no controls and no attempt to obtain any agent; it records what a chatbot said, not what a student could do.}
[^mouton2024]: `report` Mouton, C. A., Lucas, C. and Guest, E. *The Operational Risks of AI in Large-Scale Biological Attacks: Results of a Red-Team Study*. RAND Corporation, 2024. {Tested the models available at the time against fixed scenarios; a null result on those models is not a null result on later ones.}
[^ord2020]: `book` Ord, T. *The Precipice: Existential Risk and the Future of Humanity*. Bloomsbury, 2020. {A trade book for a general readership; the figures are one author's stated credences, not a peer-reviewed risk assessment.}
[^esvelt2022]: `report` Esvelt, K. M. *Delay, Detect, Defend: Preparing for a Future in which Thousands Can Release New Pandemics*. Geneva Centre for Security Policy, Geneva Paper 29, 2022. {A policy paper by an author who also runs a synthesis-screening project, arguing for measures his group works on.}
[^eo14292]: `law` Executive Order 14292, "Improving the Safety and Security of Biological Research." United States, May 2025.

==============================================================================
ARTICLE: sleep-reduction
TITLE:   Engineered sleep reduction
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/sleep-reduction
SOURCE:  https://futurehumanwiki.com/raw/sleep-reduction
==============================================================================

---
title: "Engineered sleep reduction"
slug: "sleep-reduction"
type: "concept"
status: "speculative"
horizon: "2040s"
categories: ["enhancement"]
tags: ["sleep", "enhancement", "genetics", "pharmacology", "cognition", "circadian"]
summary: "The proposal to shorten the amount of sleep a person needs without incurring the costs of sleep deprivation, based on rare short-sleep gene variants and wake-promoting drugs."
updated: "2026-07-27"
humanEvidence: "No human demonstration exists. The evidence is rare families carrying short-sleep variants, described in small pedigrees with self-report and limited cognitive testing, plus mouse models of those variants."
issues: ["The 2025 phase 3 orexin agonist result in narcolepsy is stated without a source."]
---

```infobox
{
  "caption": "Proposed form of human enhancement",
  "rows": [
    { "label": "Target", "value": "Sleep need, not sleepiness" },
    { "label": "Natural precedent", "value": "Familial natural short sleep" },
    { "label": "Genes implicated", "value": "DEC2, ADRB1, NPSR1" },
    { "label": "Drug routes", "value": "Wake-promoting agents, orexin" },
    { "label": "Human demonstration", "value": "None" },
    { "label": "Main obstacle", "value": "Sleep's functions are plural" }
  ]
}
```

**Engineered sleep reduction** is the proposal to lower how much sleep a person requires, rather than merely suppressing the feeling of sleepiness, so that waking hours increase without the deficits that follow sleep loss. It is one of the few enhancement targets with a natural existence proof: a small number of families carry variants that let them sleep four to six hours a night, apparently without impairment. Whether that biology can be transferred to anyone else, and whether the apparent absence of cost survives careful measurement, are both unresolved.

```keyfacts
[
  { "value": "≥3", "label": "Genes linked to familial short sleep", "note": "DEC2/BHLHE41, ADRB1 and NPSR1, with further candidates reported" },
  { "value": "4–6 h", "label": "Typical night in natural short sleepers", "note": "with no reported daytime sleepiness and no weekend rebound" },
  { "value": "~2 h", "label": "Daily sleep in wild African elephants", "note": "the least measured in any land mammal, showing how far need can vary across species" }
]
```

## Sleepiness is not sleep need

The distinction that organises the subject is between the drive to sleep and the functions sleep performs. Caffeine, modafinil and amphetamines act on the first. They suppress the subjective and, to a degree, the performance consequences of being awake too long, and they do not discharge the underlying pressure: the sleep that is lost is still owed, and slow-wave rebound follows when the drug wears off. A drug that removes sleepiness without removing sleep need is a stimulant, not an enhancement.

That makes the functions the real target, and they are plural. Sleep is implicated in synaptic downscaling, in memory consolidation, in metabolic and hormonal regulation, in immune function, and in clearance of interstitial waste from the brain, which ties it to the aggregation problems described in [[proteostasis]]. The clearance result was shown in mice, where interstitial space expands during sleep and removal of solutes including amyloid-beta increases.[^xie2013] In humans, a single night of deprivation increased amyloid-beta signal in the hippocampus and thalamus on positron emission tomography.[^shokri2018] An intervention that shortened sleep would have to leave every one of these functions intact, and there is no reason to expect a single lever that does so.

## Natural short sleep

The most informative evidence comes from families identified by Ying-Hui Fu and Louis Ptáček's group and others.

A mutation in the transcriptional repressor DEC2, also called BHLHE41, was found in a mother and daughter who slept about two hours less than average and reported no ill effects. Mice and flies carrying the variant slept less, which established that the effect was causal rather than a family habit.[^he2009] A variant in the beta-1 adrenergic receptor ADRB1 was later identified in a family averaging under six hours, again with a mouse model showing reduced sleep and increased activity of wake-promoting neurons in the dorsal pons.[^shi2019] A third, in the neuropeptide S receptor NPSR1, produced short sleep in mice that were also unusually resistant to the memory deficits normally caused by sleep deprivation.[^xing2019] Additional candidate genes have been reported since. The pattern is the one familiar from other complex traits: rare variants of large effect in a handful of pedigrees, alongside common variation of small effect that would be useless for [[embryo-selection|selection]].

Two features of this literature matter for the enhancement question. Familial natural short sleep is rare, and most people who describe themselves as short sleepers are chronically restricted rather than genuinely short-sleeping: they carry sleep debt, show weekend rebound, and perform worse on objective tests than they believe. True short sleepers show neither. And in mouse models of neurodegeneration, short-sleep variants have been reported to reduce rather than increase pathology, which cuts against the simple assumption that less sleep means less clearance.[^dong2022] That result is in mice, in engineered disease models, and has not been examined in the human carriers.

> [!caution] What the human evidence actually covers
> The claim that natural short sleepers suffer no cost rests on small numbers of people, self-report, and limited cognitive testing over short periods. No cohort has been followed for decades with the outcome measures — dementia incidence, cardiovascular events, mortality — that would settle it.

## Pharmacological routes

Wake-promoting drugs are the only pharmacology in clinical use, and they do not address sleep need. Modafinil and armodafinil, solriamfetol and the histamine H3 antagonist pitolisant are licensed for excessive daytime sleepiness in narcolepsy and related conditions, and their off-label use for cognitive purposes is covered in [[nootropics]].

Orexin is the most-discussed prospective target. Narcolepsy type 1 results from loss of the hypothalamic neurons that produce orexin, and orexin receptor agonists have been developed to replace the missing signal; a phase 3 programme reported positive results for one such agent in narcolepsy in 2025. This is replacement therapy in people who lack the peptide. Whether raising orexin signalling in someone with an intact system would reduce sleep need, rather than simply forcing wakefulness with an accumulating debt, is untested and not the indication these drugs are being developed for.

Deliberate torpor is a separate proposal with a separate literature, treated in [[human-hibernation]], and motivated mainly by long-duration spaceflight rather than by productivity; the physiological burden of such missions is surveyed in [[space-medicine]]. Genetic routes remain hypothetical. Introducing a short-sleep variant would require germline modification, which is prohibited for clinical use in most jurisdictions and would be a paradigm case of enhancement rather than therapy; see [[germline-editing]] and [[human-enhancement]]. Somatic delivery to the relevant hypothalamic and brainstem populations is beyond current [[aav-vectors|vector]] targeting.

## The cost of chronic restriction

The best-characterised experiment in this area is the opposite of enhancement. Van Dongen and colleagues restricted volunteers to four, six or eight hours in bed for two weeks with objective performance testing. Deficits accumulated across days in a dose-dependent way in both restricted groups, and — the finding that matters most here — subjective sleepiness plateaued while objective lapses of attention kept getting worse.[^vandongen2003] People adapt to how sleep restriction feels, not to what it does.

Epidemiology adds a weaker but consistent signal. Meta-analyses find a [[sleep-and-longevity|U-shaped association]] between habitual sleep duration and all-cause mortality, with elevated risk at both short and long durations.[^cappuccio2010] The association is confounded in obvious ways: illness shortens and lengthens sleep, and self-reported duration is unreliable. It is nonetheless the only population-scale evidence available, and it does not favour deliberate reduction. Sleep duration also correlates with several of the [[hallmarks-of-aging]] and with the functional measures used to define [[healthspan]], which means an intervention that shortened sleep would have to be assessed on outcomes the field has not yet learned to measure over short horizons.

Polyphasic schedules — dividing sleep into short naps to reduce the total — have circulated for decades with no controlled evidence of benefit. A consensus review by sleep researchers found the practice associated with adverse outcomes and no demonstrated advantage.[^weaver2021]

## Is sleep need compressible

Comparative biology shows enormous variation in sleep duration across mammals, from around two hours a day in wild African elephants to nineteen in some bats, which demonstrates that the requirement is not a fixed physical constant.[^gravett2017] It does not show that the human requirement can be moved. Species differences reflect whole-organism differences in metabolic rate, brain organisation and predation ecology, not a dial that evolution set arbitrarily.

The strongest version of the case for compressibility is the short-sleep families: within a single species, a single-gene change produces a two-hour reduction with no obvious penalty. The strongest version of the case against is that the reduction is small, that the carriers were not selected for having no deficit but for reporting none, and that a mechanism which produces two hours may not scale to four. As with [[exercise-and-aging|exercise]], the unglamorous comparator sets a high bar: adequate sleep is free, and no intervention in this area has been shown to beat it.

> [!debate] Two readings of the same data
> Enhancement advocates read familial short sleep as proof that sleep need is a modifiable set-point with unused headroom. Sleep researchers more often read it as proof that the set-point is under tight genetic control and that a rare variant shifts it slightly, in the way rare variants shift height — which would mean the trait is heritable and not, in any practical sense, engineerable.

## Outlook

The near-term work that would matter is unglamorous: longitudinal follow-up of natural short sleepers with imaging and cognitive endpoints, to establish whether the absence of cost is real over decades. Without that, the entire case for the target rests on an unmeasured assumption.

If the assumption holds, the intervention still requires acting on hypothalamic circuitry in healthy adults, which no regulator would approve on an enhancement indication and no delivery technology can currently achieve. The ethical objections would then be the familiar ones catalogued in [[bioethics-of-enhancement]], with one addition specific to this target: waking hours are a competitive resource, so a workable intervention would generate the coercion structure described in [[enhancement-arms-race]] faster than almost any other enhancement. If it does not hold — if the short sleepers turn out to be paying in outcomes nobody has looked at — then the field's existence proof disappears and sleep reduction becomes what most sleep scientists already consider it: a rebranding of sleep deprivation.

## See also

- [[human-enhancement]]
- [[nootropics]]
- [[intelligence-amplification]]
- [[germline-editing]]
- [[genetic-enhancement-of-intelligence]]
- [[human-hibernation]]
- [[exercise-and-aging]]
- [[bioethics-of-enhancement]]

## References

[^he2009]: `paper` He, Y. et al. "The Transcriptional Repressor DEC2 Regulates Sleep Length in Mammals." *Science*, 2009.
[^shi2019]: `paper` Shi, G. et al. "A Rare Mutation of β1-Adrenergic Receptor Affects Sleep/Wake Behaviors." *Neuron*, 2019.
[^xing2019]: `paper` Xing, L. et al. "Mutant neuropeptide S receptor reduces sleep duration with preserved memory consolidation." *Science Translational Medicine*, 2019.
[^dong2022]: `paper` Dong, Q. et al. "Familial natural short sleep mutations reduce Alzheimer pathology in mice." *iScience*, 2022.
[^xie2013]: `paper` Xie, L. et al. "Sleep Drives Metabolite Clearance from the Adult Brain." *Science*, 2013.
[^shokri2018]: `paper` Shokri-Kojori, E. et al. "β-Amyloid accumulation in the human brain after one night of sleep deprivation." *Proceedings of the National Academy of Sciences*, 2018. {A single night of total deprivation in healthy adults, measured by PET; it shows acute accumulation and says nothing about habitual short sleep.}
[^vandongen2003]: `paper` Van Dongen, H. P. A., Maislin, G., Mullington, J. M. and Dinges, D. F. "The cumulative cost of additional wakefulness: dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation." *Sleep*, 2003. {Healthy adults restricted in a laboratory for two weeks; it measures the cost of imposed restriction, not the experience of people who sleep little by nature.}
[^cappuccio2010]: `paper` Cappuccio, F. P., D'Elia, L., Strazzullo, P. and Miller, M. A. "Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies." *Sleep*, 2010.
[^weaver2021]: `paper` Weaver, M. D. et al. "Adverse impact of polyphasic sleep patterns in humans: Report of the National Sleep Foundation sleep timing and variability consensus panel." *Sleep Health*, 2021.
[^gravett2017]: `paper` Gravett, N. et al. "Inactivity/sleep in two wild free-roaming African elephant matriarchs." *PLOS ONE*, 2017. {Two animals, with sleep inferred from trunk actigraphy rather than from EEG, so the two-hour figure is an activity-based estimate.}

==============================================================================
ARTICLE: enhancement-arms-race
TITLE:   Enhancement arms race
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/enhancement-arms-race
SOURCE:  https://futurehumanwiki.com/raw/enhancement-arms-race
==============================================================================

---
title: "Enhancement arms race"
slug: "enhancement-arms-race"
type: "concept"
status: "contested"
horizon: "2030s"
categories: ["society", "enhancement"]
tags: ["enhancement", "coercion", "sport", "military", "economics", "policy"]
summary: "The argument that competitive pressure makes enhancement effectively compulsory, so that individually rational adoption leaves everyone worse off."
updated: "2026-07-27"
issues: ["The 2020 op-ed claim about a foreign military enhancement programme carries no citation.", "The military section states the amphetamine and friendly-fire history without sources."]
---

```infobox
{
  "caption": "Collective-action problem",
  "rows": [
    { "label": "Structure", "value": "Prisoner's dilemma" },
    { "label": "Underlying concept", "value": "Positional goods" },
    { "label": "Named by", "value": "Fred Hirsch, 1976" },
    { "label": "Worked example", "value": "Anti-doping in sport", "link": "/wiki/enhancement-in-sport" },
    { "label": "Standard remedies", "value": "Ban, disclose, or universalise" },
    { "label": "Current status", "value": "Real in sport, hypothetical elsewhere" }
  ]
}
```

**Enhancement arms race** describes a situation in which the benefit of an enhancement depends on having it when others do not, so that each person's rational decision to adopt it removes the advantage while leaving everyone with the costs. The argument is the strongest objection to enhancement that does not depend on any claim about human nature, dignity, or the value of the given. It concedes that the technology works, that people choose it freely, and that each choice is individually sensible — and shows that the outcome can still be worse for everyone.

## Positional and absolute goods

The distinction comes from Fred Hirsch, who observed that some goods are valuable in themselves and others only in relation to what others have.[^hirsch1976] Clean water is absolute: everyone having it is better than some people having it. A front-row seat is positional: its value consists in others not having it. Robert Frank extended the analysis to biological competition, noting that traits selected for relative advantage — antler size is his standing example — impose costs on every member of the population without improving any individual's rank.[^frank2011]

The modalities catalogued in [[human-enhancement]] divide along this axis, and the division determines what policy is appropriate.

**Absolute.** Resistance to infection, repair of damaged tissue, freedom from pain. Universal adoption makes everyone better off, and for infectious disease the externality runs positive — one person's immunity protects others. Nobody argues for restricting these, which is why the therapy side of [[bioethics-of-enhancement]] is comparatively uncontested.

**Positional.** Height in a society that rewards it, competitive ranking, scarce admissions places. Universal adoption produces no aggregate benefit and may produce aggregate cost.

**Mixed.** Cognition is the interesting case, and both camps claim it. It is positional in selection contexts — a fixed number of medical school places will be filled whatever the applicants' capabilities. It is absolute wherever thinking produces something others can use, and the historical spillovers from literacy, numeracy, and public education were enormous. Which component dominates for a marginal increase in a wealthy society is not known.

## The structure of the trap

For a purely positional enhancement with real costs, the payoff matrix is a prisoner's dilemma. If nobody enhances, the ranking is unchanged and nobody bears the cost. If everybody enhances, the ranking is unchanged and everybody bears the cost. If one enhances and others do not, that one gains. Defection dominates, and the equilibrium is universal adoption with no benefit.

The mechanism does not require anyone to be forced. It is the reason "voluntary" is a weak defence of enhancement in competitive settings: the availability of an option changes the situation of people who decline it, a form of pressure that liberal frameworks based on individual consent handle poorly. It also cuts against the standard fairness objection developed in [[access-and-inequality]] — the arms-race problem is worst precisely when access is *equal*, because then everyone adopts.

> [!key] Two different objections
> Unequal access to enhancement produces unfair advantage. Equal access to a positional enhancement produces universal cost with no advantage. Policies that fix the first can make the second worse, which is why "just make it available to everyone" is not a general solution.

## The worked example: sport

Athletic competition is a pure positional contest with an explicit ranking, which is why it is the only domain where the dynamic has been observed at scale and the only one with a mature regulatory response. The anti-doping regime described in [[enhancement-in-sport]] exists because athletes in several sports demonstrated in the 1970s and 1980s that in the absence of enforcement, the equilibrium is universal use of drugs with serious health costs and unchanged relative standings.

The regime's rationale is explicitly collective: it is not that doping is unnatural but that a testable prohibition is the only way to reach the outcome most athletes say they prefer. Whether it works is disputed — detection lags the pharmacology, and [[gene-doping]] would be harder to detect than any small molecule, though no confirmed case has been documented. The Enhanced Games, an event announced in 2023 and scheduled to run in 2026, proposes the opposite equilibrium: permit everything, disclose everything, and treat the resulting performances as a separate category. It is the first serious attempt to test whether an unregulated equilibrium is stable, and the medical objections are substantial.

## The military case

Militaries face the arms-race structure directly and have no international body corresponding to the anti-doping agency. Amphetamines have been issued to combat pilots for decades to sustain performance on long missions, a practice that surfaced publicly in the litigation following a 2002 friendly-fire incident in Afghanistan. US defence research agencies have funded programmes on sustained performance without sleep, on physical endurance, and on non-surgical neural interfaces intended for able-bodied operators rather than patients.

Claims about foreign programmes have been used to justify domestic ones. A senior US intelligence official asserted in a 2020 newspaper op-ed that China had conducted human testing on military personnel with the aim of producing biologically enhanced soldiers; no supporting evidence was published then or since. The pattern is familiar from earlier arms races: unverifiable claims about an adversary's capability are themselves an input to the race, and are hard to distinguish from advocacy.

The military domain is where the tools discussed in [[non-invasive-neuromodulation]], [[exoskeleton]], [[myostatin-inhibition]], and [[sleep-reduction]] research are most likely to be adopted first, because the institution can order adoption and internalise the costs. It is also where the liberty framing of [[morphological-freedom]] has least purchase, since the person modified does not choose.

## The civilian case is weaker than usually assumed

Outside sport and the military, the arms-race argument runs into an empirical obstacle: the enhancements are not good enough to race over.

Surveys of stimulant use among students and academics — including an informal poll of readers conducted by a scientific journal in 2008 that found roughly a fifth of respondents reporting non-medical use for cognitive purposes — indicate that people believe otherwise.[^maher2008] The evidence reviewed in [[nootropics]] does not support them. Prescription stimulants and modafinil produce reliable effects on wakefulness and small, inconsistent effects on cognition in rested healthy adults, with some studies finding that they increase effort and confidence more than accuracy. An arms race over an ineffective enhancement wastes resources without changing outcomes, which is a real cost but a different problem.

The genuine civilian enhancement arms race is educational and financial: tutoring, test preparation, and school selection, which have measurable effects, are purchased overwhelmingly by the wealthy, and are almost entirely unregulated. Biological enhancement is discussed as though it would introduce a dynamic that has been operating for a century.

The nearest biological candidate is reproductive rather than pharmacological. [[polygenic-embryo-screening]] is purchasable, unregulated in most jurisdictions, and confers whatever advantage it confers before the child can consent — the structure of a positional race with the participants selected in advance, which is the version the ethics literature is least equipped to address.

> [!debate] Whether the race is winnable
> Some argue that if enhancement is coming regardless, a society should provide it universally and cheaply, converting a positional race into a level baseline. Others reply that this simply relocates the race to whatever remains scarce, and that the terminal state of a fully enhanced population is the same ranking with higher costs.

## Regulatory equilibria

Four responses have been proposed, and each has a domain where it works.

**Prohibition with testing** suits closed competitive systems with an enforcement body — sport, and in principle competitive examinations. It fails where detection is infeasible.

**Mandatory disclosure** preserves choice and lets others adjust. It works badly when disclosure is itself stigmatising.

**Universal provision** removes the inequality but not the race, and is the right answer for enhancements with strong absolute components.

**Redesigning the contest** is the least discussed and possibly the most effective: changing what is being competed for so that the enhanced trait no longer determines the outcome. Weight classes and paralympic classification do this in sport; admissions lotteries above a threshold do it in education.

Which of these applies to any given enhancement depends on the positional-versus-absolute question, and that question has not been answered empirically for a single candidate technology — including the one, [[genetic-enhancement-of-intelligence]], around which most of the argument has been conducted.

## See also

- [[enhancement-in-sport]]
- [[human-enhancement]]
- [[bioethics-of-enhancement]]
- [[access-and-inequality]]
- [[nootropics]]
- [[gene-doping]]
- [[morphological-freedom]]
- [[moral-enhancement]]

## References

[^hirsch1976]: `book` Hirsch, F. *Social Limits to Growth*. Harvard University Press, 1976. {An economic argument about growth and consumption; it predates the enhancement debate and does not discuss biological modification.}
[^frank2011]: `book` Frank, R.H. *The Darwin Economy: Liberty, Competition, and the Common Good*. Princeton University Press, 2011.
[^maher2008]: `paper` Maher, B. "Poll results: look who's doping." *Nature*, 2008. {An informal poll of self-selected journal readers with no defined sampling frame, reported in a news section rather than as research.}

==============================================================================
ARTICLE: enhancement-in-sport
TITLE:   Enhancement in sport
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/enhancement-in-sport
SOURCE:  https://futurehumanwiki.com/raw/enhancement-in-sport
==============================================================================

---
title: "Enhancement in sport"
slug: "enhancement-in-sport"
type: "concept"
status: "established"
horizon: "present"
categories: ["enhancement", "society"]
tags: ["sport", "doping", "regulation", "fairness", "prosthetics", "eligibility"]
summary: "The regulation of performance enhancement in competition, the only domain that draws an operational line between therapy and enhancement and litigates it case by case."
updated: "2026-07-27"
issues: ["The Code's listing criteria and the 2009 biological passport date carry no citation.", "The 2016 exemption-record leak is described without a source."]
---

```infobox
{
  "caption": "Regulatory and ethical domain",
  "rows": [
    { "label": "Global regulator", "value": "World Anti-Doping Agency" },
    { "label": "Founded", "value": "1999" },
    { "label": "Governing instrument", "value": "World Anti-Doping Code" },
    { "label": "Liability standard", "value": "Strict liability" },
    { "label": "Listing test", "value": "Two of three criteria" },
    { "label": "Principal challenge", "value": "Open-enhancement competitions" }
  ]
}
```

**Enhancement in sport** is the use of substances, methods and equipment to raise athletic performance, and the regulatory apparatus built to prohibit some of it. Sport is unusual among human activities in maintaining a formal, enforced boundary between restoring health and exceeding it, complete with a list of banned agents, a testing infrastructure, an appeals court, and a procedure for granting exemptions. That machinery makes it the best available case study for what happens when the therapy–enhancement distinction described in [[human-enhancement]] is turned into administrative law.

```keyfacts
[
  { "value": "1999", "label": "World Anti-Doping Agency founded", "note": "following the 1998 Festina affair at the Tour de France" },
  { "value": "2 of 3", "label": "Criteria for adding to the Prohibited List", "note": "performance enhancement, health risk, contrary to the spirit of sport" },
  { "value": "2009", "label": "Athlete Biological Passport introduced", "note": "longitudinal profiling rather than detection of a specific substance" }
]
```

## The regime

Anti-doping predates the World Anti-Doping Agency by decades, but the modern system dates from the 1998 Tour de France, when a team car full of doping products was intercepted and the resulting scandal made clear that federations could not police themselves. WADA was established the following year as a joint initiative of the Olympic movement and governments, and the World Anti-Doping Code came into force in 2004. It has been revised several times since.

Its central features are an annually revised Prohibited List, accredited laboratories, out-of-competition testing with whereabouts requirements, and strict liability: an athlete is responsible for what is in their body regardless of intent, subject to reduced sanction where fault is minimal — the provision that governs the recurring disputes over contaminated [[dietary-supplements|supplements]]. Disputes go to the Court of Arbitration for Sport.

Detection has shifted from finding a molecule to characterising a person. The Athlete Biological Passport, introduced for blood variables in 2009 and extended to steroid profiles, tracks each athlete against their own longitudinal baseline and flags deviations whose cause need not be identified. This is the model anti-doping science expects to rely on as agents become harder to detect directly, and it is the approach discussed in [[gene-doping]] for interventions that leave no foreign molecule at all.

## What is prohibited, and why

A substance or method is added to the Prohibited List if it meets two of three criteria: it enhances or could enhance performance; it poses an actual or potential health risk; and it violates the spirit of sport. Masking agents are included regardless. The list therefore mixes categories that have little in common pharmacologically: anabolic agents and the muscle-directed constructs discussed in [[myostatin-inhibition]], hormones, diuretics, and the stimulants and wakefulness drugs covered in [[nootropics]].

Each criterion carries a different justification and each is contested.

**Health.** The paternalism objection is answered by the coercion argument: in a competitive setting, one athlete's use imposes a choice on everyone else, so protecting health is protecting against a collective-action failure rather than against individual foolishness. The historical case that grounds this is East Germany's state doping programme, which administered androgens to athletes including minors and produced documented long-term harm. The structure is the one set out in [[enhancement-arms-race]].

**Fairness.** The difficulty is that sport does not equalise anything else. Athletes differ in genes, altitude of birthplace, national funding and access to coaching, and none of this is levelled. The fairness argument therefore has to explain why pharmacological advantage is different from every other kind, which it does mainly by pointing to concealment and to the absence of consent.

**Spirit of sport.** The Code lists values — health, excellence, character, respect for rules, courage, community — and the criterion has been criticised as the point at which the system stops giving reasons. It is nonetheless the criterion that decides the hard cases, since many prohibited substances would fail the first two tests on their own.

> [!note] The list is not a principle
> Caffeine was prohibited until 2004 and is now merely monitored. Hypoxic tents, which raise red-cell mass by the same physiological route as a banned hormone, were examined by WADA's ethics panel in 2006 and judged contrary to the spirit of sport, and were then not prohibited. The list records a series of negotiated decisions rather than the application of a consistent rule.

## Therapeutic use exemptions

The therapeutic-use exemption is where the therapy–enhancement line becomes an actual decision by an actual committee. An athlete may use a prohibited substance if withholding it would cause significant health impairment, if it produces no enhancement beyond a return to normal health, and if no permitted alternative exists.

The middle condition is the hard one. Restoring an asthmatic athlete's lung function to normal is by construction a performance improvement relative to their untreated state, and there is no non-arbitrary way to identify the baseline the exemption is supposed to return them to. The 2016 publication of stolen exemption records, showing widespread and lawful use of powerful medications by prominent athletes, made the tension public without producing a resolution. The system's defence is that the alternative — refusing treatment to sick athletes — is worse, which is true and is not an answer to the conceptual objection.

## Equipment and bodies

Governing bodies regulate equipment on a logic that parallels doping and produces the same inconsistencies. Polyurethane swimsuits that reduced drag and improved buoyancy were permitted through a period of extraordinary record-setting in 2008 and 2009, then banned. Distance-running shoes with thick compliant midsoles and embedded plates measurably improved running economy and were restricted by a stack-height rule rather than prohibited. Neither decision was framed as protecting health. Powered assistance of the kind described in [[exoskeleton]] has never needed a rule, because no federation has had to consider admitting it.

Prosthetic athletes force the question directly. When Oscar Pistorius, a double transtibial amputee, sought to compete against non-amputees, athletics' governing body ruled him ineligible on the basis of a biomechanical study finding a mechanical advantage from his carbon-fibre blades; the Court of Arbitration for Sport overturned the ruling in 2008, holding that an overall advantage had not been proved, and he competed at the 2012 Olympics.[^cas2008] The scientists involved subsequently disagreed in print about what the measurements showed, since the blades are lighter and return energy more efficiently than a biological limb while providing far less force at the start and less stability.[^weyand2009] The unresolved case is the long jumper Markus Rehm, whose distances are competitive at the highest level and who has not been able to satisfy the requirement to prove no advantage — a burden of proof that, applied symmetrically, no athlete could meet.

The prosthetics themselves are covered in [[myoelectric-prosthetics]] and [[osseointegration]]; the relevant point here is that a device classed as a medical aid in one setting is classed as potential enhancement in another, with no change to the device. Paralympic classification faces the mirror-image problem of grouping impairments finely enough to make competition meaningful, and the disputes it generates are examined in [[disability-rights-and-enhancement]].

## Eligibility and the body's own hormones

Regulations on athletes with differences of sex development require testosterone suppression for eligibility in some women's events, on the argument that endogenous androgen levels in a typical male range confer a performance advantage that categories exist to control. Caster Semenya challenged the rules through arbitration and the Swiss courts without success, and then before the European Court of Human Rights, where a chamber found in her favour in 2023 and the Grand Chamber upheld part of her case in 2025, on the adequacy of the judicial review she had received in Switzerland rather than on the science. Neither judgment struck down the eligibility rules, which remain in force.[^echr2025]

The case is the sharpest illustration of the field's central difficulty. Every other criterion in anti-doping concerns something an athlete did. This one concerns something an athlete is, and requires a medical intervention to remove a natural characteristic in order to compete — the inverse of the therapy–enhancement relation everywhere else in the Code.

## The open-enhancement challenge

The Enhanced Games, announced in 2023 and backed by private investors, proposes competition with no anti-doping rules and medical supervision instead, on the argument that prohibition drives use underground and makes it more dangerous. An inaugural event was scheduled for Las Vegas in mid-2026; its conduct and outcome fall outside what this article can report. International federations have condemned the venture, and world governing bodies in at least one sport have moved to bar participants from their own competitions.

The argument deserves a serious answer rather than dismissal. Its weakest point is that supervision does not make unapproved agents safe, and that the people accepting the physiological risk are not the people capitalising the event. Its strongest point is that the existing regime spends heavily to enforce a line whose intellectual basis, as the exemption and equipment cases show, is not stable. The underlying dispute is the one set out in [[bioethics-of-enhancement]], with the unusual feature that one side has a testing budget and the other has a venue.

## Criticism and outlook

Anti-doping is criticised from three directions at once: as ineffective, since detection lags administration and the biological passport produces sanctions slowly and rarely; as unjust, since its burdens fall hardest on athletes from countries with the least legal support; and as incoherent, for the reasons set out above. Defenders reply that no regime with these aims could be fully coherent, and that the counterfactual is the 1990s, when entire pelotons were doped and athletes died. The claim to bodily self-determination advanced in [[morphological-freedom]] has never had much purchase here, because entry into competition is voluntary and the rules are a condition of it.

The technical pressure is toward interventions the current system cannot see: expression-modulating agents that add no foreign sequence, transient delivery that clears the body, and interventions targeting [[hallmarks-of-aging|ageing biology]] whose performance effects are indirect. If detection cannot keep pace, the choice narrows to profiling every athlete continuously — which raises the privacy problems examined in [[genetic-discrimination]] and, for neural interventions, [[mental-privacy]] — or to accepting that some enhancement will go unpunished. Neither branch preserves the position the Code currently occupies.

## See also

- [[gene-doping]]
- [[human-enhancement]]
- [[myostatin-inhibition]]
- [[enhancement-arms-race]]
- [[nootropics]]
- [[bioethics-of-enhancement]]
- [[disability-rights-and-enhancement]]
- [[exoskeleton]]

## References

[^cas2008]: `law` Court of Arbitration for Sport. *Pistorius v. International Association of Athletics Federations*, CAS 2008/A/1480. {The award turned on the federation failing to prove an overall advantage, and decides only that athlete's eligibility.}
[^weyand2009]: `paper` Weyand, P. G. and Bundle, M. W. "Point: Artificial limbs do make artificially fast running speeds possible." *Journal of Applied Physiology*, 2009, with the accompanying counterpoint by Kram, Grabowski, McGowan, Brown and Herr. {Published as a paired point and counterpoint by researchers who had worked on the same measurements and read them oppositely.}
[^echr2025]: `law` European Court of Human Rights. *Semenya v. Switzerland*, Chamber judgment 2023 and Grand Chamber judgment 2025.

==============================================================================
ARTICLE: epigenetic-clock
TITLE:   Epigenetic clocks
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/epigenetic-clock
SOURCE:  https://futurehumanwiki.com/raw/epigenetic-clock
==============================================================================

---
title: "Epigenetic clocks"
slug: "epigenetic-clock"
type: "technology"
status: "established"
horizon: "present"
trl: 6
categories: ["longevity"]
tags: ["epigenetics", "biomarkers", "aging", "dna methylation", "measurement"]
summary: "Statistical predictors that estimate age or mortality risk from DNA methylation patterns at selected sites in the genome."
updated: "2026-07-27"
humanEvidence: "Built and validated in large human cohorts, where age acceleration predicts mortality and disease modestly; no trial has shown that moving a clock reading changes an outcome."
access: "Sold direct to consumer as mail-in methylation tests and available as a research assay on standard arrays; no clock is an approved clinical test or an accepted regulatory endpoint."
reversibility: "reversible"
issues: ["The consumer-testing section makes claims about company practice with no citations."]
---

```infobox
{
  "caption": "Molecular aging measure",
  "rows": [
    { "label": "Type", "value": "Regression biomarker" },
    { "label": "Input", "value": "DNA methylation array data" },
    { "label": "First multi-tissue clock", "value": "Horvath, 2013" },
    { "label": "Typical size", "value": "3 to several hundred CpG sites" },
    { "label": "Regulatory status", "value": "No accepted surrogate endpoint" },
    { "label": "Readiness", "value": "TRL 6" }
  ]
}
```

**Epigenetic clocks** are statistical models that estimate a person's age, or their risk of death, from the pattern of methyl groups attached to cytosines at selected positions in their genome. They are the most widely used molecular measure in aging research, cheap enough to run on stored blood or saliva and accurate enough at predicting chronological age to be startling. What they measure biologically, and whether moving a clock reading means anything, remain open.

## What a clock measures

DNA methylation at CpG dinucleotides regulates gene expression and differs systematically between tissues and between young and old individuals. A clock is built by taking methylation values at hundreds of thousands of sites, measured on a commercial array, and fitting a penalised regression that predicts a target variable from a small subset of them. The selected sites are not a mechanism. They are whichever coordinates carried the most predictive signal in the training data.

For a clock trained on chronological age, the interesting quantity is not the prediction but the residual: the difference between predicted and actual age, usually called epigenetic age acceleration. A person whose blood reads five years older than their birth certificate is, on average across large cohorts, at somewhat elevated risk of death and of several chronic diseases. The effect is real and modest.

Compared with earlier candidate measures, methylation performs well. Leukocyte telomere length, discussed under [[telomeres-and-telomerase]], correlates with age so loosely in individuals that it is close to useless as a personal readout. Methylation clocks are far tighter, which is why they displaced telomere length as the field's default molecular measure of the [[hallmarks-of-aging|epigenetic hallmark]] within a few years.

## Generations of clocks

```compare
{
  "columns": ["First generation", "Second generation", "Pace-of-aging"],
  "rows": [
    { "label": "Examples", "values": ["Horvath 2013, Hannum 2013", "PhenoAge, GrimAge", "DunedinPACE"] },
    { "label": "Trained to predict", "values": ["Chronological age", "Clinical biomarkers, then mortality", "Longitudinal rate of change in organ-function measures"] },
    { "label": "Output", "values": ["Estimated age in years", "Estimated age in years", "Biological years elapsed per calendar year"] },
    { "label": "Association with mortality", "values": ["Modest", "Stronger", "Comparable to second generation"] },
    { "label": "Requires a longitudinal cohort to build", "values": ["No", "No", "Yes"] }
  ]
}
```

### First generation

Early clocks were trained directly on chronological age. [[steve-horvath]]'s 2013 multi-tissue clock used 353 CpG sites and predicted age across most human tissue types with a median error of roughly three and a half years, a result that also held for cells in culture and, notably, returned close to zero for [[induced-pluripotent-stem-cells]].[^horvath2013] Gregory Hannum's group published a blood-specific clock in the same year.[^hannum2013] Because these models are optimised to reproduce the calendar, any deviation from it is treated as error by the fitting procedure, which limits how much aging-relevant signal they can retain.

### Second generation

The second wave changed the training target. DNAm PhenoAge was fitted to a composite of clinical laboratory measures and chronological age rather than to age alone.[^levine2018] GrimAge went further, building methylation surrogates for plasma proteins and for smoking history and then regressing those on time to death.[^lu2019] Both predict mortality and disease incidence considerably better than first-generation clocks, which is unsurprising given what they were trained on.

### Pace-of-aging

DunedinPACE takes a different approach again. Rather than estimating a level, it estimates a rate, trained on two decades of repeated organ-function measurements in the Dunedin birth cohort, all of whom are the same age.[^belsky2022] Its output is a speed: roughly one biological year per calendar year for an average person. Because chronological age is constant within the training set, the model cannot be fitting the calendar.

## What the epidemiology shows

In population cohorts, age acceleration on second-generation clocks associates with smoking, obesity, low socioeconomic position, chronic infection, and markers of [[inflammaging]]. It also associates prospectively with incident cardiovascular disease, dementia and all-cause mortality, though generally with hazard ratios well below those of established clinical risk factors. GrimAge in particular carries a large share of its predictive weight in its smoking surrogate, which complicates the claim that it is measuring aging as such.

Clocks have been applied to questions about exceptional longevity, including studies of residents of the regions described under [[blue-zones]] and of supercentenarian cohorts relevant to [[maximum-human-lifespan]]. Results in such groups are limited by tiny sample sizes and by the same data-quality problems that affect the underlying age records.

## Reliability and confounding

Technical noise is a serious and underappreciated problem. Running the same DNA sample twice on the same array platform can produce clock estimates differing by more than a year, which is large relative to the effect sizes reported in intervention studies. Reconstructing clocks on principal components of the methylation data substantially improves test-retest reliability.[^higgins2022]

Cell composition is the other main confounder. Blood is a mixture, and the proportions of naive T cells, memory T cells, monocytes and granulocytes shift with age and with acute illness. A clock trained on whole blood partly measures that shift. Some clocks adjust for estimated cell counts; the adjustment is imperfect.

> [!caution] Contested
> No epigenetic clock has been shown to track a causal driver of aging rather than a downstream correlate. Methylation changes could be the mechanism, a readout of the mechanism, or an incidental consequence of cell-turnover history. The distinction is invisible to the regression.

## Response to interventions

The strongest use case for clocks would be as a short-cut endpoint, letting a trial read out in a year rather than in decades. See [[aging-biomarkers]] for the general problem and the regulatory bar. The evidence that clocks respond to interventions in humans is thin.

The most-cited human result is the TRIIM study, in which nine men received growth hormone with DHEA and [[metformin]] and showed a mean reduction of a few years in several clock readings alongside thymic changes.[^fahy2019] It had no control group and nine participants. In the CALERIE randomised trial of [[caloric-restriction]], stored samples showed a small slowing of DunedinPACE in the restricted arm but no significant effect on PhenoAge or GrimAge.[^waziry2023] That pattern, one clock moving and others not, recurs across studies and is not well explained.

[[epigenetic-reprogramming]] moves clock readings dramatically in cells and in mice, which is the strongest demonstration that the readings are manipulable. It is also the clearest case where a clock delta and a functional benefit could come apart, since the same treatment that resets methylation can erode cell identity.

## Beyond humans

The Mammalian Methylation Consortium constructed clocks that work across dozens of mammalian species using conserved CpG sites, allowing the same measure to be applied to a mouse, a bat and a bowhead whale.[^lu2023] These cross-species clocks are useful for comparative work on [[negligible-senescence]] and for testing whether interventions such as [[rapamycin]] or dietary restriction slow methylation change in animals whose lifespans can actually be measured. Mouse clocks have become a routine secondary endpoint in geroprotector studies, including work on [[senolytics]] and [[parabiosis-and-young-blood|heterochronic blood exchange]], with the same interpretive caveat as in humans.

## Consumer testing

Several companies sell direct-to-consumer methylation age tests, through the same mail-in channels as
the [[consumer-blood-testing|direct-to-consumer laboratory panels]]. The assays are usually real, but the reported precision often exceeds what the underlying reliability supports, the clocks used are frequently proprietary and unpublished, and a result expressed as a single number invites over-interpretation of what is a noisy estimate with wide individual uncertainty. Repeat testing of the same person over short intervals can produce swings that reflect assay variation rather than any change in [[biological-age]]. Guidance from professional bodies has been consistently cautious about clinical use.

## Outlook

Clocks are firmly established as research instruments and firmly unestablished as endpoints. The work that would change that is unglamorous: standardised assays, published reliability statistics, replication of intervention effects across independent cohorts, and demonstrations that a clock change predicts a later functional change in the same person. Competitions such as [[xprize-healthspan]] have deliberately chosen functional endpoints instead, on the view that measured muscle, cognitive and immune restoration is what the [[geroscience-hypothesis|geroscience]] argument ultimately has to deliver.

## See also

- [[biological-age]]
- [[aging-biomarkers]]
- [[epigenetic-reprogramming]]
- [[hallmarks-of-aging]]
- [[geroscience-hypothesis]]
- [[caloric-restriction]]
- [[healthspan]]

## References

[^horvath2013]: `paper` Horvath, S. "DNA methylation age of human tissues and cell types." *Genome Biology*, 2013. {Fitted to publicly available methylation datasets; the accuracy reported is against chronological age, the variable the model was trained on.}
[^hannum2013]: `paper` Hannum, G. et al. "Genome-wide methylation profiles reveal quantitative views of human aging rates." *Molecular Cell*, 2013.
[^levine2018]: `paper` Levine, M.E. et al. "An epigenetic biomarker of aging for lifespan and healthspan." *Aging*, 2018.
[^lu2019]: `paper` Lu, A.T. et al. "DNA methylation GrimAge strongly predicts lifespan and healthspan." *Aging*, 2019.
[^belsky2022]: `paper` Belsky, D.W. et al. "DunedinPACE, a DNA methylation biomarker of the pace of aging." *eLife*, 2022.
[^higgins2022]: `paper` Higgins-Chen, A.T. et al. "A computational solution for bolstering reliability of epigenetic clocks." *Nature Aging*, 2022.
[^fahy2019]: `paper` Fahy, G.M. et al. "Reversal of epigenetic aging and immunosenescent trends in humans." *Aging Cell*, 2019. {An uncontrolled study of nine men, run to test thymus regeneration rather than to test an intervention against a clock.}
[^waziry2023]: `paper` Waziry, R. et al. "Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial." *Nature Aging*, 2023. {An analysis of stored samples from a trial designed to test caloric restriction, not to validate the clocks.}
[^lu2023]: `paper` Lu, A.T. et al. "Universal DNA methylation age across mammalian tissues." *Nature Aging*, 2023.

==============================================================================
ARTICLE: epigenetic-reprogramming
TITLE:   Epigenetic reprogramming
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/epigenetic-reprogramming
SOURCE:  https://futurehumanwiki.com/raw/epigenetic-reprogramming
==============================================================================

---
title: "Epigenetic reprogramming"
slug: "epigenetic-reprogramming"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 3
categories: ["longevity", "genetics"]
tags: ["reprogramming", "epigenetics", "rejuvenation", "gene therapy", "aging", "stem cells"]
summary: "The use of pluripotency factors to reset age-associated marks on DNA and chromatin, restoring youthful gene expression in cells without converting them into stem cells."
updated: "2026-07-27"
humanEvidence: "No person has been treated with reprogramming as a rejuvenation therapy: those results are from mice or cultured human cells, where transient reprogramming lowered methylation age without loss of fibroblast identity. Full reprogramming to iPSCs is separate."
access: "Nothing to obtain outside the laboratory; no partial-reprogramming therapy has published clinical data, and companies have so far stated only an intention to file for first-in-human trials."
reversibility: "context"
issues: ["The company-authored old-mouse lifespan result is described without a citation."]
---

```infobox
{
  "caption": "Experimental rejuvenation technology",
  "rows": [
    { "label": "Type", "value": "Cell-state manipulation" },
    { "label": "Key factors", "value": "OSKM / OSK", "link": "/wiki/yamanaka-factors" },
    { "label": "First shown in vivo", "value": "2016 (mice)" },
    { "label": "Delivery", "value": "AAV, transgene, mRNA" },
    { "label": "Human trials", "value": "None reported as of 2026" },
    { "label": "Main risk", "value": "Teratoma, loss of cell identity" },
    { "label": "Readiness", "value": "TRL 3" }
  ]
}
```

**Epigenetic reprogramming** is the deliberate resetting of the chemical marks on DNA and histones that determine which genes a cell expresses. Applied briefly, it restores several youthful molecular features in old cells and old animals without turning them into stem cells; applied for too long, it erases cell identity and produces tumours. That narrow window between rejuvenation and dedifferentiation is the central technical problem of the field, and it has not yet been shown to be navigable in a human being.

## The epigenetic view of aging

Every cell in a body carries nearly the same genome but reads a different portion of it. Which portion is set by cytosine methylation, histone modification, chromatin accessibility, and the transcription-factor networks that maintain them. These patterns drift with age in ways that are reproducible enough to be read as a clock: methylation at a few hundred sites predicts chronological age across tissues to within a few years, which is the basis of the [[epigenetic-clock|epigenetic clocks]].

The drift has structure. Across mammals, CpG islands controlling developmental genes, many of them targets of the polycomb repressive complexes, gain methylation with age, while the genome as a whole loses it. Heterochromatin at repetitive regions relaxes. Chromatin that should be closed in a given cell type opens, and lineage-inappropriate genes become detectably transcribed at low levels. The result is not random noise but a loss of definition in the cell's regulatory state.

The reprogramming hypothesis treats this drift as more than a record. It proposes that the loss of a cell's correct expression programme is itself a driver of aging, one of the [[hallmarks-of-aging|hallmarks of aging]] classified as an "epigenetic alteration", and that restoring the programme restores function. The strong version of the claim is that the youthful information is still present in old cells and merely obscured, so that aging is partly a software problem rather than a matter of accumulated physical damage. David Sinclair's laboratory has argued this position most forcefully, including through a mouse model in which repeated non-mutagenic DNA breaks accelerated aging phenotypes; the interpretation of that work has been disputed by other groups. See [[david-sinclair]].

> [!caution] Contested
> No experiment has separated epigenetic drift as a cause of aging from epigenetic drift as a consequence of it. Reprogramming can improve function while leaving the causal question open, because the same intervention that rewrites methylation also changes transcription, metabolism, and chromatin structure at once.

## How it works

Reprogramming uses the [[yamanaka-factors]]: Oct4, Sox2, Klf4, and c-Myc, abbreviated OSKM. Sustained expression in a differentiated cell drives it backwards through an intermediate state and into pluripotency, producing [[induced-pluripotent-stem-cells]]. Full reprogramming resets methylation age to approximately zero, which is why iPSC lines derived from an elderly donor score as embryonic on methylation clocks.

The rejuvenation application stops well short of that endpoint. In vitro, cells passing through reprogramming lose methylation age steadily from the first days, but do not lose their somatic identity until a later transition point.[^olova2019] The gap between those two curves is the operating window. [[partial-reprogramming]] exploits it by expressing the factors transiently, cyclically, or at low dose, so that the epigenome is partially reset while the cell remains a fibroblast, a hepatocyte, or a neuron.

Three delivery strategies dominate. Doxycycline-inducible transgenes built into a mouse line give the cleanest experimental control and are useless clinically. [[aav-vectors|AAV vectors]] carrying inducible cassettes work in vivo and persist, which is a liability when the payload is a pluripotency factor. mRNA delivered in [[lipid-nanoparticles]] is inherently transient and is the approach most often proposed for human use, at the cost of poor delivery to most tissues other than liver.

Many groups now drop c-Myc, the most oncogenic of the four, and use OSK alone. Others screen for entirely different transcription-factor combinations that move a cell toward a younger state of its own type rather than toward pluripotency, an approach that avoids the identity-loss problem by construction. Chemical cocktails that substitute small molecules for some or all of the factors exist and have produced human iPSCs, but chemical rejuvenation claims specifically have drawn methodological criticism.

## Development history

```timeline
[
  { "year": "1962", "title": "Differentiation is reversible", "text": "John Gurdon obtains swimming tadpoles from frog eggs carrying nuclei taken from intestinal cells, showing that specialisation does not delete genetic information." },
  { "year": "2006", "title": "Four factors reprogram a fibroblast", "text": "Takahashi and Yamanaka convert mouse fibroblasts into induced pluripotent stem cells using Oct4, Sox2, Klf4 and c-Myc." },
  { "year": "2013", "title": "In vivo reprogramming produces tumours", "text": "Abad and colleagues express OSKM throughout living mice and obtain pluripotent cells in multiple organs, along with lethal teratomas." },
  { "year": "2016", "title": "Cyclic partial reprogramming", "text": "Ocampo and colleagues extend median lifespan in a progeroid mouse strain using intermittent OSKM induction, and improve injury repair in wild-type mice." },
  { "year": "2020", "title": "Vision restored in mice", "text": "Lu and colleagues deliver OSK by AAV to retinal ganglion cells and recover visual function after optic nerve injury and in aged animals." },
  { "year": "2022", "title": "Capital arrives", "text": "Altos Labs launches with about $3 billion in committed funding and a research programme centred on reprogramming; NewLimit and others pursue narrower transcription-factor screens." },
  { "year": "2024–2026", "title": "Approach to the clinic", "text": "Companies state intentions to file for first-in-human trials of OSK gene therapy in eye disease. As of mid-2026 no partial-reprogramming therapy has published clinical efficacy data." }
]
```

## Evidence in animals

The founding in vivo result came from Alejandro Ocampo and colleagues in the laboratory of Juan Carlos Izpisua Belmonte. Mice carrying a progeria-causing lamin A mutation, given two days of OSKM induction per week, lived longer than untreated littermates and showed improved tissue histology.[^ocampo2016] In normally aging wild-type mice, the same protocol improved regeneration of pancreatic beta cells and muscle after injury but did not report a lifespan extension. The distinction matters: accelerated-aging mouse models are not miniature versions of normal aging, and interventions that rescue them frequently do nothing in ordinary animals.

The eye produced the result that has since shaped the field's clinical strategy. Yuancheng Lu and colleagues, working in David Sinclair's laboratory, recovered lost sight in three separate mouse models by delivering OSK to retinal ganglion cells, and showed the recovery required the TET enzymes that strip methyl groups from DNA, which points at a specifically epigenetic mechanism rather than a general trophic effect.[^lu2020] The retina is an unusually favourable target: small, immune-privileged, dosable by direct injection, and readable without touching the animal. Every clinical intention announced in reprogramming so far has followed that logic rather than aiming at systemic treatment.

Longer-term dosing has been tested. Continuous low-level partial reprogramming over months in middle-aged and old mice altered age-associated transcriptional and methylation signatures in several tissues without gross toxicity.[^browder2022] At least one company-authored study has reported that inducible OSK delivered to very old mice extended their remaining lifespan; that result has not been reproduced by an independent group. In human cells, transient reprogramming taken to the maturation phase and then withdrawn reduced transcriptomic and methylation age substantially while the cells retained fibroblast identity and, in that study, some functional improvements.[^gill2022]

What no animal study has yet shown is the combination that would matter most: a large lifespan extension in normal, genetically unmodified, wild-type mice from partial reprogramming alone, replicated across laboratories. Much of the published in vivo work uses a small number of transgenic mouse lines built in a handful of laboratories, which limits how much independent replication has been possible. The [[senolytics]] field went through a comparable phase, with dramatic mouse results followed by human trials whose results have been modest.

Cell-type coverage is another gap. Reprogramming effects have been characterised in fibroblasts, retinal ganglion cells, hepatocytes, muscle satellite cells, and a few immune populations. Post-mitotic tissue with high metabolic demand and no regenerative reserve, notably cardiac muscle and most of the brain, is both the hardest to reach and the least forgiving of a transient loss of identity.

## Organisations and money

Reprogramming attracted more capital between 2021 and 2023 than any other approach in [[geroscience-hypothesis|geroscience]]. [[altos-labs]] launched publicly in January 2022 with roughly $3 billion in committed funding, reported backing from Jeff Bezos and Yuri Milner, and a roster including Izpisua Belmonte, [[steve-horvath]], and [[shinya-yamanaka]] as a senior scientific adviser. It has published research but announced no clinical programme.

[[newlimit]], founded by Brian Armstrong and Blake Byers, takes a different route: high-throughput screening of transcription-factor combinations for effects on a specific cell type, initially hepatocytes, with an unusually open posture about releasing negative and intermediate data. [[retro-biosciences]] funds a reprogramming programme alongside work on [[autophagy]] and plasma-derived factors. [[calico]] and several smaller companies maintain programmes. Life Biosciences has stated an intention to bring an OSK gene therapy for optic neuropathy into first-in-human testing, which would be the first clinical test of the approach.

## Limitations and risks

The tumour risk is not hypothetical. Systemic sustained OSKM expression in mice produces teratomas and kills the animals.[^abad2013] Every therapeutic proposal therefore depends on dose control, and gene therapy is not a modality with a dose dial. A viral cassette that misfires, or an inducible promoter that leaks, does so for the life of the transduced cell. Doxycycline-controlled systems solve this in mice by making expression depend on a drug the experimenter withholds; no equivalent system has been shown to be reliable over decades in a human.

Loss of cell identity is the subtler failure, because a cell that has drifted part of the way out of its role can score younger on a clock while performing its job worse, and tissues differ sharply in how much drift they tolerate. The dose-response evidence behind that asymmetry, and the failure modes that sit inside the apparently tolerated window, are set out under [[partial-reprogramming]].

Reprogramming also does not address the majority of age-related damage. It does not repair somatic mutations, remove extracellular aggregates, reverse crosslinked collagen, or clear the cells that drive [[cellular-senescence]] and [[inflammaging]]. It rewrites regulatory marks. If the dominant causes of aging in humans are genomic and structural rather than epigenetic, a perfect reprogramming therapy would help less than its advocates expect.

Finally, delivery is the binding constraint on almost every version of the idea, as it is for [[somatic-gene-therapy]] generally. Reaching most brain, muscle, and immune cells at controllable dose remains unsolved.

## What "age reversal" means here

Press coverage of reprogramming routinely reports that mice or cells were made "younger by X years". The measurement behind that phrase is almost always a methylation clock reading, and the relationship between a clock reading and [[biological-age|biological age]] is correlational. Reprogramming demonstrably moves clock readings. Whether it moves the underlying process the clock was trained to approximate is a separate claim that requires functional endpoints: grip strength, cognition, immune competence, mortality.

> [!key] The distinction that matters
> A clock reading is a prediction of chronological age fitted to methylation data. An intervention that changes methylation can move the prediction without changing anything the prediction was a proxy for. Functional and mortality endpoints, not clock deltas, are what regulators and [[aging-biomarkers|biomarker consortia]] treat as evidence.

## Outlook

The near-term path runs through narrow, high-value, locally deliverable indications rather than systemic rejuvenation: optic neuropathies, corneal endothelium, possibly liver. Those trials would establish whether the mouse safety window survives translation, which is the single question with the most information value in the field.

The wider claim, that partial reprogramming is a general-purpose rejuvenation platform, rests on animal results that are genuinely striking and genuinely narrow. Whether it becomes the mechanism behind [[longevity-escape-velocity|escape-velocity]] arguments or a useful ophthalmic therapy depends on results that do not exist yet. A cautionary comparison is [[telomeres-and-telomerase|telomerase therapy]], which also reversed a clear molecular marker of aging in mice and has spent two decades failing to become a human medicine because the cancer risk could not be bounded.

## See also

- [[partial-reprogramming]]
- [[yamanaka-factors]]
- [[epigenetic-clock]]
- [[induced-pluripotent-stem-cells]]
- [[biological-age]]
- [[altos-labs]]
- [[gene-therapy-for-aging]]
- [[hallmarks-of-aging]]

## References

[^olova2019]: `paper` Olova, N., Simpson, D.J., Marioni, R.E., Chandra, T. "Partial reprogramming induces a steady decline in epigenetic age before loss of somatic identity." *Aging Cell*, 2019.
[^ocampo2016]: `paper` Ocampo, A. et al. "In Vivo Amelioration of Age-Associated Hallmarks by Partial Reprogramming." *Cell*, 2016. {The lifespan result is in a progeria mouse model; in wild-type mice the same protocol improved injury repair with no lifespan gain reported.}
[^lu2020]: `paper` Lu, Y. et al. "Reprogramming to recover youthful epigenetic information and restore vision." *Nature*, 2020. {Mouse retina dosed by direct injection into an immune-privileged tissue, which is the easiest possible delivery target.}
[^browder2022]: `paper` Browder, K.C. et al. "In vivo partial reprogramming alters age-associated molecular changes during physiological aging in mice." *Nature Aging*, 2022.
[^gill2022]: `paper` Gill, D. et al. "Multi-omic rejuvenation of human cells by maturation phase transient reprogramming." *eLife*, 2022. {Cultured human skin fibroblasts, not people; the rejuvenation is measured as transcriptomic and methylation age in the dish.}
[^abad2013]: `paper` Abad, M. et al. "Reprogramming in vivo produces teratomas and iPS cells with totipotency features." *Nature*, 2013.

==============================================================================
ARTICLE: epigenome-editing
TITLE:   Epigenome editing
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/epigenome-editing
SOURCE:  https://futurehumanwiki.com/raw/epigenome-editing
==============================================================================

---
title: "Epigenome editing"
slug: "epigenome-editing"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 6
categories: ["genetics"]
tags: ["crispr", "epigenetics", "gene regulation", "gene therapy", "dna methylation", "biotechnology"]
summary: "Targeted modification of the chemical marks that control gene expression, switching a gene off or on without altering the underlying DNA sequence."
updated: "2026-07-27"
humanEvidence: "Early-phase trials in chronic hepatitis B and facioscapulohumeral muscular dystrophy are reported to have dosed patients in 2024 and 2025, with no published results; the year-long silencing results are from mice."
access: "Not obtainable outside company-run clinical trials; no epigenetic editor has been approved by any regulator."
reversibility: "context"
issues: ["Programme statuses in Current state are described without citations to registry records."]
---

```infobox
{
  "caption": "Gene regulation platform",
  "rows": [
    { "label": "Type", "value": "Programmable transcriptional control" },
    { "label": "Targeting modules", "value": "dCas9, zinc fingers, TALEs" },
    { "label": "First demonstrated", "value": "2013 (dCas9 repression)" },
    { "label": "Changes DNA sequence", "value": "No" },
    { "label": "Durability", "value": "Locus-dependent" },
    { "label": "First human trials", "value": "2024–2025" },
    { "label": "Readiness", "value": "TRL 6" }
  ]
}
```

**Epigenome editing** changes how much of a gene a cell makes without changing the gene itself. A programmable DNA-binding protein is aimed at a promoter or enhancer and carries with it an enzyme that writes or erases a regulatory mark — DNA methylation, histone modification — or that simply recruits the cell's own repression machinery. The sequence is left intact, which means the change is in principle reversible, and which also means it can be lost. That distinguishes it sharply from [[base-editing]] and [[prime-editing]], which alter the genome permanently and cannot be undone once installed.

## How it works

Every epigenome editor is two interchangeable halves bolted together: something that finds a chosen stretch of DNA, and something that acts on the chromatin once it arrives. The halves are developed independently, which is why advances in targeting from nuclease work transfer directly to this field.

### The DNA-binding module

The most common targeting scaffold is catalytically dead Cas9, produced by disabling both nuclease domains of the enzyme described in [[crispr-cas9]]. It retains full guide-RNA-programmed binding and none of the cutting. Zinc-finger proteins and TALE arrays, the platforms that preceded CRISPR, remain in use for the same purpose and have two advantages for therapy: they are considerably smaller, which eases packaging, and being closer to human protein scaffolds they may provoke less of an immune response than a bacterial enzyme. The tradeoff is that retargeting them requires protein engineering rather than swapping twenty nucleotides of RNA.

### The effector

What is fused to the binding module determines the effect.

- **Repressors.** A KRAB domain recruits heterochromatin machinery and shuts transcription down within hours. Adding DNA methyltransferase domains lays down methylation at CpG sites, which the cell's maintenance methyltransferase copies to daughter strands during replication.
- **Activators.** Transcriptional activation domains, alone or in multiplexed arrays, raise expression from a silent or weakly expressed locus. Histone acetyltransferase fusions open chromatin at enhancers rather than acting on the promoter directly.
- **Erasers.** Demethylases remove existing methylation and can reactivate a silenced gene, which is the basis of most reversibility claims.

The distinction that matters clinically is between effects that persist only while the editor is present and effects that persist after it is gone. Simple repressor fusions do the former. Combining a KRAB domain with methyltransferases produces silencing that survives cell division and differentiation after a single transient exposure — the property that turned epigenome editing from a laboratory technique into a therapeutic proposition.[^nunez2021]

```timeline
[
  { "year": "2013", "title": "Programmable repression", "text": "Several groups show that catalytically dead Cas9, alone or fused to a KRAB repressor domain, can silence a targeted gene in bacteria and in human cells." },
  { "year": "2015–2016", "title": "Writers and erasers", "text": "Fusions to histone acetyltransferase, DNA methyltransferase and TET demethylase domains demonstrate direct editing of regulatory marks at chosen loci." },
  { "year": "2016", "title": "Hit-and-run silencing", "text": "Transient delivery of combined repressor and methyltransferase effectors is shown to produce inheritable silencing of endogenous genes in cultured cells." },
  { "year": "2021", "title": "Single-construct memory", "text": "A single fusion protein is shown to install durable, reversible transcriptional silencing at a large fraction of the genome's promoters." },
  { "year": "2024", "title": "Durable silencing in animals", "text": "A single in vivo dose targeting a cholesterol-regulating gene produces silencing lasting many months in mice." },
  { "year": "2024–2025", "title": "First patients", "text": "Epigenetic silencing candidates for chronic hepatitis B and for facioscapulohumeral muscular dystrophy are reported dosed in early-phase trials." }
]
```

## Development history

The tools arrived in a predictable order: repression first, because recruiting existing silencing machinery is easy;[^gilbert2013] activation next, because it needed multiplexed activation domains to reach useful expression levels; and durable, heritable marks last, because those required combining several enzymatic activities at once.[^liu2016] The pivotal demonstration for medicine was that a transient dose could leave a permanent regulatory state — a "hit-and-run" edit — since a therapy that requires continuous expression of a bacterial protein is difficult to make safe.[^amabile2016]

Work in mice then showed that a single systemic dose could silence a liver gene for the better part of a year without touching the DNA sequence.[^cappelluti2024] That result, together with the observation that reversal is possible with a demethylating editor, defined the commercial pitch: the durability of gene therapy with an off switch.

## Current state

As of 2026 the field has moved into first-in-human testing, with results still limited. Several venture-funded companies pursue durable epigenetic silencing — one focused on chronic hepatitis B, where the aim is to silence both the viral episome and integrated viral DNA that antivirals cannot clear; another on facioscapulohumeral muscular dystrophy, where the disease is caused by inappropriate expression of a normally silent gene and where restoring silence is a more natural fix than cutting. Zinc-finger repressors have been developed for chronic pain and for prion disease, where lowering expression of a single protein is the whole therapeutic goal, though how far each has advanced into human testing is not firmly established. None of these has produced anything comparable to the approval record of nuclease-based [[somatic-gene-therapy]], and no epigenetic editor has reached the regulatory stage of [[casgevy]].

Consolidation has been rapid. Companies in the sector have merged or refocused since 2024, and the small number of programmes in the clinic means that a single safety event would change the field's trajectory. Readers should treat any specific programme status here as possibly outdated.

> [!key] Why silencing is the natural first target
> For a large class of diseases the problem is too much of something — a viral genome, a toxic protein, a gene that should be off. Turning expression down requires no template, no repair pathway and no correction of a specific mutation, so one editor can serve every patient with the condition regardless of their individual genotype.

## Limitations

Durability is locus-dependent and imperfectly predictable. Promoters rich in CpG dinucleotides accept and propagate methylation well; CpG-poor promoters often do not, and some loci resist stable silencing entirely. Activation is harder than repression: raising a gene's output to a therapeutically meaningful, sustained level is much less reliable than shutting it down, which is why activation-based programmes lag repression-based ones.

Delivery imposes the same constraints as elsewhere in the field. The fusion proteins are large — larger than Cas9 alone — so single-vector packaging in [[aav-vectors]] is often impossible, and most in vivo work uses [[lipid-nanoparticles]] carrying messenger RNA, which mainly reaches the liver. Repeat dosing, if durability proves shorter than hoped, would run into anti-Cas9 immunity.

## Risks

Because dCas9 binds tolerantly at sequences that would not be cut, epigenome editors can perturb transcription at sites where a nuclease would produce no [[crispr-off-target-effects|off-target lesion]] at all. Off-target binding is transient and reversible in principle, but if the effector deposits durable methylation the perturbation may not be. Silencing can also spread beyond the intended element to neighbouring genes, and long-range enhancer contacts mean the affected gene is not always the nearest one.

A structural concern is verification. Sequencing shows that the genome is unchanged, which is reassuring but also means the standard safety assay is blind to the edit. Confirming that an epigenetic therapy did what was intended, and nothing else, requires genome-wide methylation and expression profiling of the treated tissue — feasible in a mouse, difficult in a patient.

> [!caution] Reversibility is a design goal, not a guarantee
> Erasing an installed mark has been demonstrated in cultured cells and in animals. Whether a clinician could reliably reverse a silencing therapy in a patient, in the tissue where it was delivered, has not been shown.

## Relation to reprogramming and aging

Epigenome editing is often confused with [[epigenetic-reprogramming]], which uses [[yamanaka-factors|pluripotency transcription factors]] to reset a cell's epigenome globally rather than at a chosen locus. The two share a premise — that cell state is written in modifiable marks — and little else. Targeted editing changes one element and leaves identity alone; [[partial-reprogramming]] changes identity and leaves the specific targets unchosen. Companies such as [[newlimit]] work between the two, screening transcription factor combinations for effects on cell state rather than editing single regulatory elements, and readings from an [[epigenetic-clock]] are sometimes used as an outcome measure in both approaches despite the unresolved question of what such clocks actually track.

## Outlook

The near-term test is durability in humans. Mouse liver is a permissive setting: hepatocytes divide slowly and the delivery problem is solved. Whether a single dose can hold a gene off for years in a dividing tissue, and whether the immune system tolerates the machinery, will determine whether epigenome editing becomes a category of medicine or a laboratory technique with a brief clinical detour. A heritable-germline application is not on any credible roadmap, and the frameworks in [[governance-of-genome-editing]] would treat a durable epigenetic change to an embryo no differently from [[germline-editing]] proper, since the marks would be transmitted to every cell of the resulting person.

## See also

- [[crispr-cas9]]
- [[base-editing]]
- [[prime-editing]]
- [[epigenetic-reprogramming]]
- [[partial-reprogramming]]
- [[epigenetic-clock]]
- [[somatic-gene-therapy]]
- [[crispr-off-target-effects]]

## References

[^gilbert2013]: `paper` Gilbert, L. A. et al. "CRISPR-mediated modular RNA-guided regulation of transcription in eukaryotes." *Cell*, 2013.
[^liu2016]: `paper` Liu, X. S. et al. "Editing DNA methylation in the mammalian genome." *Cell*, 2016.
[^amabile2016]: `paper` Amabile, A. et al. "Inheritable silencing of endogenous genes by hit-and-run targeted epigenetic editing." *Cell*, 2016.
[^nunez2021]: `paper` Nuñez, J. K. et al. "Genome-wide programmable transcriptional memory by CRISPR-based epigenome editing." *Cell*, 2021. {Demonstrated in cultured human cells; the memory reported is measured across cell divisions in a dish, not in an organism.}
[^cappelluti2024]: `paper` Cappelluti, M. A. et al. "Durable and efficient gene silencing in vivo by hit-and-run epigenome editing." *Nature*, 2024. {The long-lasting silencing is in mouse liver, the tissue lipid nanoparticles reach best and one that divides slowly.}

==============================================================================
ARTICLE: eric-drexler
TITLE:   Eric Drexler
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/eric-drexler
SOURCE:  https://futurehumanwiki.com/raw/eric-drexler
==============================================================================

---
title: "Eric Drexler"
slug: "eric-drexler"
type: "person"
status: "contested"
horizon: "present"
categories: ["people", "nanomedicine"]
tags: ["nanotechnology", "molecular manufacturing", "grey goo", "foresight", "artificial intelligence"]
summary: "American engineer who proposed molecular nanotechnology, wrote Engines of Creation and Nanosystems, and later worked on AI strategy at Oxford."
updated: "2026-07-27"
issues: ["Bill Joy's 2000 relinquishment essay is described without a citation."]
---

```infobox
{
  "caption": "Engineer and theorist",
  "rows": [
    { "label": "Born", "value": "25 April 1955, Alameda, California" },
    { "label": "Nationality", "value": "American" },
    { "label": "Education", "value": "PhD, MIT, 1991" },
    { "label": "Known for", "value": "Molecular nanotechnology", "link": "/wiki/molecular-assembler" },
    { "label": "Key books", "value": "Engines of Creation (1986); Nanosystems (1992)" },
    { "label": "Co-founded", "value": "Foresight Institute, 1986" },
    { "label": "Later affiliation", "value": "Future of Humanity Institute, Oxford" }
  ]
}
```

**Eric Drexler** is an American engineer and theorist who argued that mechanical control of chemical
synthesis at the molecular scale is physically permitted, and that a technology built on it would
allow manufacturing with atomic precision. His 1986 book *Engines of Creation* introduced molecular
nanotechnology to a general audience along with the [[grey-goo]] scenario; his 1992 technical
treatise *Nanosystems* attempted to establish the physical case. Neither the machines nor the
catastrophe has materialized, and the resulting argument over why is the central fact of his career.

## Career

Drexler studied at MIT, initially working on space systems including solar sails, and completed a
doctorate there in 1991 under Marvin Minsky at the Media Lab — the first doctorate awarded in
molecular nanotechnology. His first technical publication on the subject appeared a decade
earlier, proposing that protein engineering could bootstrap a route to general molecular
machinery.[^drexler1981] With Christine Peterson he co-founded the Foresight Institute in 1986 to
promote and monitor the field.

After the American nanotechnology programme moved decisively away from his agenda, he shifted focus.
He joined the Future of Humanity Institute at Oxford as a research fellow, working with
[[nick-bostrom]] and [[anders-sandberg]] on technology strategy and, latterly, on artificial
intelligence.

## Molecular nanotechnology

The core claim of *Nanosystems* is mechanical rather than chemical.[^nanosystems1992] Biology
demonstrates that molecular machines — ribosomes, motor proteins, polymerases — work; Drexler argued
that machines built from stiff covalent materials such as diamondoid could work far better, and he
provided calculations of stiffness, error rates, thermal noise, energy dissipation and throughput for
hypothetical bearings, gears and positioning devices. From these he derived the concept of a
[[molecular-assembler]]: a device that positions reactive molecular fragments to fabricate specified
structures, including copies of itself.

> [!note] Terminology
> "Nanotechnology" was coined by Norio Taniguchi in 1974 for precision machining at nanometre
> tolerances. Drexler adopted it in 1986 for something quite different — mechanical control of
> matter atom by atom. When American funding agencies took the word over in the late 1990s, they
> restored a meaning closer to Taniguchi's. Three senses of one term now circulate, which is a
> recurring source of confusion in the literature.

The book's method is engineering analysis of the physically permissible, deliberately excluding the
question of how to build the first device. That exclusion is both its strength — it made a
falsifiable physical argument rather than a promise — and the source of every subsequent objection.

## The Smalley debate

The most consequential criticism came from Richard Smalley, a Nobel laureate in chemistry and a
central figure in the American nanotechnology programme. Smalley argued in 2001 that mechanically
guided synthesis fails on two counts: a manipulator large enough to hold a reactive atom would be
too big to fit in the space where the reaction must occur, and an atom held by a manipulator would
adhere to it rather than transfer to the target — the "fat fingers" and "sticky fingers"
objections.[^smalley2001] The two exchanged open letters, published in *Chemical & Engineering News*
in 2003.

The exchange settled nothing technically. Drexler answered that his proposals do not involve
grasping bare atoms, and that mechanosynthesis by controlled positioning of reactive molecular
tips is a different operation from the one Smalley was refuting; Smalley responded that chemistry is
not a matter of pushing pieces into place and accused Drexler of frightening children. What the
exchange did settle was institutional. Smalley's position carried the day inside the United States
National Nanotechnology Initiative, whose funding went to materials science, nanoparticles and
characterization tools rather than to molecular manufacturing.

```compare
{
  "columns": ["Drexler's programme", "What nanotechnology became"],
  "rows": [
    { "label": "Core object", "values": ["Positional molecular assembler", "Engineered nanoparticles and materials"] },
    { "label": "Method", "values": ["Mechanical positioning of reactants", "Solution chemistry and self-assembly"] },
    { "label": "Medical form", "values": ["Autonomous nanorobots in the bloodstream", "Lipid nanoparticles, drug conjugates, DNA origami"] },
    { "label": "Status", "values": ["No device built", "Approved medicines and commercial materials"] }
  ]
}
```

## Grey goo and its retraction

*Engines of Creation* included a scenario in which self-replicating assemblers escape control and
consume the biosphere. Drexler introduced it as a hazard to be designed against, but it detached
from the book and became the popular image of nanotechnology, amplified by fiction and by Bill Joy's
widely read 2000 essay on relinquishment, which treated self-replicating nanomachines alongside
genetic engineering and AI as a novel class of [[existential-risk]].

Drexler subsequently argued that the scenario had been a strategic error on his part. With Chris
Phoenix he set out the case that efficient molecular manufacturing does not require free-floating
self-replicators at all — factory architectures with fixed machinery and external control are both
more practical and inherently incapable of ecological escape.[^dp2004] He has since described the
grey-goo framing as an unnecessary distraction that made the field harder to fund and easier to
dismiss. The episode is a standard case study in how a risk illustration can capture a technology's
public identity, and in the difficulty of applying the [[precautionary-principle]] to a hazard whose
mechanism was hypothetical.

## Later work

*Radical Abundance* (2013) restates the programme under the name atomically precise manufacturing,
dropping the nanorobot imagery and emphasizing factory-scale production with atomic precision as an
economic rather than a medical proposition. At Oxford he turned to artificial intelligence, arguing
in a 2019 technical report that superintelligence is more likely to arrive as a collection of
specialized, bounded services than as a unified agent with goals — a direct alternative to the
agent-centred framing in Bostrom's *Superintelligence*.[^cais2019] The argument has been influential
among researchers who find the classic agent model unrepresentative of how machine learning
systems are actually deployed, and it bears on [[artificial-general-intelligence]] more than on
anything in his nanotechnology work.

## Reception and legacy

Chemists have largely not adopted Drexler's programme, and no molecular assembler exists. What
exists instead is a large and productive field of molecular machines built by chemists on their own
terms: rotaxane-based motors and switches recognized by the 2016 Nobel Prize in Chemistry,
[[ai-protein-design|computational protein design]] that now produces novel folded structures to
specification, [[dna-nanotechnology|DNA origami]]
devices, and [[microrobots-in-medicine|magnetically steered microrobots]]. Medicine's nanoscale
successes — [[lipid-nanoparticles]], [[targeted-drug-delivery|antibody-drug conjugates]] — arrived
through chemistry and immunology rather than through mechanical engineering.

Drexler's supporters read this as vindication of the physical claim by other means; his critics read
it as evidence that the mechanical framing was the wrong abstraction from the start, and that the
bottleneck was never whether atoms can be arranged but how. The distinction matters for
[[medical-nanorobots]] and for [[respirocytes]], whose design studies assume the assembler
Drexler described. Forty years after *Engines of Creation*, no experiment has demonstrated
positional mechanosynthesis of a useful structure, and no experiment has shown it to be
impossible.

## See also

- [[molecular-assembler]]
- [[grey-goo]]
- [[medical-nanorobots]]
- [[dna-nanotechnology]]
- [[respirocytes]]
- [[nick-bostrom]]
- [[anders-sandberg]]
- [[precautionary-principle]]

## References

[^drexler1981]: `paper` Drexler, K. E. "Molecular engineering: An approach to the development of general capabilities for molecular manipulation." *Proceedings of the National Academy of Sciences*, 1981.
[^nanosystems1992]: `book` Drexler, K. E. *Nanosystems: Molecular Machinery, Manufacturing, and Computation*. Wiley, 1992.
[^smalley2001]: `news` Smalley, R. E. "Of Chemistry, Love and Nanobots." *Scientific American*, 2001. {A magazine essay by a party to the dispute; the fat-fingers and sticky-fingers objections are argued, not demonstrated experimentally.}
[^dp2004]: `paper` Phoenix, C. and Drexler, K. E. "Safe exponential manufacturing." *Nanotechnology*, 2004.
[^cais2019]: `report` Drexler, K. E. *Reframing Superintelligence: Comprehensive AI Services as General Intelligence*. Future of Humanity Institute Technical Report, University of Oxford, 2019. {An institutional technical report from the author's own institute, not a peer-reviewed publication.}

==============================================================================
ARTICLE: exercise-and-aging
TITLE:   Exercise as a geroprotector
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/exercise-and-aging
SOURCE:  https://futurehumanwiki.com/raw/exercise-and-aging
==============================================================================

---
title: "Exercise as a geroprotector"
slug: "exercise-and-aging"
type: "intervention"
status: "established"
horizon: "present"
trl: 9
categories: ["longevity"]
tags: ["exercise", "cardiorespiratory fitness", "sarcopenia", "aging", "healthspan", "clinical trials"]
summary: "Structured physical activity, the intervention with the strongest evidence for extending healthy life in humans and the benchmark any geroprotective drug must beat."
updated: "2026-07-28"
humanEvidence: "A randomized trial in sedentary adults aged 70 to 89 lowered incident mobility disability, and large human cohorts link fitness to lower mortality; no trial has tested lifespan."
access: "Freely available and close to costless: guideline-level activity requires no prescription, and adherence rather than price or supply is the binding constraint."
reversibility: "reversible"
issues: ["The claim that two trials found metformin blunted training adaptations carries no citation.", "The per-decade decline in VO2max is stated without a source."]
---

```infobox
{
  "caption": "Behavioural intervention",
  "rows": [
    { "label": "Type", "value": "Aerobic, resistance, balance training" },
    { "label": "Standard guideline", "value": "150–300 min moderate activity weekly" },
    { "label": "Best-evidenced outcome", "value": "Reduced all-cause mortality" },
    { "label": "Randomized functional evidence", "value": "Yes (mobility disability)" },
    { "label": "Effect on maximum lifespan", "value": "Not demonstrated" },
    { "label": "Cost", "value": "Near zero" },
    { "label": "Readiness", "value": "TRL 9" }
  ]
}
```

**Exercise as a geroprotector** is the proposition that structured physical activity acts on the
underlying biology of aging rather than only on the fitness of the individual muscles and vessels
trained. It has better human evidence than any drug in [[geroscience-hypothesis|geroscience]],
including randomized trials with functional endpoints, and it is the comparator against which every
candidate geroprotective compound is implicitly measured. Its limits are equally clear: it has not
been shown to extend maximum lifespan in laboratory rodents, and adherence, not efficacy, is the
binding constraint in practice.

```keyfacts
[
  { "value": "3–4 yrs", "label": "Associated life expectancy gain", "note": "pooled cohort analysis of leisure-time activity at guideline levels; observational" },
  { "value": "No plateau", "label": "Fitness–mortality relationship", "note": "in a large treadmill-testing cohort, mortality kept falling with higher fitness" },
  { "value": "~10%", "label": "Per-decade decline in VO2max", "note": "in sedentary adults, accelerating after 70" }
]
```

## The evidence

The observational base is unusually large and unusually consistent. Pooled cohort analyses associate
leisure-time activity at guideline levels with a gain in life expectancy on the order of three to
four years.[^moore2012] In a retrospective study of more than a hundred thousand patients referred for
treadmill testing, all-cause mortality fell steadily as measured cardiorespiratory fitness rose, with
no observed level beyond which additional fitness ceased to help.[^mandsager2018] Grip strength, a
cheap proxy for whole-body muscle function, predicts all-cause and cardiovascular mortality across
countries and income levels.[^leong2015]

Observational data on exercise carry a specific hazard: people become sedentary because they are
ill, so low activity can be a symptom rather than a cause. What distinguishes exercise from most
geroprotective candidates is that this objection has been tested. The LIFE trial randomized
sedentary adults aged 70 to 89 to a structured physical activity programme or a health education
control and found a lower incidence of major mobility disability, defined as the loss of the ability
to walk a quarter mile, in the exercise group.[^pahor2014] That is a randomized functional outcome
in an older population, which no drug in [[senolytics|the senolytic]], [[rapamycin|mTOR-inhibitor]],
or [[nad-precursors|NAD-precursor]] classes has achieved.

## Mechanisms

Exercise engages most of the processes catalogued as [[hallmarks-of-aging]], which is the substance of
the claim that it is geroprotective rather than merely healthful.

Endurance training increases mitochondrial content and quality through PGC-1α signalling and
improves oxidative capacity, acting on the machinery whose decline defines
[[mitochondrial-dysfunction|mitochondrial aging]]. Acute exercise induces [[autophagy]] in muscle,
liver, and other tissues, and in mice that capacity is required for several of exercise's metabolic
benefits. Regular training lowers circulating inflammatory markers over time, opposing
[[inflammaging]], despite provoking a transient inflammatory response acutely. Resistance training
counteracts [[sarcopenia]], the loss of muscle mass and, faster still, of strength, which begins in the
fourth decade and accelerates after sixty. Mechanical loading maintains bone density. Aerobic
training in older adults has been associated with increased hippocampal volume in randomized
studies, and exercise alters the circulating proteome in ways that transfer benefits between
animals, which connects it to the [[parabiosis-and-young-blood|circulating-factor]] literature.
Recovery practice can work against the same adaptations: regular post-exercise cold-water immersion
reduces long-term gains in muscle size relative to active recovery, one of the better-replicated
findings in the [[heat-and-cold-exposure|thermal exposure]] literature.

> [!key] The comparator problem
> A geroprotective drug tested in a sedentary population is competing against a placebo. The harder
> and more informative question is whether it adds anything on top of training. At least two trials
> found that [[metformin]] blunted the mitochondrial and hypertrophic adaptations to exercise in older
> adults, which is the first documented case of a candidate geroprotector subtracting from the
> benchmark.

## Limits

The rodent evidence is where the strong version of the claim weakens. Voluntary wheel running and
forced treadmill protocols generally improve median survival in mice and rats while leaving maximum
lifespan largely unchanged. Under the standard interpretation, that pattern indicates a reduction in
extrinsic mortality and morbidity rather than a change in the rate of aging itself — the opposite of
what [[caloric-restriction]] does in the same animals. Exercise, on this reading, is an extraordinary
health intervention whose credentials as a *rate-of-aging* intervention are weaker than its
credentials as a survival intervention.

Cognitive outcomes are mixed. Some randomized trials of exercise in older adults report improvements
in executive function or hippocampal measures; others, including trials in people who already have
dementia, have found no benefit on cognitive decline. Aggregating them yields a small and
inconsistent effect, which is a reminder that "exercise is good for the brain" is a much better
supported claim about risk than about treatment.

Exercise also does not prevent all age-related disease. Highly fit people still develop cancer,
neurodegeneration, and atherosclerosis, and endurance athletes are not exempt from age-related
mortality. Adherence is the practical ceiling: guideline levels are reached by a minority of adults
in most countries, and physical limitation, time, and environment are the reasons, not ignorance.

## Measurement

Cardiorespiratory fitness, expressed as VO2max, declines by roughly ten per cent per decade in
sedentary adults, with the decline steepening after seventy. Because it integrates cardiac output,
vascular function, muscle oxidative capacity, and pulmonary performance, it functions as one of the
better-validated single-number summaries of physiological reserve, and it predicts outcomes more
strongly than most molecular measures used in [[aging-biomarkers|aging biomarker]] research. Gait
speed, grip strength, and chair-stand time serve a similar role at lower cost and are the endpoints
around which competitions such as [[xprize-healthspan]] are built.

This is a useful corrective to clock-based measurement. A training programme reliably improves
functional capacity while moving [[epigenetic-clock|methylation clock]] estimates little, which
suggests either that the clocks miss something important or that
[[biological-age|composite age scores]] and functional reserve are measuring different things.

## Enhancement and its boundary

The same physiology drives the enhancement debate in athletics. Attempts to reproduce training
adaptations pharmacologically, using AMPK and PPARδ agonists studied as "exercise mimetics",
produced compounds that were banned by anti-doping authorities before they were licensed for any
human indication, one of them after showing carcinogenicity in rodents. No exercise mimetic has been
approved, and the gap between activating a signalling pathway and reproducing the systemic effects
of loading remains wide. The regulatory and ethical questions this raises are treated under
[[gene-doping]] and [[enhancement-in-sport]], while assistive technologies such as
[[exoskeleton|powered exoskeletons]] address the same functional decline from the outside.

## Outlook

The open questions are about specificity and dose. Which components of training (endurance volume,
intensity, resistance loading, balance work) map onto which age-related outcomes, and whether
prescriptions can be individualized by measurable response rather than by age. Whether the
circulating factors released during exercise can be isolated and used in people who cannot train,
which is the population that would benefit most. And whether any pharmacological geroprotector, once
tested properly, adds to what training already provides — a question now most pressing for the
[[glp-1-receptor-agonists|incretin drugs]], whose weight loss includes the lean tissue that
resistance training preserves. Until one does, the honest position is that
the most effective intervention against human aging is also the oldest, the cheapest, and the least
commercially interesting.

## See also

- [[hallmarks-of-aging]]
- [[healthspan]]
- [[compression-of-morbidity]]
- [[caloric-restriction]]
- [[rapamycin]]
- [[aging-biomarkers]]
- [[geroscience-hypothesis]]
- [[enhancement-in-sport]]

## References

[^moore2012]: `paper` Moore, S.C. et al. "Leisure time physical activity of moderate to vigorous intensity and mortality: a large pooled cohort analysis." *PLoS Medicine*, 2012. {Pooled observational cohorts with self-reported activity; the life-expectancy gain is an association, not a trial result.}
[^mandsager2018]: `paper` Mandsager, K. et al. "Association of cardiorespiratory fitness with long-term mortality among adults undergoing exercise treadmill testing." *JAMA Network Open*, 2018. {A retrospective cohort of patients referred for clinical treadmill testing, so not a sample of the general population.}
[^leong2015]: `paper` Leong, D.P. et al. "Prognostic value of grip strength: findings from the Prospective Urban Rural Epidemiology (PURE) study." *The Lancet*, 2015.
[^pahor2014]: `paper` Pahor, M. et al. "Effect of structured physical activity on prevention of major mobility disability in older adults: the LIFE study randomized clinical trial." *JAMA*, 2014. {Randomized against health education in sedentary elders, with a walking endpoint; it was not powered for mortality.}

==============================================================================
ARTICLE: existential-risk
TITLE:   Existential risk
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/existential-risk
SOURCE:  https://futurehumanwiki.com/raw/existential-risk
==============================================================================

---
title: "Existential risk"
slug: "existential-risk"
type: "risk"
status: "contested"
horizon: "indefinite"
categories: ["society", "foundations"]
tags: ["risk", "governance", "biosecurity", "longtermism", "forecasting", "policy"]
summary: "The category of risks that would end humanity or permanently destroy its potential, and the research programme built around identifying and reducing them."
updated: "2026-07-27"
humanEvidence: "No existential catastrophe has occurred, so the anthropogenic probabilities are stated credences rather than estimates from data; only the natural background rate is bounded, by the fossil and geological record."
reversibility: "irreversible"
issues: ["The 2022 asteroid deflection result is stated without a citation."]
---

```infobox
{
  "caption": "Category in risk analysis and philosophy",
  "rows": [
    { "label": "Defined by", "value": "Nick Bostrom, 2002", "link": "/wiki/nick-bostrom" },
    { "label": "Distinguishing feature", "value": "Unrecoverable outcome" },
    { "label": "Leading anthropogenic cases", "value": "Engineered pandemics, AI" },
    { "label": "Natural background", "value": "Very low per century" },
    { "label": "Key monograph", "value": "The Precipice, 2020" },
    { "label": "Institutional home", "value": "Dispersed since 2024" }
  ]
}
```

**Existential risk** is a risk of an outcome that would either annihilate Earth-originating intelligent life or permanently and drastically curtail its potential. The definition is [[nick-bostrom]]'s, from a 2002 paper that gave the field its name, and its distinguishing feature is not severity but irreversibility.[^bostrom2002] A catastrophe that kills a large fraction of humanity and from which the survivors recover is, in this framework, categorically different from one that forecloses the future — however similar the two look in the immediate aftermath.

## The definition and its taxonomy

Bostrom's original classification sorted outcomes by how the potential is lost. **Bangs** are sudden extinction events. **Crunches** leave humanity alive but permanently unable to develop further. **Shrieks** reach a form of posthuman existence realising only a fraction of what was possible. **Whimpers** are gradual erosions. A later restatement replaced the taxonomy with two axes — scope and severity — with existential risk in the far corner of both.[^bostrom2013]

Treating a permanent curtailment of potential as equivalent to extinction requires the view that the value at stake is mostly in the future rather than the present, the premise connecting this research to longtermism and one many critics reject. The related term *global catastrophic risk* is broader and does not require irreversibility.

## The mechanism

What makes a risk existential rather than merely catastrophic is the absence of a recovery path. Three properties tend to produce that.

**Self-propagation.** An engineered pathogen, unlike a bomb, continues to act after release. The same is true of the self-spreading edits described in [[gene-drive]] and of the chirally inverted organisms discussed in [[mirror-life]], where the specific concern is that immune systems evolved against ordinary chirality might not recognise the invader at all. The original self-replication scenario, [[grey-goo]], turned out to rest on engineering assumptions its own author later withdrew — a case worth remembering when assessing the current ones.

**Loss of the capacity to respond.** A sufficiently capable optimising system pursuing objectives that diverge from human ones need not be malicious to be unrecoverable; it need only be better at acting in the world than the people trying to stop it. This is the argument treated in [[artificial-general-intelligence]] and [[technological-singularity]], and it is where the field's attention has concentrated since 2020.

**Lock-in.** A stable global order that permanently prevents change — through surveillance, through control of the means of coordination, or through value entrenchment — produces a crunch without killing anyone. This branch receives the least analysis, partly because it is hard to distinguish in advance from ordinary political failure. It is also where the field's concerns intersect with the trajectories set out in [[future-of-humanity]] and with the question, raised in [[machine-consciousness]], of whether the entities inheriting such an order would be moral patients at all.

Extinction from a single physical event is comparatively hard: humans occupy every continent, and the record includes survived population bottlenecks.

## Plausibility

The most-cited quantitative estimates are Toby Ord's, published in 2020 and presented explicitly as his own subjective credences rather than as calculations.[^ord2020] He puts the probability of an existential catastrophe in the twenty-first century at roughly one in six; unaligned artificial intelligence at roughly one in ten; engineered pandemics at roughly one in thirty; and all natural risks combined at something like one in ten thousand.

The natural-risk figure is the only one with a defensible derivation. A species facing a high per-century probability of extinction from asteroids, supervolcanism, or naturally arising disease would be unlikely to persist for hundreds of thousands of years — and mammalian species typically endure on the order of a million. The fossil and geological record bounds the natural rate from above, and the bound is low.

No comparable derivation exists for the anthropogenic figures, and this is the field's central methodological problem. There is no base rate for technologies never deployed, no frequency data, and no way to calibrate a forecaster on events that happen at most once. Estimates correlate strongly with the estimator's prior commitments. As numbers they carry a precision the reasoning does not support; as ordinal rankings — anthropogenic risks dominate natural ones, and biology and AI dominate within them — they command much broader agreement.

> [!caution] What the probabilities are
> Ord's figures are stated credences from one careful analyst, not scientific estimates. They are frequently quoted without that qualification. Their most defensible use is comparative: they say which risks their author thinks deserve the most attention, not how likely the world is to end.

The biological case is the one where the reasoning is most concrete, because the enabling capabilities are documented rather than projected. [[dual-use-research]] traces the specific experiments — enhanced transmissibility, viral synthesis from published sequence, evasion of countermeasures — that turn a general worry into a mechanism, and the whole-genome construction methods in [[synthetic-genomes]] supply the assembly step. The design tools described in [[ai-protein-design]] raise a further question: whether screening of synthesised sequences can recognise a protein that was designed rather than copied. No comparable chain of demonstrated capabilities exists for any other candidate risk.

## The vulnerable world hypothesis

Bostrom's sharpest contribution to policy is a thought experiment.[^bostrom2019] Imagine technological discovery as drawing balls from an urn: most are white and beneficial, some grey and mixed. The hypothesis asks what happens if the urn contains a black ball — a technology that by default destroys the civilisation that discovers it. His illustrative case is a method for building nuclear weapons from readily available materials, which would make mass destruction available to anyone sufficiently motivated.

The argument's force is in what follows. If such a technology exists and is drawn, only four responses are available: stop drawing, ensure no actor has the motive, achieve preventive policing effective enough to intercept every attempt, or achieve global governance capable of coordinating a halt. Bostrom notes without enthusiasm that the last two require surveillance and centralisation most people would regard as intolerable, and that the first two are not achievable. The paper is best read as a demonstration that a certain class of world admits no acceptable solution, and therefore as an argument for finding out early whether this is such a world.

## Criticism

**The numbers are unfalsifiable.** A probability assigned to a unique unobserved event cannot be scored, and there is no mechanism by which persistently wrong estimates would be corrected.

**Pascal's mugging.** Multiplying a very small probability by an astronomically large value can make any intervention appear to dominate, a decision-theoretic pathology that Bostrom himself named. Any framework valuing the whole future is vulnerable to it, and no accepted resolution exists.

**The values are contested.** Émile Torres and Timnit Gebru have argued that the cluster of ideas linking existential risk, longtermism, and [[transhumanism]] rests on assumptions about what counts as valuable — vast future populations, expansion beyond Earth, digital minds — that are neither neutral nor widely shared, and that the framework licenses discounting present harms against speculative future ones.[^torres2024] Defenders reply that reducing extinction risk is supported by almost any value system and does not require the longtermist premises.

**Institutional credibility.** The field's funding and reputation were damaged by the 2022 collapse of a major donor in the effective-altruism ecosystem, and Oxford's Future of Humanity Institute — its original home — was closed by the university in 2024 after protracted administrative disputes. Work continues at Cambridge's Centre for the Study of Existential Risk, at the Future of Life Institute, and at biosecurity centres, but the centre of gravity has moved toward AI policy organisations with narrower remits.

## Mitigation and governance

The field's characteristic policy proposal is [[differential-technological-development]]: rather than accelerating or halting technology in general, deliberately advance defensive and stabilising capabilities ahead of destabilising ones. Applied to biology this means faster pathogen detection, broad-spectrum countermeasures, and better protective equipment before further transmissibility research. It is a coherent principle and faces an obvious coordination problem, since sequencing requires agreement among actors who do not agree. It is also the field's answer to the objection that its position amounts to a blanket [[precautionary-principle]]: the proposal is to change the order in which capabilities arrive, not to stop them arriving.

Existing instruments are thin. The Biological Weapons Convention has no verification mechanism, and nuclear arms control has eroded. AI governance assembled rapidly and lightly: the EU's AI Act entered into force in 2024, national AI safety and standards bodies were established in the UK and US from late 2023, and intergovernmental summits from Bletchley through Seoul and Paris produced declarations rather than obligations. The UN's 2024 Summit of the Future adopted a Pact for the Future including a declaration on future generations — the first UN document of that standing to carry an existential-risk framing, and non-binding.

Planetary defence is the one area with a demonstrated capability: a spacecraft impact in 2022 measurably altered an asteroid's orbital period, establishing that deflection works for the class of object tested. It is also the least severe of the risks, which says something about the relation between tractability and importance here.

## Open problems

The unresolved question is whether existential risk can be made an ordinary policy category. Governments manage risks by frequency and expected cost, and a risk that has never occurred, cannot be insured, and would leave nobody to assign blame fits none of those procedures. The proposals for institutional fixes — standing bodies for future generations, mandatory long-horizon impact assessment, an international scientific panel on AI of the kind the UN General Assembly moved to establish in 2025 — are all attempts to create a constituency for outcomes that no living voter will experience. None has yet been shown to work.

## See also

- [[nick-bostrom]]
- [[dual-use-research]]
- [[differential-technological-development]]
- [[mirror-life]]
- [[precautionary-principle]]
- [[artificial-general-intelligence]]
- [[future-of-humanity]]
- [[grey-goo]]

## References

[^bostrom2002]: `paper` Bostrom, N. "Existential Risks: Analyzing Human Extinction Scenarios and Related Hazards." *Journal of Evolution and Technology*, 2002. {The paper that named the field appeared in a small transhumanist journal, not in a mainstream risk or policy venue.}
[^bostrom2013]: `paper` Bostrom, N. "Existential Risk Prevention as Global Priority." *Global Policy*, 2013.
[^ord2020]: `book` Ord, T. *The Precipice: Existential Risk and the Future of Humanity*. Bloomsbury, 2020. {A trade book for a general readership; the figures are one author's stated credences, not a peer-reviewed risk assessment.}
[^bostrom2019]: `paper` Bostrom, N. "The Vulnerable World Hypothesis." *Global Policy*, 2019.
[^torres2024]: `paper` Gebru, T. and Torres, É.P. "The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence." *First Monday*, 2024. {A critique of the ideological cluster around the field rather than of any particular risk estimate.}

==============================================================================
ARTICLE: extropianism
TITLE:   Extropianism
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/extropianism
SOURCE:  https://futurehumanwiki.com/raw/extropianism
==============================================================================

---
title: "Extropianism"
slug: "extropianism"
type: "concept"
status: "historical"
horizon: "historical"
categories: ["people"]
tags: ["transhumanism", "movements", "libertarianism", "extropy", "intellectual history", "mailing lists"]
summary: "The first organized transhumanist movement, built around a set of optimistic principles for self-directed human transformation and dissolved as an institution in 2006."
updated: "2026-07-27"
issues: ["The mailing-list section names many participants without a citation.", "The Extropy Institute's 2006 closing statement is paraphrased with no source."]
---

```infobox
{
  "caption": "Intellectual movement",
  "rows": [
    { "label": "Founded", "value": "1988 (journal), 1992 (institute)" },
    { "label": "Founders", "value": "Max More, Tom Bell", "link": "/wiki/max-more" },
    { "label": "Core text", "value": "The Principles of Extropy" },
    { "label": "Institution", "value": "Extropy Institute" },
    { "label": "Dissolved", "value": "2006" },
    { "label": "Successor body", "value": "Humanity+", "link": "/wiki/humanity-plus" },
    { "label": "Political tendency", "value": "Libertarian, technophilic" }
  ]
}
```

**Extropianism** is the first organized [[transhumanism|transhumanist]] movement, formulated in California in the late 1980s around a set of principles favouring perpetual self-improvement, rational optimism, decentralized order, and the deliberate use of technology to overcome biological limits. Its institutional life was short — the Extropy Institute closed in 2006 — but its mailing list functioned as an incubator for a set of ideas and people that went on to shape [[artificial-general-intelligence|AI]] risk research, cryptocurrency, and the modern longevity industry.

## The name

"Extropy" was coined as an informal antonym of entropy: a measure of a system's intelligence, functional order, vitality, and capacity for improvement. Its authors were explicit that it is not a physical quantity and not measurable, and warned against treating it as one.[^more1998] The term functioned as a banner rather than a concept, signalling that the movement's orientation was toward growth and organization in a universe that tends the other way.

## Origins

Max More, then a philosophy graduate student who had changed his name from Max O'Connor, and Tom Bell, writing as T. O. Morrow, launched *Extropy: The Journal of Transhumanist Thought* in 1988. Its early issues collected material on [[cryonics]], nanotechnology as [[eric-drexler]] had described it, [[nootropics]], space settlement, artificial intelligence, and libertarian political economy — a combination that had existed as scattered enthusiasms and had not previously been given a name and a coherent philosophy.

More's 1990 essay "Transhumanism: Toward a Futurist Philosophy" articulated the position as a philosophy of life rather than a technology forecast, and set out the first version of the principles.[^more1990] The Extropy Institute was incorporated in 1992 as a nonprofit to run conferences and publish the journal. Its meetings, held through the 1990s, were among the first venues where cryonicists, AI researchers, nanotechnology advocates, and life-extension proponents met as a single community. Ed Regis's 1994 *Wired* profile brought the group to a general readership, and fixed a public image — chosen names, self-experimentation, and a taste for the ninety-ninth percentile of technological optimism — that the movement never entirely shed.[^regis1994]

```timeline
[
  { "year": "1988", "title": "Journal launched", "text": "Max More and Tom Bell publish the first issue of Extropy, giving a scattered set of technophilic enthusiasms a shared name." },
  { "year": "1990", "title": "The philosophy stated", "text": "More's 'Transhumanism: Toward a Futurist Philosophy' sets out extropy as a stance rather than a forecast, and drafts the principles." },
  { "year": "1991", "title": "The extropians mailing list", "text": "An email list is established that becomes the movement's real centre of gravity for the next decade." },
  { "year": "1992", "title": "Extropy Institute founded", "text": "A California nonprofit is incorporated to publish the journal and run conferences." },
  { "year": "1994", "title": "Mainstream attention", "text": "Ed Regis's Wired feature 'Meet the Extropians' introduces the movement to a wide audience, in a tone that is affectionate and skeptical in roughly equal measure." },
  { "year": "1998", "title": "A rival institution", "text": "Nick Bostrom and David Pearce found the World Transhumanist Association, positioning transhumanism as academically respectable and politically non-aligned." },
  { "year": "2004", "title": "The proactionary principle", "text": "More drafts an explicit counterweight to the precautionary principle at an Extropy Institute summit." },
  { "year": "2006", "title": "Institute closes", "text": "The board announces the organization's mission is complete, arguing that its ideas have dispersed into the wider culture and no longer need a dedicated body." }
]
```

## The principles

The Principles of Extropy went through several revisions between 1990 and 2003. The final version listed seven: perpetual progress, self-transformation, practical optimism, intelligent technology, open society (information and democracy), self-direction, and rational thinking.[^more2003] Earlier versions included "boundless expansion", "dynamic optimism" and "spontaneous order", the last drawn directly from Friedrich Hayek.

The principles were framed as an attitude to be adopted rather than propositions to be verified, and the documents said so. Several are worth distinguishing because they became points of divergence within transhumanism:

- **Self-transformation** holds that a person may legitimately redesign their own body, mind and character, and is the ancestor of the [[morphological-freedom]] argument.
- **Practical optimism** was explicitly distinguished from faith or blind hope; it asserts that optimism is instrumentally useful because pessimism suppresses effort, not that good outcomes are likely.
- **Spontaneous order**, and the open-society principle that replaced it, encoded a preference for decentralized coordination over central planning that made early extropianism politically libertarian in a way that later transhumanist organizations deliberately were not.

More's [[precautionary-principle|proactionary principle]], drafted in 2004, extended this into risk policy: it asks decision-makers to weigh the costs of restricting a technology, including the lives lost to its absence, against the costs of permitting it, rather than treating inaction as the default safe option.

> [!note] Terminology
> Extropianism is not a synonym for transhumanism. It is one movement within it, distinguished by an explicit libertarian politics and a stated attitude of optimism. Later transhumanist organizations such as [[humanity-plus]] were founded partly to separate the broader philosophical position from these commitments.

## The mailing list

The extropians email list, running from 1991, mattered more than the institute. It was a small, high-volume forum with an unusual concentration of people who later became consequential in separate fields.

Participants included [[nick-bostrom]] and [[anders-sandberg]], who carried the [[existential-risk]] and [[whole-brain-emulation]] research programmes into academia; Eliezer Yudkowsky, whose arguments about AI risk grew out of and then against the list's optimism; the economist Robin Hanson; [[ray-kurzweil]] and Drexler as occasional presences; and the cryptographers Hal Finney, Nick Szabo and Wei Dai. Szabo's bit gold and Dai's b-money proposals, both direct antecedents of Bitcoin, were circulated in this milieu, and Finney received the first Bitcoin transaction and was later cryopreserved at [[alcor]].

This concentration is the strongest claim extropianism has on historical attention. A list with a few hundred subscribers produced, within twenty years, the intellectual seeds of contemporary AI safety research, of cryptocurrency, and of a substantial part of the [[longevity-escape-velocity|life-extension]] funding ecosystem.

## Dissolution and afterlife

The Extropy Institute closed in 2006. Its board's stated reason was that the mission had succeeded: the ideas no longer needed an advocacy organization because they had entered general circulation. A less generous reading is that the institute had been overtaken. The World Transhumanist Association, founded in 1998 by Bostrom and David Pearce and later renamed Humanity+, offered an academically credentialed and politically neutral version of the same programme, and drew both members and legitimacy away from a body with an explicit libertarian identity.

More went on to lead Alcor for most of the 2010s. [[natasha-vita-more]], who had developed a parallel strand of transhumanist art and body design and who married More, took leadership roles in Humanity+ and edited the movement's principal anthology.[^reader2013]

The ideas dispersed rather than died. Contemporary [[effective-accelerationism]] restates practical optimism and the proactionary principle in the vocabulary of thermodynamics and startup culture, largely without acknowledging the lineage. [[biohacking|Grinder]] subculture inherited self-transformation as a physical practice. [[singularitarianism]] took the technology forecast and dropped the politics. The [[posthuman|posthuman]] end-state that extropian writing described is now discussed mainly in bioethics, where it is treated as a claim to be evaluated rather than a goal to be pursued.

## Criticism

Three lines of criticism have held up.

The first is that the movement's optimism was a stance rather than a conclusion, and that stating this openly does not immunize it. Deciding in advance to expect good outcomes is a poor procedure for estimating probabilities, and the extropian forecast record — [[medical-nanorobots|nanomedicine]], [[mind-uploading]], and radical life extension all expected within decades — reflects it.

The second is political. Bioconservative critics and left-leaning transhumanists agreed, from opposite directions, that a movement combining private [[human-enhancement|enhancement]] with hostility to regulation had no account of who would be left behind. The [[access-and-inequality]] problem was visible in extropian writing mainly as an objection to be rebutted. James Hughes's argument for a democratic transhumanism was written explicitly against this tendency.

The third is that "extropy" itself never did any work. Presented as the movement's organizing concept, it turned out to be a mood rather than a measure, and the documents defining it conceded the point.

## Legacy

Extropianism's significance is as a social formation rather than a body of theory. It assembled, named, and networked a set of positions that had previously been separate hobbies, and it did so early enough that the people it assembled were still young. Judged by its own principles it failed: the technologies it anticipated arrived slowly or not at all, and the institution dissolved. Judged by the second-order question of which conversations it started, its influence on AI risk, digital currency and longevity funding is out of all proportion to its size — which raises the harder question of whether the small, self-selected online community remains an efficient way to originate ideas, or whether that was a feature of a particular decade of the internet.

## See also

- [[transhumanism]]
- [[max-more]]
- [[natasha-vita-more]]
- [[humanity-plus]]
- [[singularitarianism]]
- [[effective-accelerationism]]
- [[morphological-freedom]]
- [[precautionary-principle]]

## References

[^more1990]: `paper` More, M. "Transhumanism: Toward a Futurist Philosophy." *Extropy*, 1990. {Published in the movement's own journal rather than an academic venue, and framed as a stance to adopt rather than a claim to test.}
[^more1998]: `statement` More, M. "The Extropian Principles: A Transhumanist Declaration," version 3.0, Extropy Institute, 1998.
[^more2003]: `statement` More, M. "Principles of Extropy," version 3.11, Extropy Institute, 2003.
[^regis1994]: `news` Regis, E. "Meet the Extropians." *Wired*, October 1994.
[^reader2013]: `book` More, M. and Vita-More, N. (eds.) *The Transhumanist Reader*. Wiley-Blackwell, 2013.

==============================================================================
ARTICLE: fm-2030
TITLE:   FM-2030
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/fm-2030
SOURCE:  https://futurehumanwiki.com/raw/fm-2030
==============================================================================

---
title: "FM-2030"
slug: "fm-2030"
type: "person"
status: "historical"
horizon: "historical"
categories: ["people"]
tags: ["transhumanism", "futurism", "cryonics", "identity", "life extension"]
summary: "Iranian-American futurist, born Fereidoun M. Esfandiary, who defined the transhuman as a transitional being and was cryopreserved on his death in 2000."
updated: "2026-07-28"
issues: ["Sources disagree on when he took the name FM-2030; the article gives the range rather than a date."]
---

```infobox
{
  "caption": "Futurist and novelist",
  "rows": [
    { "label": "Born", "value": "15 October 1930, Brussels" },
    { "label": "Born as", "value": "Fereidoun M. Esfandiary" },
    { "label": "Nationality", "value": "Iranian-American" },
    { "label": "Known for", "value": "The transhuman as transitional being", "link": "/wiki/transhumanism" },
    { "label": "Key work", "value": "Are You a Transhuman? (1989)" },
    { "label": "Taught at", "value": "The New School; UCLA Extension; Florida International University" },
    { "label": "Died", "value": "8 July 2000, New York City" },
    { "label": "Status", "value": "Cryopreserved at Alcor", "link": "/wiki/alcor" }
  ]
}
```

**FM-2030** was an Iranian-American futurist, novelist and teacher, born Fereidoun M. Esfandiary,
who argued that some people alive in the twentieth century were already transitional beings — no
longer fully bound by the biological, national and family categories that had defined human life,
and not yet whatever comes next. He called such people transhumans, and the usage passed into
[[transhumanism]] as a movement. He took the name FM-2030 in place of the one he was born with, a
designation pointing at the year he would turn one hundred.

## Life

Esfandiary was born in Brussels in October 1930 to Iranian diplomats and spent his childhood moving
between postings in Europe and Asia; biographical accounts repeat his claim to have lived in
seventeen countries as a child.[^olympedia] He played basketball for Iran at the 1948 Olympic Games
in London,[^olympedia] and from 1952 to 1954 served on the United Nations Conciliation Commission
for Palestine.[^nypl] That combination of early rootlessness and direct exposure to the machinery
of international institutions shaped a lifelong argument that national and ethnic identity was a
residue rather than a fact.

He published three novels in English: *The Day of Sacrifice* (1959), *The Beggar* (1965) and
*Identity Card* (1966), the last a study of bureaucratic identity in Iran. From the late 1960s he
turned to futurism, teaching long-running courses on the future at The New School in New York, at
UCLA Extension and later at Florida International University.[^nypl] His futurist books are
*Optimism One* (1970), *Up-Wingers: A Futurist Manifesto* (1973), *Telespheres* (1977) and *Are You
a Transhuman?* (1989).

Accounts of the name disagree. The New York Public Library, which holds his papers, dates the
change to 1988; a profile in *Bidoun* places the adoption around 1985 with the paperwork filed in
1988; shorter biographical sketches give 1970.[^nypl][^tiven2010][^olympedia] The stated reason is
consistent across them: a surname presumed a family and a nationality he did not accept, and 2030
marked the hundredth year after his birth.[^tiven2010]

## Key ideas

### Upwing politics

FM-2030 held that the left–right axis was obsolete because both wings argued about the distribution
of a fixed set of goods within a fixed human condition. He proposed an "upwing" orientation
concerned instead with expanding what is possible: abolishing scarcity, disease and death rather
than redistributing their burdens.[^esfandiary1973] The framing anticipates arguments now made under
[[effective-accelerationism]] and, in a more careful form, under [[longevity-dividend|economic]]
cases for aging research, though FM-2030's version was closer to a countercultural utopianism than
to either.

### The transhuman as transitional being

His most durable idea is definitional. A transhuman, in his usage, is not an enhanced person but a
person whose way of living already points beyond the traditional human template. The markers he
listed were mundane and deliberately so: prosthetics and implants, reproductive technology used
without embarrassment, life without fixed geographic roots, non-religious orientation, androgyny,
comfort with telecommunication as a primary mode of contact, and a rejection of the family as
destiny.[^bostrom2005] The claim was that transition is gradual and already underway, not a future
event.

> [!note] Two meanings of "transhuman"
> FM-2030's transhuman is a *transitional human* — a person in the middle of a species-scale change.
> The later movement's usage folds this into a general commitment to enhancement, and the
> intermediate status between human and [[posthuman]] is what survives. The two senses are often
> conflated.

### The self-assessment book

*Are You a Transhuman?* (1989) is structured as a set of self-assessment instruments, asking
readers to score their own rate of change, exposure to new technology, physical modification,
mobility and openness to future institutions.[^fm1989] The tests are not psychometrics in any validated sense, and were not offered
as such; their purpose was to make the argument concrete by forcing the reader to locate themselves
on a trajectory. The device also explains much of the book's reach outside academic futures studies:
it turned an abstract thesis into something a reader could do.

## Death and cryopreservation

FM-2030 died of pancreatic cancer in Manhattan on 8 July 2000 and was cryopreserved by the
[[alcor|Alcor Life Extension Foundation]] in Scottsdale, Arizona.[^nyt2000] He had been an advocate
of [[cryonics]] and had argued that death from disease should be treated as a solvable engineering
failure rather than an occasion for consolation. His choice is often cited as the movement's most
literal statement of principle, and as an illustration of the gap the practice depends on. An
account of protocol history published in Alcor's own library dates the switch from glycerol
perfusion to vitrification to late 2000,[^dewolfprotocols] so the procedures available when he died
were not those of current [[brain-preservation]] work, and no revival of a cryopreserved human has
been demonstrated or is close to being demonstrated.

He did not live to see the year he named. Nothing about 2030 in his writing amounts to a technical
forecast; he described it as a symbolic marker, a period he expected to be "a magical time" rather
than a dated prediction.[^olympedia]

## Reception and legacy

FM-2030 was a public figure in a way later transhumanists mostly have not been, appearing on
American television — including a long session with Larry King — as a professional optimist at a
moment when futurism was dominated by resource-depletion pessimism.[^tiven2010] Critics then and
since have noted that his work is exhortation rather than analysis: it contains almost no
quantitative argument, no engagement with the biology of aging as it stood, and no account of how
any of the changes it anticipates would be paid for or governed. Even sympathetic historians of the
movement have found the transhuman criteria arbitrary, since nothing in them explains why cosmetic
surgery and frequent air travel should put a person closer to posthumanity than anyone
else.[^bostrom2005]

What survived is the vocabulary and the stance. [[max-more]] and the [[extropianism|extropians]]
took the transhuman terminology and gave it a philosophical structure; [[natasha-vita-more]], who
worked alongside him in the Los Angeles futurist scene, carried the aesthetic and design side of
the project forward. His deeper claim — that identity is a variable rather than a fixed inheritance
— now runs through arguments over [[morphological-freedom]] and
[[personal-identity-and-continuity]] that are considerably more rigorous than anything he wrote, and
that have not settled the question he raised: whether a person mid-transition is still the same
person at the end of it.

His prediction record is poor in the ordinary way and interesting in a specific one. The 2030 date
attached to his name was never a forecast so much as a target, and almost none of the concrete
changes he expected by it — the end of nation-states, the dissolution of the nuclear family into
fluid affiliation, the routine reversal of aging — is close. What he got closer to is the direction
of travel on identity: the assumption that ethnicity, nationality, family structure and eventually
biology are things a person negotiates rather than inherits is now unremarkable across much of the
world he wrote for, arrived at through social change rather than the technology he expected to
drive it.

That gap is the most useful thing about him. The futurist habit he exemplified — reading a
cultural shift as evidence that a technological one is imminent — recurs throughout this field, and
the record suggests the two decouple more often than not. Read as a diagnosis of what people were
becoming willing to want, his work holds up; read as a schedule, it does not, and he insisted on
being read as a schedule.

## See also

- [[transhumanism]]
- [[posthuman]]
- [[cryonics]]
- [[alcor]]
- [[max-more]]
- [[natasha-vita-more]]
- [[extropianism]]
- [[morphological-freedom]]

## References

[^olympedia]: `report` Olympedia. "Fereidoun Esfandiary." Athlete record, olympedia.org. {Confirms the 1948 basketball appearance and the birth date; the attached biographical note is the database's own summary, not a primary document.}
[^nypl]: `report` Manuscripts and Archives Division, The New York Public Library. "F. M. Esfandiary / FM-2030 papers." Collection finding aid. {The repository that holds his papers; it gives the teaching dates, the UN commission years, and 1988 for the name change.}
[^nyt2000]: `news` Martin, D. "Futurist Known as FM-2030 Is Dead at 69." *The New York Times*, 11 July 2000.
[^tiven2010]: `news` Tiven, B. "The Future Takes Forever." *Bidoun*, issue 19, Winter 2010.
[^bostrom2005]: `paper` Bostrom, N. "A History of Transhumanist Thought." *Journal of Evolution and Technology*, vol. 14, 2005. {Bostrom writes as a participant in the movement whose history he traces, which makes the passage criticising FM-2030's criteria an internal objection rather than an outside one.}
[^esfandiary1973]: `book` Esfandiary, F. M. *Up-Wingers: A Futurist Manifesto*. John Day, 1973.
[^fm1989]: `book` FM-2030. *Are You a Transhuman?: Monitoring and Stimulating Your Personal Rate of Growth in a Rapidly Changing World*. Warner Books, 1989.
[^dewolfprotocols]: `statement` de Wolf, A. and Phaedra, C. "Advances in Cryonics Protocols, 1990–2006." Alcor Life Extension Foundation library. {Written by cryonics practitioners and published by the organization whose protocols it describes; it dates the switch to vitrification to late 2000.}

==============================================================================
ARTICLE: future-of-humanity
TITLE:   Future of humanity
PORTAL:  Foundational Concepts
URL:     https://futurehumanwiki.com/wiki/future-of-humanity
SOURCE:  https://futurehumanwiki.com/raw/future-of-humanity
==============================================================================

---
title: "Future of humanity"
slug: "future-of-humanity"
type: "concept"
status: "speculative"
horizon: "indefinite"
categories: ["foundations", "society"]
tags: ["forecasting", "existential risk", "longtermism", "demography", "trajectories", "philosophy"]
summary: "The study of humanity's possible long-run trajectories, conventionally grouped into extinction, recurrent collapse, plateau, and posthuman transformation."
updated: "2026-07-27"
issues: ["The demographic figures in keyfacts and the life-expectancy claim carry no footnote."]
---

```infobox
{
  "caption": "Field of inquiry in philosophy and futures studies",
  "rows": [
    { "label": "Standard taxonomy", "value": "Four scenario families" },
    { "label": "Proposed by", "value": "Nick Bostrom, 2009", "link": "/wiki/nick-bostrom" },
    { "label": "Method", "value": "Scenario analysis, base rates, physical limits" },
    { "label": "Best-known survey", "value": "Ord, The Precipice, 2020" },
    { "label": "Chief difficulty", "value": "No reference class" },
    { "label": "Predictive track record", "value": "Poor at all horizons" }
  ]
}
```

**Future of humanity** names the attempt to reason systematically about which long-run trajectories are open to the human species and its descendants, and how likely each is. It is not forecasting in the ordinary sense: the events at issue are unprecedented, the relevant timescales exceed any dataset, and the most consequential variables are the least predictable. What the field offers instead is a discipline of constraint — physical limits that no technology can exceed, historical base rates that any specific claim must beat, and a taxonomy that keeps arguments from talking past each other.

```keyfacts
[
  { "value": "4", "label": "Scenario families in the standard taxonomy", "note": "extinction, recurrent collapse, plateau, posthumanity" },
  { "value": "~10.3bn", "label": "Projected peak world population", "note": "UN World Population Prospects, 2024 revision; peak in the mid-2080s" },
  { "value": "~1 in 6", "label": "Ord's estimate of existential catastrophe this century", "note": "an explicit subjective judgement, not a measurement" }
]
```

## Four trajectories

[[nick-bostrom]]'s 2009 essay set out the taxonomy the field still uses.[^bostrom2009] Every long-run scenario, he argued, falls into one of four families, distinguished by where the trajectory ends up rather than by how it gets there.

**Extinction.** Humanity ceases to exist, permanently. The category includes both sudden events and slow declines, and its defining feature is irreversibility: unlike every other outcome, it forecloses all subsequent ones. Derek Parfit's argument that the moral difference between near-extinction and extinction is far larger than the difference in immediate casualties turns on exactly this asymmetry.[^parfit1984]

**Recurrent collapse.** Civilisation repeatedly reaches roughly its present level of technology and repeatedly falls back, without either dying out or advancing further. This scenario is unusual in requiring a mechanism that both prevents permanent extinction and prevents permanent advance, and Bostrom notes that no plausible mechanism has been described. Resource depletion works against it: a collapsed civilisation would rebuild without the accessible fossil energy that fuelled the first ascent.

**Plateau.** Development continues to some ceiling and stops there, with technology and human capacities remaining broadly at levels a person today would recognise. The plateau might be imposed by physics, by economics, or by deliberate global choice. It is the implicit assumption of most economic and demographic projection, and it is the scenario that receives the least explicit defence.

**Posthumanity.** At least one central human capacity is raised far beyond its current maximum, producing beings that fall outside the current concept of human — the condition defined in [[posthuman]]. This family includes radically extended healthspan, greatly enhanced cognition, and non-biological substrates.

The taxonomy is exhaustive by construction and says nothing about probability. Its value is in forcing an argument to specify which transition it is claiming, and what would have to be true for that transition to occur.

## What is reasonably well established

A short list of things about the next several decades is on firm ground, and it is worth separating from everything else in this article.

Global population growth is decelerating. Fertility has fallen below replacement in more than half of all countries, and standard projections show a peak around the 2080s followed by decline — a reversal of the assumption that shaped twentieth-century futurism, and one with direct bearing on the objection examined in [[overpopulation-and-longevity]]. Ageing populations, not growing ones, define the demographic problem of the next fifty years, which is the premise of the argument in [[longevity-dividend]].

Health gains have shifted from mortality to morbidity. Global life expectancy at birth is around seventy-three years, most of the twentieth century's gains came from reducing early-life mortality, and further gains must come from late-life disease — where progress has been slower and where the [[compression-of-morbidity|compression of morbidity]] that some public-health models predict has not clearly materialised.

Physical limits are calculable and binding. Energy budgets, waste-heat dissipation, the speed of light, and the accelerating expansion of the universe together bound what any civilisation can do; [[anders-sandberg]] and colleagues have argued that intergalactic settlement is not obviously ruled out by physics, while the reachable volume shrinks over cosmological time.[^armstrong2013] These constraints are among the few genuinely reliable inputs to long-horizon reasoning.

Climate change is a well-characterised risk that is not, on mainstream assessment, an extinction risk. It is a large-scale harm to human welfare and ecosystems and a plausible contributor to instability, which is a different claim from civilisational termination and should not be merged with it.

## The technological levers

This wiki covers the specific capabilities through which any transition out of the plateau would have to run. Their maturity varies enormously, and the variation matters more than the list.

*Aging.* The [[geroscience-hypothesis]] holds that the processes catalogued in [[hallmarks-of-aging]] drive most chronic disease jointly, so that intervening upstream would compress many diseases at once. Multiple interventions extend lifespan in mice. None has demonstrated an effect on human aging in a controlled trial, and the interventions with the best human evidence remain exercise and the avoidance of known harms. [[longevity-escape-velocity]] is the limiting case of this lever and is rejected as a near-term prospect by most biogerontologists.

*Genomes.* [[crispr-cas9]] and its successors made precise editing routine in the laboratory and have produced approved somatic therapies. Heritable modification is a separate matter, restricted almost everywhere and covered in [[germline-editing]]; selection rather than editing is what is commercially available, and polygenic embryo screening can shift the odds on common traits only slightly, because thousands of variants of small effect underlie the traits people most want to influence.

*Reproduction.* [[in-vitro-gametogenesis]] would, if it worked in humans, decouple gamete supply from the ovary and change the arithmetic of embryo selection completely. It works in mice. [[artificial-womb]] devices support extremely premature lambs and have not been tried in humans.

*Nervous systems.* [[brain-computer-interface|Brain–computer interfaces]] restore communication and control to people with paralysis at rates well below unimpaired function. They do not currently enhance healthy users, and the arguments for merging human and machine cognition rest on capabilities that do not exist.

*Minds.* [[whole-brain-emulation]] and mind uploading depend on two premises — that the relevant causal structure is capturable by achievable imaging, and that mind is substrate-independent — which are respectable positions rather than established results. Most neuroscientists regard both as very distant.

*Bodies and matter.* [[organ-bioprinting]] and xenotransplantation address a defined shortage, with real clinical progress and unsolved problems in vascularisation and immunology. Nanomedicine has produced working DNA-based devices in the laboratory and nothing resembling the cell-scale machines proposed in the 1980s.

*Space.* Decades of spaceflight medicine document what microgravity and radiation do to human physiology over months; the engineered adaptations imagined under the heading of pantropy, and the multigenerational biology a generation ship would require, are unstudied outside fiction and thought experiment.

*Machine intelligence.* [[artificial-general-intelligence]] is the lever most likely to change the others, because it acts on research productivity rather than on any single problem. It is also the one whose timeline is least constrained by evidence.

> [!key] The asymmetry between levers
> Every lever above except machine intelligence has a bounded effect on a specific system. AI is the only candidate that could plausibly compound the rate at which the others advance, which is why AGI timelines dominate every serious long-horizon estimate — and why the weakness of the evidence about them propagates into everything else.

## The risks

The case for treating this as urgent rather than academic rests on the claim that anthropogenic risks now exceed natural ones by a wide margin. Toby Ord's survey estimates roughly a one-in-six chance of existential catastrophe this century, dominated by unaligned artificial intelligence and engineered pandemics, with natural risks such as asteroid impact and supervolcanism orders of magnitude smaller.[^ord2020] The estimates are explicitly subjective and have been criticised as unfalsifiable; the *ranking* is more robust than the numbers, since natural risk rates can be estimated from the geological record and have evidently been low enough to permit two hundred thousand years of human existence.

Engineered pathogens are the clearest biological case, and the mechanisms are covered under [[dual-use-research|dual-use research of concern]] and [[mirror-life]]. Martin Rees's earlier assessment reached a similar conclusion about the century by a different route, emphasising the falling cost and rising accessibility of destructive capability.

The countervailing observation is that the field has a systematic bias toward the dramatic. Slow trajectories — declining fertility, entrenched inequality, institutional decay, value lock-in through durable technology — are less discussed than sudden ones, and there is no reason to think they matter less.

## Reasoning across long horizons

Four methodological commitments distinguish careful work in this area from speculation.

**Separate mechanism from timeline.** The argument that a capability is physically possible is independent of any claim about when it will exist. Most disputes conflate them, and the timeline half has by far the worse track record, as [[accelerating-change]] documents.

**Use base rates, and admit when none exists.** Reference-class forecasting works when a reference class exists. For extinction, it does not, and the historical absence of extinction is weak evidence because of observation selection: only surviving civilisations produce records.

**Beware conjunctive detail.** Detailed scenarios feel more plausible and are strictly less probable than their components, a well-documented reasoning failure. A five-step story about how a technology transforms society is less likely than any of its steps.

**Distinguish demonstrated from extrapolated.** The recurring failure in this literature is treating mouse results, laboratory demonstrations and design studies as though they were positions on a timeline rather than facts about what has been done. Maturity scales such as the technology readiness levels used throughout this wiki are a blunt instrument for this, and better than nothing.

> [!caution] What the record shows
> Confident long-horizon technology predictions have been wrong in both directions and at roughly comparable rates. Nuclear energy, space settlement and machine translation were all over-forecast in the mid-twentieth century; the internet, genomic sequencing costs and [[ai-protein-design|protein structure prediction]] were all under-forecast. The lesson is not that optimists are wrong but that specific dates carry almost no information.

## Where the disagreements actually are

Participants agree on far more than the public argument suggests. Nearly everyone accepts that the technologies covered here are physically possible in some form, that timelines are uncertain, and that irreversible outcomes deserve extra weight. The live disputes are narrower.

*How much do we owe the far future?* Longtermist arguments hold that the number of potential future people is so large that reducing existential risk dominates other priorities. Critics object that this reasoning is dominated by tiny probabilities of enormous values, and that it licenses present sacrifices for speculative benefit.

*Is technological progress the right lever?* Transhumanists and accelerationists treat capability growth as the primary route to good outcomes. Bioconservatives and the [[precautionary-principle]] tradition hold that some capabilities should not be developed. A third position proposes to alter the ordering of arrivals rather than the total, which most participants nominally endorse and few institutions implement.

*Who benefits?* The distributional question examined in [[access-and-inequality]] cuts across all four trajectories. A posthuman transition available to a small fraction of people is a different outcome from a universal one, and the taxonomy above does not distinguish them.

## Open problems

Three problems are unresolved and structurally hard rather than merely unfinished.

The **plateau's mechanism** is undertheorised. If neither collapse nor transformation occurs, something must be holding development at a ceiling for a very long time, and no one has described what. The scenario receives the least attention precisely because it is the least interesting to argue about, which is a poor reason.

The **evidence base for probability estimates** does not exist. Numbers like one in six are elicited judgements presented in the notation of measurement, and the field has not developed a method for improving them that is testable within a human lifetime. Forecasting tournaments have shown short-horizon skill; nothing shows that skill extends to century-scale questions about unprecedented events.

The **observer selection problem** is unsolved and it undermines the most common empirical argument. That humanity has survived every past hazard tells us little, because we could not be here to observe otherwise — so the historical record cannot distinguish a safe world from a lucky one, and the sharpest disagreement in the field is over how much that consideration should move any estimate.

## See also

- [[existential-risk]]
- [[posthuman]]
- [[nick-bostrom]]
- [[differential-technological-development]]
- [[technological-singularity]]
- [[artificial-general-intelligence]]
- [[longevity-escape-velocity]]
- [[transhumanism]]

## References

[^bostrom2009]: `book` Bostrom, N. "The Future of Humanity." In Berg Olsen, J.-K., Selinger, E. and Riis, S. (eds), *New Waves in Philosophy of Technology*. Palgrave Macmillan, 2009.
[^parfit1984]: `book` Parfit, D. *Reasons and Persons*. Oxford University Press, 1984.
[^armstrong2013]: `paper` Armstrong, S. and Sandberg, A. "Eternity in Six Hours: Intergalactic Spreading of Intelligent Life and Sharpening the Fermi Paradox." *Acta Astronautica*, 2013. {An argument about physical limits that assumes technologies, self-replicating probes among them, which do not exist.}
[^ord2020]: `book` Ord, T. *The Precipice: Existential Risk and the Future of Humanity*. Bloomsbury, 2020.

==============================================================================
ARTICLE: gene-doping
TITLE:   Gene doping
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/gene-doping
SOURCE:  https://futurehumanwiki.com/raw/gene-doping
==============================================================================

---
title: "Gene doping"
slug: "gene-doping"
type: "concept"
status: "experimental"
horizon: "late 2020s"
categories: ["enhancement", "genetics"]
tags: ["sport", "enhancement", "gene therapy", "doping", "detection", "regulation"]
summary: "The non-therapeutic use of gene transfer or gene regulation to improve athletic performance, prohibited in sport since the early 2000s with no publicly confirmed case."
updated: "2026-07-27"
issues: ["The Enhanced Games section, including the 2025 swim claim, carries no citation."]
---

```infobox
{
  "caption": "Prohibited method in sport",
  "rows": [
    { "label": "WADA class", "value": "M3, gene and cell doping" },
    { "label": "Prohibited since", "value": "Early 2000s" },
    { "label": "Candidate targets", "value": "EPO, myostatin, IGF-1, PPARδ" },
    { "label": "Delivery", "value": "Viral vectors", "link": "/wiki/aav-vectors" },
    { "label": "Detection", "value": "Transgene PCR, biological passport" },
    { "label": "Confirmed cases", "value": "None publicly documented" }
  ]
}
```

**Gene doping** is the use of gene transfer, gene editing or gene regulation to improve athletic performance rather than to treat disease. It has been prohibited in competitive sport since the early 2000s, before any technique for doing it existed, and it remains prohibited today. No case has been publicly confirmed in an elite athlete. The gap between a prohibition written in anticipation and an offence nobody has proved is the defining feature of the subject.

## The targets

The candidate genes come from exercise physiology and from muscle-disease research, where the same constructs are studied as therapies.

**Erythropoietin.** EPO drives red-cell production and therefore oxygen delivery. Recombinant EPO is the classic endurance drug; a gene-transfer version would in principle produce it endogenously and continuously. That constancy is also the main hazard, since expression from a viral vector cannot easily be turned down.

**Myostatin and its antagonists.** Blocking myostatin increases muscle mass, an effect established in cattle, dogs and mice, and pursued clinically for muscle-wasting disease. Follistatin constructs delivered by adeno-associated virus are the most plausible muscle-directed doping agent, for the simple reason that they have been administered to humans outside clinical trials. The biology and its disappointments are set out in [[myostatin-inhibition]].

**IGF-1.** Local delivery of insulin-like growth factor 1 by viral vector increased muscle mass and preserved function in aging mice, a result from the late 1990s that gave gene doping its earliest popular image.[^barton1998]

**Metabolic switches.** Transgenic mice overexpressing the nuclear receptor PPARδ in muscle showed a shift toward oxidative fibres and greatly increased running endurance, and pharmacological agonists reproduced part of the effect.[^wang2004] Follow-up work on AMPK and PPARδ agonists framed these compounds as exercise mimetics, and the small molecules involved were added to the prohibited list well before any gene-transfer version was attempted.[^narkar2008]

**Angiogenic and hypoxia pathways.** VEGF and stabilised HIF constructs would increase capillary density or mimic altitude adaptation. These remain conjectural in the doping context.

## Why it is prohibited

The World Anti-Doping Agency lists gene and cell doping as method class M3. The wording is deliberately broad, covering the use of nucleic acid polymers or analogues that alter genome sequences or gene expression by any mechanism, and the use of normal or genetically modified cells.[^wada] The 2018 revision of the list extended the class explicitly to gene-editing agents, which brought [[crispr-cas9]] and its derivatives inside the rule, so an [[epigenome-editing|epigenetic]] intervention that raised expression of a native gene without introducing foreign DNA is prohibited even though it would leave no transgene to find.

Prohibition rests on the same rationale set out in [[enhancement-in-sport]]: health risk, the integrity of competition, and the coercion of athletes who would otherwise decline. The rule is unusual in one respect. Nearly every other class on the list names substances that exist and are available; M3 anticipates a method, and the anticipated version of it is closer to [[gene-therapy-for-aging]] in its ambitions than to a steroid. Gene transfer sharpens the first of these, because a viral vector cannot be withdrawn. An athlete who over-responds to an EPO construct has no route back short of therapeutic intervention against their own transgene.

## Detection

Detection strategies fall into three families.

- **Direct transgene detection.** A therapeutic construct is usually built from complementary DNA, which lacks the introns present in the genomic copy of the same gene. A polymerase chain reaction assay spanning an exon-exon junction distinguishes the two, and anti-doping laboratories have developed and validated such assays for the leading candidates. The method works well on blood, and less well if the vector was injected into a single muscle and never entered circulation in quantity.
- **Vector detection.** Residual viral capsid sequences and anti-capsid antibodies persist after administration, which turns [[aav-vectors|pre-existing vector immunity]] into a potential forensic signal.
- **Indirect signatures.** The Athlete Biological Passport tracks haematological and steroidal variables over time and flags deviations from an athlete's own baseline, regardless of cause. Transcriptomic and proteomic signatures of altered gene expression have been proposed as a complementary approach, though establishing the specificity needed for a sanction is difficult.

> [!caution] Absence of evidence
> No publicly confirmed positive test for gene doping in an elite athlete exists as of 2026. Whether that reflects absence of the practice or the difficulty of catching it is unresolved; anti-doping scientists generally argue both contribute.

## Documented and alleged cases

```timeline
[
  { "year": "1998", "title": "Muscle gene transfer in mice", "text": "Viral delivery of IGF-1 to mouse muscle increases mass and preserves function with age, and is immediately discussed as a doping risk." },
  { "year": "2002–2003", "title": "Anti-doping response", "text": "WADA convenes scientific meetings on gene transfer and adds gene doping to the Prohibited List before any means of doing it is available." },
  { "year": "2004", "title": "Endurance transgenics", "text": "Transgenic mice overexpressing PPARδ show large increases in running endurance, producing the 'marathon mouse' framing." },
  { "year": "2006", "title": "Repoxygen", "text": "In the trial of German athletics coach Thomas Springstein, correspondence emerges in which he seeks Repoxygen, an experimental EPO gene-therapy construct. No evidence of administration to an athlete is established." },
  { "year": "2015–2025", "title": "Unregulated self-administration", "text": "Follistatin gene-transfer products are offered outside clinical trials by companies operating in permissive jurisdictions and taken by self-experimenters, demonstrating that access does not depend on approval." }
]
```

The Repoxygen episode remains the closest thing to a documented attempt, and even there the evidence is of intent rather than of use. The more consequential development has been outside sport altogether: as described in [[biohacking]], gene-transfer products have been sold directly to individuals by companies based in jurisdictions with light regulation, which establishes that an athlete determined to try this would not need a laboratory of their own.

## Risks

The safety profile is inherited from [[somatic-gene-therapy]], where it has been characterised in patients who had a disease justifying the risk. Immune responses to the vector and to the transgene product are the dominant hazard; high-dose systemic AAV has caused serious liver injury and deaths in clinical trials for muscle disease. Gene transfer of erythropoietin in macaques produced, in several animals, an immune response that cross-reacted with native erythropoietin and caused severe anaemia, precisely inverting the intended effect.[^gao2004] Insertional mutagenesis is a smaller but real concern with integrating vectors, the genome-level damage catalogued in [[crispr-off-target-effects]] applies to any editing-based version, and uncontrolled expression is a structural problem rather than a dosing error.

None of these risks are acceptable in a healthy adult seeking a competitive margin, which is the standard bioethical objection and also the practical reason clinicians decline to provide it. The comparison that anti-doping physicians draw is with [[exercise-and-aging|training itself]], which produces adaptations in the same pathways with a known safety profile and no vector.

## The Enhanced Games and the open-doping argument

The Enhanced Games, announced in 2023 and backed by private investors, proposes competition without anti-doping rules, on the argument that supervised enhancement is safer than the clandestine kind and that records suppressed by prohibition would otherwise be achievable. In 2025 the organisers publicised a swim faster than a long-standing world-record mark, achieved outside sanctioned conditions; world federations rejected the comparison and several moved to bar participants from their events.

The venture's stated emphasis has been pharmacological, and its position on gene transfer has been less clearly articulated than its position on drugs. It nonetheless supplies the sharpest version of the argument in [[enhancement-arms-race]]: if a parallel competition removes the rule, the collective-action structure that makes prohibition stable in the first place weakens, and the question of who is protected by the ban becomes concrete. Critics reply that medical supervision does not make an unapproved AAV construct safe, and that the athletes bearing the risk are not the people capitalising the event. The venture is also the clearest current test of whether [[human-enhancement]] can be normalised by staging it in public rather than by arguing for it.

## Outlook

Two developments would change the picture. The first is the maturation of delivery: as [[lipid-nanoparticles|non-viral delivery]] improves and transient expression becomes routine, a doping agent that clears the body would be far harder to detect than an integrated or episomal transgene that persists. The second is the spread of expression-modulating tools that add no foreign sequence at all, for which the forensic question is not whether a construct is present but whether an athlete's own regulatory state is natural.

Anti-doping science is preparing for both, mainly through longitudinal profiling that looks for anomalous stability or change in an athlete's own biology rather than for a foreign molecule. Whether a sanction can ever be sustained on that basis, against the evidentiary standard arbitration panels apply, has not been tested.

## See also

- [[enhancement-in-sport]]
- [[myostatin-inhibition]]
- [[somatic-gene-therapy]]
- [[aav-vectors]]
- [[human-enhancement]]
- [[biohacking]]
- [[enhancement-arms-race]]
- [[exercise-and-aging]]

## References

[^barton1998]: `paper` Barton-Davis, E. R. et al. "Viral mediated expression of insulin-like growth factor I blocks the aging-related loss of skeletal muscle function." *Proceedings of the National Academy of Sciences*, 1998.
[^wang2004]: `paper` Wang, Y.-X. et al. "Regulation of muscle fiber type and running endurance by PPARδ." *PLoS Biology*, 2004. {The mice overexpressed the receptor in muscle from conception, which is not evidence that an intervention in an adult would do the same.}
[^narkar2008]: `paper` Narkar, V. A. et al. "AMPK and PPARδ agonists are exercise mimetics." *Cell*, 2008. {The compounds were given to mice; GW1516 was never approved for any human use, and anti-doping authorities have since warned athletes about its toxicity in animal studies.}
[^wada]: `regulator` World Anti-Doping Agency. *The Prohibited List*, section M3: Gene and Cell Doping. Published annually. {The list is reissued each January, so the class wording cited here is the current text rather than the original from the early 2000s.}
[^gao2004]: `paper` Gao, G. et al. "Erythropoietin gene therapy leads to autoimmune anemia in macaques." *Blood*, 2004.

==============================================================================
ARTICLE: gene-drive
TITLE:   Gene drives
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/gene-drive
SOURCE:  https://futurehumanwiki.com/raw/gene-drive
==============================================================================

---
title: "Gene drives"
slug: "gene-drive"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 5
categories: ["genetics", "society"]
tags: ["crispr", "malaria", "ecology", "biosafety", "governance", "mosquitoes"]
summary: "Genetic elements engineered to bias their own inheritance above the Mendelian half, allowing an introduced trait to spread through a wild population."
updated: "2026-07-27"
humanEvidence: "No drive-carrying organism has been released anywhere as of 2026, so the human outcome claimed for the technology, reduced malaria transmission, is untested; every suppression result comes from caged mosquito populations."
access: "Nothing to obtain: drives exist only in contained laboratory and cage colonies, and no regulator anywhere has authorised a field release."
reversibility: "irreversible"
issues: ["Governance section cites the 2016 National Academies report and the CBD decisions without footnotes."]
---

```infobox
{
  "caption": "Population-scale genetic technology",
  "rows": [
    { "label": "Type", "value": "Super-Mendelian inheritance element" },
    { "label": "Proposed", "value": "2003 (Austin Burt)" },
    { "label": "CRISPR version", "value": "2014–2015" },
    { "label": "Leading target", "value": "Malaria mosquitoes" },
    { "label": "Field release", "value": "None as of 2026" },
    { "label": "Governance forum", "value": "Convention on Biological Diversity" },
    { "label": "Readiness", "value": "TRL 5" }
  ]
}
```

**Gene drives** are genetic elements that copy themselves into the matching chromosome in the germline, so that a heterozygous parent passes the element to far more than half of its offspring. An engineered drive can therefore spread through a wild population from a small release, in principle carrying a trait to fixation against natural selection. The idea long predates [[crispr-cas9]], but the ability to program a nuclease with an RNA guide is what made building one straightforward, and what turned a theoretical proposal into a governance problem.

## How it works

A standard homing drive consists of a nuclease, a guide RNA that specifies a site in the host genome, and the whole cassette inserted at that same site. In a heterozygote, the drive-bearing chromosome expresses the nuclease, which cuts the corresponding site on the intact homologous chromosome. If the cell repairs that break by homologous recombination, it uses the drive-bearing chromosome as template and copies the entire cassette across. The organism, formerly heterozygous, is now homozygous, and every gamete carries the drive.

The two useful designs differ in what they carry. **Population modification** drives spread a cargo that makes the organism harmless — in mosquitoes, an effector that blocks development of the malaria parasite — while leaving the population intact. **Population suppression** drives target a gene required for female fertility or sex determination; the drive spreads while it is rare and heterozygous, and the population collapses as homozygotes accumulate. Suppression is more powerful and less reversible.

> [!key] Why anyone is attempting this
> Malaria kills roughly six hundred thousand people a year, most of them African children under five, and insecticide resistance has eroded the effectiveness of bed nets and spraying.[^who2024] A drive is one of very few proposals that would not require sustained delivery of anything to the households at risk.

## Development history

The concept was worked out in theory a decade before anyone could build one, and the decisive change was not an insight about population genetics but the arrival of a nuclease that could be aimed at an arbitrary sequence.

```timeline
[
  { "year": "2003", "title": "The proposal", "text": "Austin Burt argues that site-specific selfish genetic elements could be engineered to spread deliberately through wild populations, and sets out the resistance problem that would follow." },
  { "year": "2014", "title": "CRISPR drives described", "text": "Esvelt, Church and colleagues describe how RNA-guided nucleases would make drives easy to build, and publish the proposal alongside a call for containment standards before anyone builds one." },
  { "year": "2015", "title": "First working drives", "text": "Drives are demonstrated in fruit flies and then in Anopheles stephensi, the latter carrying anti-parasite effector genes." },
  { "year": "2016", "title": "Suppression drives in Anopheles gambiae", "text": "Drives targeting female fertility genes spread efficiently in cages but select for resistant alleles that halt them." },
  { "year": "2018", "title": "Resistance overcome in cages", "text": "A drive targeting an ultraconserved region of the doublesex gene collapses caged populations within a small number of generations without generating functional resistance." },
  { "year": "2021", "title": "Large-cage trials", "text": "The same drive suppresses larger, more ecologically structured indoor cage populations, the closest existing approximation to field conditions." }
]
```

The technical arc has been consistent: drives work in insects, resistance is the main obstacle, and careful target choice can defeat resistance in a cage. The unresolved questions are all about what happens outside one.

One feature of the field's history is unusual. The 2014 paper by Kevin Esvelt, [[george-church]] and colleagues that described CRISPR drives also argued for containment standards and public discussion before any drive was built, and was published alongside a policy piece to that effect. The technology was therefore proposed and criticised by the same people at the same time, an inversion of the sequence at the [[asilomar-conference]] four decades earlier, where the moratorium followed the capability.

## Resistance

The drive's own mechanism generates its principal failure mode. When the nuclease cuts the homologous chromosome, homologous recombination copies the drive across, but end joining instead produces a small insertion or deletion at the cut site. That altered sequence is no longer recognised by the guide RNA. If the resulting allele still works, it is a resistance allele that is immune to the drive and favoured by selection whenever the drive imposes a fitness cost — and suppression drives impose a very large one.

Two countermeasures have worked in the laboratory. One is to target a sequence so functionally constrained that almost any indel destroys the gene, so resistance alleles are themselves lethal or sterile. The other is multiplexing several guide RNAs against the same gene, so that resistance requires simultaneous escape at multiple sites. Neither has been tested against the standing genetic variation of a wild population, which is far larger than any cage contains.

## Current state

No gene drive organism has been released anywhere as of 2026. The most advanced programme, Target Malaria, a not-for-profit research consortium working on *Anopheles gambiae* in West and East Africa, has followed a deliberately staged path: first a release of genetically modified sterile males that could not persist, then male-biasing strains, with a drive release contemplated only after regulatory and community processes conclude. The releases conducted so far involved no drive.

Two adjacent technologies are already deployed and are frequently confused with drives. Self-limiting modified male *Aedes aegypti* have been released at scale in Brazil and, in a smaller pilot, in the Florida Keys; they reduce local populations but disappear within generations and must be re-released. Mosquitoes carrying *Wolbachia* bacteria, which distort inheritance without any engineered nuclease, have been established in several cities and were associated with a large reduction in dengue incidence in a randomised trial in Indonesia.[^utarini2021] Both illustrate that population-level intervention is possible without a self-propagating edit.

In mammals, drives remain far weaker. A drive built in mice copied itself only in the female germline and at rates too low to spread, so proposals to control invasive rodents on islands have no working molecular basis yet.[^grunwald2019] The gap between what is possible in an insect germline and in a mammalian one is a recurring theme in applied genetics, and it constrains conservation proposals as much as it constrains the edited proxy animals pursued under [[de-extinction]] by firms such as [[colossal-biosciences]].

## Confinement and reversibility

Because a standard homing drive is designed to spread indefinitely, most current research effort goes into designs that do not.

- **Split drives** separate the nuclease from the guide RNA, so the element spreads only where the missing component has been supplied — usually only in a laboratory strain.
- **Daisy-chain drives** arrange elements in a series in which each drives the next; the chain is consumed from one end, and the drive exhausts itself after a predictable number of generations.[^noble2019]
- **Threshold-dependent designs** spread only if released above a frequency high enough that they cannot cross into a neighbouring population by ordinary migration.
- **Reversal drives** are intended to overwrite a released drive with a neutral sequence. They restore function but do not restore the original sequence, and no field demonstration exists.

The logic parallels the synthetic auxotrophy used for biocontainment in [[recoded-organisms]]: rather than trusting physical barriers, the organism is engineered so that escape is self-defeating. The honest position is that reversibility here is a design intention rather than a demonstrated capability. Nothing engineered has yet been recalled from a wild population.

## Risks and governance

The ecological concerns are of three kinds: effects on species that eat or are pollinated by the target; spread of the construct into related species through hybridisation, which is a live issue in the *Anopheles gambiae* species complex; and the possibility that a suppression drive succeeds more completely or more widely than intended. The [[crispr-off-target-effects|off-target problem]] takes an unusual form here, since a drive amplifies itself over generations and any unintended edit it carries travels with it.

Governance has settled into an uneasy pattern. A 2016 United States National Academies report endorsed continued research with staged testing rather than prohibition. Proposals for a global moratorium at the Convention on Biological Diversity were rejected in 2018 in favour of decisions calling for case-by-case risk assessment and for the free, prior and informed consent of indigenous peoples and local communities potentially affected. The World Health Organization has published a testing framework for genetically modified mosquitoes. None of these instruments is binding in the way that domestic biosafety law is, and a drive released anywhere will not respect the jurisdiction that authorised it — a structural problem shared with the concerns raised in [[mirror-life]] and, more distantly, in [[dual-use-research]].

The consent question has no clean answer. The population that must agree to a release is not the population that will be affected, since mosquitoes cross borders and generations — a problem formally similar to the objection that no future person can consent to [[germline-editing]], but at ecosystem scale. Frameworks developed for [[governance-of-genome-editing]] in medicine assume an identifiable patient and do not transfer, and arguments from the [[precautionary-principle]] cut both ways when the status quo is a disease that kills hundreds of thousands of people annually. Drives are also a standard case in the [[existential-risk]] literature, less because a mosquito drive threatens humanity than because it is the first cheap technology whose intended behaviour is unbounded spread. Proposals for staged, reversible deployment are the clearest applied instance of [[differential-technological-development]] in biology.

## Outlook

The technical questions that remain are ecological rather than molecular: how a drive behaves across a structured, seasonally fluctuating wild population with far more genetic variation than any cage, and whether modelling can be trusted before a release that cannot be undone. Field trials of any drive-carrying organism would be the first deliberate release of a self-propagating engineered element, and the decision is not principally a scientific one. Whether the first such release happens this decade depends less on the drive's performance in cages than on whether an African regulator, a national government and an affected community all decide, in that order, that the alternative is worse.

## See also

- [[crispr-cas9]]
- [[crispr-off-target-effects]]
- [[governance-of-genome-editing]]
- [[mirror-life]]
- [[dual-use-research]]
- [[precautionary-principle]]
- [[de-extinction]]
- [[recoded-organisms]]

## References

[^who2024]: `report` World Health Organization. *World Malaria Report*. Annual series; figures cited are from the 2024 edition. {WHO malaria deaths are modelled estimates rather than counted deaths, and carry wide uncertainty intervals.}
[^utarini2021]: `paper` Utarini, A. et al. "Efficacy of Wolbachia-infected mosquito deployments for the control of dengue." *New England Journal of Medicine*, 2021. {A cluster-randomised trial in Yogyakarta, Indonesia. Wolbachia is not a gene drive and involves no engineered nuclease.}
[^grunwald2019]: `paper` Grunwald, H. A. et al. "Super-Mendelian inheritance mediated by CRISPR–Cas9 in the female mouse germline." *Nature*, 2019.
[^noble2019]: `paper` Noble, C. et al. "Daisy-chain gene drives for the alteration of local populations." *Proceedings of the National Academy of Sciences*, 2019. {The design is worked out in population-genetic models; the paper does not report a daisy-chain drive built in an organism.}

==============================================================================
ARTICLE: gene-therapy-for-aging
TITLE:   Gene therapy for aging
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/gene-therapy-for-aging
SOURCE:  https://futurehumanwiki.com/raw/gene-therapy-for-aging
==============================================================================

---
title: "Gene therapy for aging"
slug: "gene-therapy-for-aging"
type: "intervention"
status: "experimental"
horizon: "2040s"
trl: 3
categories: ["longevity", "genetics"]
tags: ["gene therapy", "aav", "telomerase", "klotho", "follistatin", "aging", "self-experimentation"]
summary: "The use of delivered genes to modify aging biology systemically, demonstrated in mice with telomerase and other constructs and untested in any controlled human trial."
updated: "2026-07-28"
humanEvidence: "No controlled trial has tested any of these constructs against aging in people; the human record is unregulated self-administration reported by press release, plus early-phase follistatin trials in muscular dystrophy."
access: "Available through no approved route; pay-to-participate ventures in permissive jurisdictions have offered telomerase and follistatin constructs, one at a reported price of about a million dollars per participant."
reversibility: "context"
issues: ["The BioViva, Libella, and Honduras episodes are described without citations.", "The klotho monkey result and the OSK mouse lifespan claim are stated without sources."]
---

```infobox
{
  "caption": "Experimental therapeutic strategy",
  "rows": [
    { "label": "Delivery", "value": "AAV, plasmid, lipid nanoparticle" },
    { "label": "Common targets", "value": "TERT, klotho, follistatin, OSK" },
    { "label": "Best animal result", "value": "AAV-telomerase lifespan gain in mice, 2012" },
    { "label": "Controlled human trials", "value": "None for aging" },
    { "label": "Unregulated human use", "value": "Documented since 2015" },
    { "label": "Main obstacle", "value": "Delivery, permanence, no approved indication" },
    { "label": "Readiness", "value": "TRL 3" }
  ]
}
```

**Gene therapy for aging** is the delivery of genetic material to a large fraction of an adult's
cells with the aim of changing the trajectory of aging itself rather than treating a single inherited
disease. It differs from ordinary [[somatic-gene-therapy]] in ambition rather than in method: the
same viral vectors and nucleic-acid payloads, aimed at systemic targets like telomere maintenance,
muscle mass, or epigenetic state, and given to people who are not sick by any current definition. As
of 2026 the animal results are real and the human record consists almost entirely of unregulated
self-experimentation.

## The animal evidence

The founding result is telomerase. A 2012 study delivered the mouse TERT gene in an adeno-associated
viral vector to adult mice and reported extended median lifespan, with a larger effect in animals
treated at one year of age than at two, and without an observed increase in cancer
incidence.[^bernardes2012] That last clause is the striking part, because telomerase reactivation is
a near-universal feature of tumours, and the finding has not been reproduced widely enough to settle
the safety question. Its interpretation depends on the broader argument covered under
[[telomeres-and-telomerase]], where telomere attrition is a weaker driver of human aging than of
mouse aging in several respects.

Two other targets recur. [[klotho|Klotho]], a protein whose loss in mice produces a syndrome resembling
accelerated aging and whose overexpression extends mouse lifespan,[^kurosu2005] has been pursued both
as gene therapy and as injected protein; a study in aged rhesus monkeys reported cognitive improvement
after peripheral klotho administration, in small numbers and without long-term follow-up. Follistatin
blocks myostatin signalling and reliably increases muscle mass when delivered by AAV to mice and
monkeys; early-phase trials in muscular dystrophy have reported modest and inconsistent functional
change,[^mendell2015] and the relationship between added muscle mass and added strength is not
straightforward — the issue treated in [[myostatin-inhibition]].

The newest line delivers reprogramming factors rather than a single protein. Vectors carrying OSK, the
three-factor subset of the [[yamanaka-factors]] that omits the oncogene MYC, have been reported to
extend remaining lifespan in mice treated in old age, in studies with small cohorts that have not been
independently replicated. This work sits at the intersection of gene therapy and
[[epigenetic-reprogramming]]; the dosing and duration questions are those of
[[partial-reprogramming]].

> [!caution] Mice are not the hard case
> A mouse lives under thirty months, weighs thirty grams, and can receive a vector dose per kilogram
> that is impossible to manufacture for an adult human. Every result in this section is a result about
> a small, short-lived, inbred animal treated in a specific-pathogen-free facility.

## Why aging is a difficult target

Gene therapy works best where a single defective gene in a reachable tissue causes a defined disease.
Aging has none of those properties.

*No single target.* The processes catalogued as [[hallmarks-of-aging]] are numerous and partly
independent. A construct that addresses telomere length leaves the others untouched,
which is why proposals in this space tend toward multi-gene combinations, each adding immunological
and manufacturing complexity.

*Delivery.* [[aav-vectors|AAV]] reaches the liver efficiently and most other tissues poorly, carries
a payload limit of about five kilobases, and provokes an immune response that prevents redosing with
the same capsid. A substantial share of adults already carry neutralizing antibodies to common
serotypes and are ineligible. Engineered capsids and non-viral routes such as
[[lipid-nanoparticles]] address parts of this, but no delivery system currently reaches a majority of
cells in a majority of tissues in an adult.

*Permanence.* An AAV genome persists episomally in non-dividing cells for years and cannot be
withdrawn. For a transgene that promotes growth, blocks a tumour suppressor, or reactivates
telomerase, the absence of an off switch is a serious safety problem. Regulatable systems and
transient mRNA delivery are the responses, at the cost of requiring repeated administration through a
route that immunity tends to close.

*Risk-benefit.* High-dose systemic AAV has caused severe hepatotoxicity and deaths in trials for
serious childhood diseases. That risk is defensible when the alternative is a fatal disorder. Applied
to a healthy sixty-year-old seeking a probabilistic benefit decades hence, the same risk is not, and
no regulator has an approval pathway for the indication — the same structural obstacle described under
[[geroscience-hypothesis]] and in [[metformin|the TAME proposal]].

## Unregulated human use

Because no legitimate trial exists, the human record has been written by people outside the system.

In 2015 the chief executive of the company BioViva, Elizabeth Parrish, received AAV constructs
carrying telomerase and a myostatin inhibitor outside the United States and reported subsequent
lengthening of leukocyte telomeres measured by a commercial assay. The result was never peer
reviewed, the assay is noisy, and one person is not a study.

In 2019 Libella Gene Therapeutics announced a pay-to-participate telomerase gene therapy offered in
Colombia at around a million dollars per participant, a structure that inverts the ethics of clinical
research by charging subjects for an unproven intervention. No results have been published.

Since 2023 a company has offered plasmid-based follistatin gene therapy in a special jurisdiction in
Honduras, with participants including a prominent longevity entrepreneur. Plasmid delivery is less
persistent than AAV, which reduces some risks, and the venture has produced self-collected data rather
than controlled evidence.

These ventures share a pattern: jurisdiction shopping, absent control groups, surrogate endpoints
chosen after the fact, and outcomes announced by press release. They generate no usable knowledge and
raise the regulatory temperature for the researchers doing the work properly.

## What a legitimate path would look like

The most plausible route runs through narrow indications where a systemic genetic intervention is
already defensible. Gene therapies targeting the APOE4 variant in carriers at high risk of
Alzheimer's disease have entered early-phase trials, and one-time editing of PCSK9 to lower lifelong
cholesterol, using [[base-editing]] delivered in lipid nanoparticles and first shown to work durably
in non-human primates,[^musunuru2021] is the closest existing test of a durable genetic intervention
given for prevention rather than rescue. Neither is marketed as an aging therapy, and both would
establish precisely the precedents an aging therapy needs: acceptable risk in relatively healthy adults,
durable effect from a single administration, and a regulatory framework for prevention.

Progress also depends on measurement. Without an accepted [[aging-biomarkers|biomarker of aging]] a
trial cannot read out in less than a decade, and neither an [[epigenetic-clock]] shift nor an
improvement in a [[biological-age|composite age score]] is currently sufficient for approval.

## Outlook

The near-term question is not whether aging can be treated genetically but whether any of the
candidate constructs are worth the delivery risk. Telomerase carries an unquantified cancer liability;
follistatin adds muscle of uncertain functional value; reprogramming factors carry dedifferentiation
and teratoma risk that has been demonstrated in mice. Combination approaches multiply these problems
rather than averaging them.

If the field does mature, its main constraint may be economic rather than biological. One-time gene
therapies for rare disease already carry prices in the millions, and a therapy intended for everyone
above sixty would face a manufacturing and distribution problem of a different order — the concern
argued out under [[access-and-inequality]].

## See also

- [[somatic-gene-therapy]]
- [[aav-vectors]]
- [[epigenetic-reprogramming]]
- [[telomeres-and-telomerase]]
- [[myostatin-inhibition]]
- [[hallmarks-of-aging]]
- [[senolytics]]
- [[geroscience-hypothesis]]

## References

[^bernardes2012]: `paper` Bernardes de Jesus, B. et al. "Telomerase gene therapy in adult and old mice delays aging and increases longevity without increasing cancer." *EMBO Molecular Medicine*, 2012. {Laboratory mice have far longer telomeres than humans and express telomerase in more tissues, so the model may not transfer.}
[^kurosu2005]: `paper` Kurosu, H. et al. "Suppression of aging in mice by the hormone Klotho." *Science*, 2005. {A transgenic overexpression model rather than a delivered therapy: the mice carried the extra klotho from conception.}
[^mendell2015]: `paper` Mendell, J.R. et al. "A phase 1/2a follistatin gene therapy trial for Becker muscular dystrophy." *Molecular Therapy*, 2015. {An open-label early-phase trial in a small number of men with Becker muscular dystrophy, with no placebo group.}
[^musunuru2021]: `paper` Musunuru, K. et al. "In vivo CRISPR base editing of PCSK9 durably lowers cholesterol in primates." *Nature*, 2021.

==============================================================================
ARTICLE: generation-ship-biology
TITLE:   Generation ship biology
PORTAL:  Space & Extreme Environments
URL:     https://futurehumanwiki.com/wiki/generation-ship-biology
SOURCE:  https://futurehumanwiki.com/raw/generation-ship-biology
==============================================================================

---
title: "Generation ship biology"
slug: "generation-ship-biology"
type: "concept"
status: "speculative"
horizon: "indefinite"
categories: ["space", "reproduction"]
tags: ["interstellar", "population genetics", "reproduction", "closed ecology", "consent", "spaceflight"]
summary: "The biological requirements of a crewed voyage lasting generations: population genetics, reproduction and development off Earth, and consent for people born in transit."
updated: "2026-07-27"
humanEvidence: "No human has been conceived, gestated, or born beyond Earth, and no agency has studied conception in flight; the mammalian record is mouse embryos cultured in orbit and rodents flown during part of gestation."
issues: ["The Biosphere 2 oxygen decline and the medaka and rat spaceflight results carry no citations."]
---

```infobox
{
  "caption": "Problem in population biology and mission design",
  "rows": [
    { "label": "Scenario", "value": "Multi-generation crewed voyage" },
    { "label": "Core constraints", "value": "Genetics, gestation, ecology" },
    { "label": "Modelled minimum crew", "value": "~100 to several hundred" },
    { "label": "Mammals born in space", "value": "None" },
    { "label": "Longest closed human test", "value": "Two years (Biosphere 2)" },
    { "label": "Status", "value": "Analysis only" }
  ]
}
```

**Generation ship biology** is the study of what a crewed interstellar voyage would require of human biology when the voyage outlasts the people who begin it. The engineering problems of such a ship — propulsion, shielding, structure — are severe but conventionally analysable. The biological problems are different in kind: they involve a human population small enough to lose genetic diversity, a reproductive process never tested off Earth, an ecosystem with no external buffer, and several generations of people who are born into a mission they did not choose.

```keyfacts
[
  { "value": "~100", "label": "Smallest crew found viable in modelling", "note": "under strict breeding rules; several hundred with realistic accident and infertility rates" },
  { "value": "0", "label": "Mammals conceived and born beyond Earth", "note": "no mammalian pregnancy has been carried to term in space" },
  { "value": "50/500", "label": "Classical conservation genetics rule", "note": "later argued to require revision upward" }
]
```

## Minimum viable population

The starting question is how many people are needed to avoid genetic collapse. Conservation biology supplies the traditional heuristic: an effective population of about 50 avoids severe short-term inbreeding depression, while about 500 is needed to retain enough variation for long-term adaptive potential. Later analyses argued both figures should be raised substantially, and all of them refer to *effective* population size, which in a real group with unequal reproduction is considerably smaller than the headcount.[^frankham2014]

Explicit modelling of interstellar crews has produced numbers in the same range. Monte Carlo simulations of a multi-thousand-year voyage found that a founding group on the order of a hundred could remain genetically viable under strict pairing rules that minimise relatedness, and that several hundred were needed once realistic rates of infertility, accident, and free mate choice were allowed.[^marin2018] Anthropological estimates for shorter voyages of a few centuries have landed in the same territory.

Small founding populations produce predictable genetics. Heterozygosity is lost through drift at a rate inversely proportional to effective population size. Rare recessive alleles carried by a founder rise to high frequency, which is why real founder populations on Earth — Finnish, Ashkenazi, Old Order Amish — each carry a characteristic set of otherwise rare disorders. A ship's population would acquire its own within a few generations, and it would be a different set for each ship.

Mitochondrial lineages compound the problem, since they descend only through founding women and cannot recombine. A crew founded with few maternal lineages would carry whatever mitochondrial variants those women had, and the techniques of [[mitochondrial-replacement-therapy]] would be one of the few available corrections.

## Why frozen gametes change the arithmetic

The population-size problem is far less severe than the headline numbers suggest, because genetic diversity need not be carried in living bodies.

A gamete or embryo bank drawn from thousands of donors decouples effective population size from crew size entirely. A small crew with a well-stocked cryobank commands the genetic diversity of a far larger population, and the bank can be sampled deliberately to counteract drift rather than left to chance. The relevant technology is established: oocyte and embryo vitrification is routine clinical practice, discussed under [[reproductive-longevity]], and [[embryo-selection]] would allow screening against the recessive burden that a small population accumulates.

Two caveats matter. Cryogenic storage over centuries is chemically plausible, but radiation dose accumulates in stored samples and the empirical work is thin. Freeze-dried mouse sperm stored aboard the International Space Station for several years produced healthy offspring after return, which is encouraging for years and silent about centuries.[^wakayama2021] And [[in-vitro-gametogenesis]], if it matured in humans, would let a small crew generate large numbers of gametes from somatic cells, though it adds no diversity beyond what the founders and the bank already contain.

## Reproduction beyond Earth

No mammal has been conceived, gestated, and born outside Earth. This is the single largest gap in the subject.

What exists is fragmentary. Medaka fish mated in orbit during a Shuttle mission and produced viable offspring, the first vertebrates to complete a reproductive cycle in space. Mouse embryos cultured aboard the station developed to the blastocyst stage at rates somewhat below ground controls.[^wakayama2023] Pregnant rats flown late in gestation delivered normally after return, and their pups showed differences in vestibular development consistent with the otolith organs having formed under altered gravitational input. Attempts to breed rodents in orbit have not produced confirmed pregnancies.

Human data are absent by design: no space agency has flown a pregnant crew member, and none has studied conception in flight. The relevant physiology from [[space-medicine]] gives reasons for concern rather than reassurance — fluid shifts, immune dysregulation, and altered bone metabolism are all directly relevant to pregnancy, and obstetric emergency care is not available on any spacecraft.

Radiation adds a separate hazard. Oocytes are among the most radiosensitive cells in the body and are not replaced; the ovarian reserve a woman launches with is the reserve she keeps. Chronic exposure at the levels described in [[radiation-hardening-humans]] raises questions about both fertility and heritable mutation that cannot be settled with current data.

> [!caution] The unmeasured parameter
> Every projection about generation ships assumes that human gestation and infant development proceed acceptably in whatever gravity the vessel provides. There is no mammalian evidence for this at any gravity level between zero and one. A ship providing artificial gravity by rotation sidesteps the question; one that does not is betting a mission on an untested developmental biology.

## Development, gravity, and radiation

Vertebrate development uses gravity as a signal at several points. Otolith formation, the calibration of vestibular reflexes, and the postnatal loading history that shapes bone architecture all depend on it. Fish and amphibian work indicates that vestibular development proceeds differently under altered gravity; the mammalian case is largely unexamined and the human case entirely so.

A child raised in low gravity would develop a skeleton adapted to that load, which is an appropriate response to the environment and a permanent barrier to visiting Earth. Over generations, selection would act on the shipboard environment rather than the terrestrial one — the point at which generation-ship biology merges with [[pantropy]], and at which the ship's population becomes, in a limited sense, a distinct population rather than a travelling sample of an existing one.

## Ecosystem and microbiome over centuries

The crew is one compartment of a closed ecology, and the arguments in [[closed-loop-life-support]] apply with an added time dimension. Biosphere 2 lost atmospheric oxygen over sixteen months through a soil-and-concrete interaction nobody anticipated; a voyage lasting centuries gives such couplings far more time to run, with no resupply and nowhere to vent.

The human microbiome is part of the system. Isolated populations lose microbial diversity, and a closed habitat's microbial community drifts under selection pressures — antimicrobial use, surface materials, a restricted diet — that have no terrestrial analogue. Pathogen evolution in a small, immunologically homogeneous host population is a recognised risk, and immune naivety compounds it: a population that has not met a pathogen for ten generations retains no acquired immunity to it.

## Consent and the multi-generation problem

The founders volunteer. Everyone afterwards is born into a sealed vehicle with a fixed destination, a fixed diet, a small mating pool, and no exit. The ethical literature treats this as a genuinely hard case rather than a rhetorical one.

Several objections are usually run together. The consent objection is weak in its simple form, since no one consents to the circumstances of their birth. The stronger objection concerns the *narrowness* of the resulting life: the second generation's options are constrained not by nature but by a deliberate prior decision, and the constraint is total. And there is a question about reproductive autonomy, since a ship with a modelled population target cannot leave reproduction entirely to individual choice without risking the demographic collapse the model was built to prevent. Selection of who reproduces, or screening of which embryos are used, converts the voyage into a permanent regime of the kind [[procreative-beneficence]] debates in a much milder setting, with the coercion problem raised in [[genetic-discrimination]] applied to an entire population.

Fiction has probed this more thoroughly than the academic literature. Kim Stanley Robinson's *Aurora* argues that generation ships fail for compounding biological and social reasons — ecological drift, disease, skill loss, and the accumulated resentment of people who did not choose the mission — and the argument has been answered in earnest by advocates.

## Alternatives to carrying people

Several proposals avoid the population problem by not sending a population.

Embryo space colonisation would send cryopreserved embryos rather than adults, gestating them on arrival. This requires [[artificial-womb|full ectogenesis]], which is far beyond the partial systems discussed under [[ectogenesis]], plus machine rearing of infants, plus cultural transmission from no humans at all. It trades an intractable social problem for two intractable technical ones.

Induced [[human-hibernation|torpor]] would reduce consumable demand on shorter voyages and does nothing for a journey longer than a human lifespan. Sending information rather than bodies, to be reconstituted on arrival, belongs to the [[whole-brain-emulation]] literature and inherits every objection raised there.

## Outlook

Nothing here is testable at present, and the honest assessment is that generation-ship biology is analysis rather than research. The parameters that would matter most — whether mammals reproduce and develop normally at partial gravity, whether a closed ecology is stable over decades, whether cryopreserved gametes survive centuries of cosmic radiation — could all be investigated in low Earth orbit and largely have not been.

That gap is itself informative. A civilisation seriously planning to leave the solar system would be running rodent breeding colonies in a rotating orbital facility now. That none exists suggests the question is being treated as a thought experiment, and any argument about the [[future-of-humanity]] beyond this system that assumes otherwise is assuming data that does not exist. Whether interstellar settlement is a live option or a permanently receding one is, at this point, an empirical question nobody is measuring.

## See also

- [[closed-loop-life-support]]
- [[pantropy]]
- [[space-medicine]]
- [[artificial-womb]]
- [[in-vitro-gametogenesis]]
- [[radiation-hardening-humans]]
- [[human-hibernation]]
- [[existential-risk]]

## References

[^frankham2014]: `paper` Frankham, R., Bradshaw, C.J.A., Brook, B.W. "Genetics in conservation management: Revised recommendations for the 50/500 rules, Red List criteria and population viability analyses." *Biological Conservation*, 2014.
[^marin2018]: `paper` Marin, F. and Beluffi, C. "Computing the Minimal Crew for a Multi-Generational Space Journey Towards Proxima Centauri b." *Journal of the British Interplanetary Society*, 2018. {A Monte Carlo model whose crew numbers follow from assumed rates of infertility, accident, and mate choice.}
[^wakayama2021]: `paper` Wakayama, S. et al. "Evaluating the long-term effect of space radiation on the reproductive normality of mammalian sperm preserved on the International Space Station." *Science Advances*, 2021. {The samples sat inside a station within the magnetosphere for years, a far lower dose than a voyage outside it over centuries.}
[^wakayama2023]: `paper` Wakayama, S. et al. "Effect of microgravity on mammalian embryo development evaluated at the International Space Station." *iScience*, 2023. {Embryos were cultured only to the blastocyst stage in orbit; none was transferred or carried to term.}

==============================================================================
ARTICLE: recoded-organisms
TITLE:   Genetic code expansion and recoding
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/recoded-organisms
SOURCE:  https://futurehumanwiki.com/raw/recoded-organisms
==============================================================================

---
title: "Genetic code expansion and recoding"
slug: "recoded-organisms"
type: "technology"
status: "experimental"
horizon: "2040s"
trl: 4
categories: ["genetics"]
tags: ["synthetic biology", "genetic code", "protein engineering", "biocontainment", "virus resistance"]
summary: "Altering how a cell reads DNA so that reassigned codons encode amino acids not found in nature, producing organisms with chemistry and immunity no natural cell has."
updated: "2026-07-27"
humanEvidence: "No recoded mammal or person exists; genome recoding has been demonstrated only in bacteria, and the nearest contact with patients is investigational antibody conjugates built using non-standard amino acids."
access: "Not a clinical or consumer technology: recoded strains exist in a handful of academic laboratories, and the only products near patients are investigational antibody conjugates."
reversibility: "irreversible"
issues: ["The clinical progress of site-specific antibody conjugates is stated without a source"]
---

```infobox
{
  "caption": "Field of synthetic biology",
  "rows": [
    { "label": "Type", "value": "Translation-system engineering" },
    { "label": "First expanded code", "value": "2001, Escherichia coli" },
    { "label": "First recoded genome", "value": "2013, all amber codons replaced" },
    { "label": "Codons freed to date", "value": "Three, in bacteria" },
    { "label": "Notable consequence", "value": "Virus resistance" },
    { "label": "Demonstrated in mammals", "value": "Code expansion only" },
    { "label": "Readiness", "value": "TRL 4" }
  ]
}
```

**Genetic code expansion and recoding** are two halves of one project: changing the rules by which a cell translates DNA into protein. Expansion adds a new amino acid to an organism's repertoire by assigning it a codon; recoding clears that codon first, by replacing every natural instance of it across the genome and deleting the machinery that reads it. The result is a cell that builds proteins from a chemistry unavailable to any natural organism — and, as a side effect that has proved at least as interesting, one that viruses cannot read.

## How it works

Changing the genetic code means changing two things that normally match: the meaning a cell assigns to a codon, and the sequences written in the old meaning. Expansion handles the first and is comparatively easy. Recoding handles the second and is the reason the field depends on genome-scale DNA synthesis.

### Orthogonal translation

Protein synthesis assigns amino acids to codons through aminoacyl-tRNA synthetases, each of which charges a particular transfer RNA with a particular amino acid. To add a twenty-first amino acid, an engineer supplies a synthetase–tRNA pair that ignores, and is ignored by, everything already in the cell — an **orthogonal pair**. The synthetase is then evolved in the laboratory to accept the desired non-standard amino acid, and the tRNA is given an anticodon matching whichever codon has been set aside.

The first such system, reported in 2001, used a pair borrowed from an archaeon and assigned a modified tyrosine to the amber stop codon.[^wang2001] The pair most used since derives from the pyrrolysine system of methanogenic archaea, which is orthogonal in bacterial, yeast and mammalian cells alike and tolerates an unusually wide range of substrates.

### Freeing a codon

Assigning a stop codon to a new amino acid works, but imperfectly: the cell's release factor still competes for the same codon, so incorporation is inefficient and native proteins are occasionally read through. The clean solution is to remove every instance of the codon from the genome and then delete the release factor. That is genome surgery at thousands of positions, well beyond the reach of one-site-at-a-time tools such as [[crispr-cas9]] or [[prime-editing]], and it is why the field is inseparable from [[synthetic-genomes]].

The first genomically recoded organism, reported in 2013 by a group including [[george-church]], replaced all 321 amber stop codons in *Escherichia coli* with an alternative stop and deleted the corresponding release factor.[^lajoie2013] A later effort rebuilt the entire four-megabase genome from synthetic fragments to remove two serine codons as well as the amber codon, then deleted the transfer RNAs that read them, freeing three codons at once.[^fredens2019][^robertson2021]

> [!key] Why viruses cannot read a recoded genome
> A bacteriophage genome is written in the standard code. Injected into a cell that has deleted the tRNAs or release factor for particular codons, the viral messenger RNAs stall or are mistranslated, and no functional virions are produced. Resistance is therefore structural rather than a defence the virus can evolve around, since escaping it would require the phage to rewrite its own coding sequence throughout.

```timeline
[
  { "year": "2001", "title": "First expanded code", "text": "Schultz's group incorporates a non-standard amino acid into a protein in living E. coli using an orthogonal synthetase–tRNA pair assigned to the amber codon." },
  { "year": "2010", "title": "Orthogonal ribosomes", "text": "Engineered ribosomes that read four-base codons allow multiple distinct non-standard amino acids to be encoded in one protein." },
  { "year": "2013", "title": "First genomically recoded organism", "text": "All 321 amber codons are removed from the E. coli genome and the release factor deleted, giving efficient incorporation and partial phage resistance." },
  { "year": "2015", "title": "Synthetic auxotrophy", "text": "Essential proteins are redesigned to require a synthetic amino acid, making the strain dependent on a compound available only in the laboratory." },
  { "year": "2019–2021", "title": "Three codons freed", "text": "A fully synthetic 4-megabase E. coli genome removes two serine codons and the amber codon; deleting the corresponding tRNAs yields broad resistance to bacteriophages." },
  { "year": "2023", "title": "Code swapping", "text": "Reassigning freed codons to new meanings is shown to block viral propagation and to prevent engineered genes from functioning if transferred to natural organisms." }
]
```

## Current state

Recoding at genome scale exists only in bacteria. Code expansion — adding one or two non-standard amino acids without clearing their codons — works routinely in yeast, in cultured mammalian cells, and in whole animals including mice, where it is used to make proteins that can be switched on with light or cross-linked to their binding partners on command. No recoded mammalian genome exists, and none is close: the recoding of a bacterial genome required total synthesis of four megabases, and a human genome is close to a thousand times that size.

The commercial application that reached medicine first is protein conjugation. Placing a non-standard amino acid with a chemically distinct handle at a chosen position in an antibody allows a drug payload to be attached at exactly that site and nowhere else, giving a homogeneous product rather than the mixture produced by conventional chemistry. Antibody–drug conjugates built this way have advanced through clinical development, and the approach is now one of the more mature strands of [[targeted-drug-delivery]].

## Limitations

Recoded strains generally grow more slowly than their parents, and each additional freed codon compounds the fitness cost. Incorporation efficiency falls as more non-standard amino acids are encoded in the same protein, and the amino acids themselves must be supplied in the medium or synthesized by an engineered pathway. Orthogonality is never perfect: engineered synthetases retain some activity on natural substrates, and engineered tRNAs are occasionally charged by native enzymes.

The deeper constraint is that recoding requires writing a genome, so the technology inherits every limit of DNA synthesis and assembly. Until mammalian chromosome construction is routine, extending recoding beyond microbes is a proposal rather than a programme.

## Biocontainment and risk

The most developed safety application inverts the usual worry about engineered organisms. Rather than relying on physical containment, essential proteins in a recoded strain are redesigned so that they fold only when a synthetic amino acid is incorporated. The organism then cannot survive outside a laboratory that supplies the compound, and measured escape rates in the founding studies fell below the limit of detection.[^mandell2015] Code swapping adds a second layer: engineered genes written in a non-standard code are mistranslated if they are transferred horizontally into a natural organism, so the construct does not function even if the DNA escapes.[^nyerges2023]

These are the strongest existing demonstrations that engineered biology can be made intrinsically, rather than procedurally, contained — a contrast with the self-spreading elements described in [[gene-drive]], where containment is the unsolved problem. They remain laboratory results at modest scale, and no regulator currently treats synthetic auxotrophy as sufficient grounds for reduced physical containment.

The risks run in the other direction too. A cell immune to all natural viruses is also a cell that natural ecosystems have no mechanism to control, which is one of the concerns that motivated the warnings collected in [[mirror-life]] about organisms placed outside the reach of existing biology. Recoding is discussed in [[dual-use-research]] assessments mainly as a capability multiplier rather than a direct hazard: the same synthesis and assembly infrastructure serves both, and the [[precautionary-principle]] arguments applied to genome writing apply here with the same force.

## Outlook

Two goals define the near-term agenda. The first is bacterial: strains with several freed codons that can polymerise non-biological monomers, turning cells into programmable chemical factories for molecules the ribosome was never built to make. The second is mammalian, and much further off — a virus-resistant human cell line for biomanufacturing, proposed as the flagship of human genome writing, which would remove the contamination risk that shuts down production runs. A related and more tractable version of the same logic is already in clinical use in [[xenotransplantation]], where donor pigs have had endogenous retroviral sequences inactivated wholesale rather than recoded.

Whether recoding ever reaches a human organism rather than a human cell line is a separate question and not an active one. It would require rewriting every cell of a person, which means doing it at the embryo stage, and so falls squarely within the prohibitions on [[germline-editing]] surveyed in [[governance-of-genome-editing]]. The proposal occasionally surfaces in discussions of engineered virus immunity; nobody has proposed a credible route, and the mismatch between a laboratory strain that grows slowly on a defined medium and a functioning mammal is the reason.

## See also

- [[synthetic-genomes]]
- [[crispr-cas9]]
- [[mirror-life]]
- [[gene-drive]]
- [[dual-use-research]]
- [[george-church]]
- [[targeted-drug-delivery]]
- [[xenotransplantation]]

## References

[^wang2001]: `paper` Wang, L., Brock, A., Herberich, B., Schultz, P. G. "Expanding the genetic code of Escherichia coli." *Science*, 2001.
[^lajoie2013]: `paper` Lajoie, M. J. et al. "Genomically recoded organisms expand biological functions." *Science*, 2013. {The codons were replaced by editing a living strain rather than by synthesizing the genome, and the resulting resistance was to particular phages, not the broad resistance reported later.}
[^mandell2015]: `paper` Mandell, D. J. et al. "Biocontainment of genetically modified organisms by synthetic protein design." *Nature*, 2015. {A detection-limit result rather than a measured escape rate: the assay bounds escape at the population sizes tested in laboratory culture.}
[^fredens2019]: `paper` Fredens, J. et al. "Total synthesis of Escherichia coli with a recoded genome." *Nature*, 2019.
[^robertson2021]: `paper` Robertson, W. E. et al. "Sense codon reassignment enables viral resistance and encoded polymer synthesis." *Science*, 2021.
[^nyerges2023]: `paper` Nyerges, A. et al. "A swapped genetic code prevents viral infections and gene transfer." *Nature*, 2023.

==============================================================================
ARTICLE: genetic-discrimination
TITLE:   Genetic discrimination
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/genetic-discrimination
SOURCE:  https://futurehumanwiki.com/raw/genetic-discrimination
==============================================================================

---
title: "Genetic discrimination"
slug: "genetic-discrimination"
type: "concept"
status: "established"
horizon: "present"
categories: ["society", "genetics"]
tags: ["law", "privacy", "insurance", "employment", "gina", "genomics"]
summary: "Differential treatment of people on the basis of genetic information, and the uneven legal regime that restricts it in employment and health cover but rarely elsewhere."
updated: "2026-07-27"
issues: ["The consumer-genomics breach, the 2025 bankruptcy, and the Florida and Australia changes are uncited."]
---

```infobox
{
  "caption": "Legal and policy problem",
  "rows": [
    { "label": "Main US statute", "value": "GINA, 2008" },
    { "label": "GINA covers", "value": "Health insurance, employment" },
    { "label": "GINA excludes", "value": "Life, disability, long-term care" },
    { "label": "Canada", "value": "Criminal prohibition, 2017" },
    { "label": "United Kingdom", "value": "Industry code, not law" },
    { "label": "Enforcement volume", "value": "Low" }
  ]
}
```

**Genetic discrimination** is the use of a person's genetic information — a test result, a family history, or a predicted risk — to treat them less favourably in insurance, employment, education, or another domain where the information is not relevant to present ability. The concern predates the ability to sequence a genome cheaply and has produced a patchwork of protections that is strongest where the risk was anticipated and weakest where the information is now actually used.

## What is being protected

The canonical case is a person who learns they carry a pathogenic variant for a late-onset condition — Huntington's disease is the standard example because penetrance is near total and onset is decades away — and is then refused cover or employment on the strength of a prediction about a future they have not yet reached. The harm has two parts. One is the adverse decision. The other is the chilling effect: people who fear the consequences decline testing that would benefit them clinically, which is the effect surveys of at-risk families most consistently find. The same reluctance could extend to predictive tests that are not genetic at all, including the direct-to-consumer panels covered in [[consumer-blood-testing]] and the multi-cancer screening assays described in [[nanoparticle-diagnostics]], where a positive result may reach an underwriter long before it becomes a diagnosis.

Whether genetic information deserves special treatment is contested. The position called genetic exceptionalism holds that it is distinctively sensitive, because it is predictive, shared with relatives, and immutable — the last of which ceases to be true if the editing methods described in [[somatic-gene-therapy]] mature.[^murray1997] Critics note that family history, cholesterol, and an HIV test are also predictive and shared, that most genetic variants carry weak probabilistic information rather than a verdict, and that carving out genetics produces an incoherent privacy regime in which the same fact is protected or unprotected depending on how it was learned.

Sequencing has strengthened the critics' case. Everyone carries variants associated with elevated risk of something. Once whole-genome data is routine, protecting "genetic test results" as a category protects almost nothing, because the discriminating input is the interpretation rather than the sequence — a point that also complicates [[polygenic-embryo-screening]], where the output is a probability rather than a diagnosis.

## GINA and its gaps

The US Genetic Information Nondiscrimination Act, enacted in 2008 after more than a decade in Congress, prohibits health insurers from using genetic information in eligibility, premiums, or underwriting, and prohibits employers with fifteen or more employees from using it in hiring, firing, or terms of employment. It also bars employers from requesting or purchasing genetic information, including family medical history, with narrow exceptions.

Its exclusions define the current problem.

**Life, disability, and long-term care insurance are not covered.** These are the products for which a predictive genetic result is most actuarially useful, and in most US states insurers may lawfully ask for and act on test results. Florida became the first state to close the gap in 2020, barring life, disability, and long-term-care insurers from using genetic test information; a small number of other states have since restricted it in narrower ways.

**Manifest disease is not covered.** Once a condition has appeared, it is a medical condition rather than genetic information, and GINA no longer applies. Protection is therefore strongest for the asymptomatic and disappears when the risk materialises.

**Small employers, the military, and several federal programmes fall outside.** The armed services are not bound by GINA, and genetic findings can affect assignment and retention. The exclusion matters more than its size suggests, since military medicine is where several enhancement-adjacent technologies are first adopted, as [[enhancement-arms-race]] describes.

**Selection is untouched.** GINA governs what may be done with a living person's genetic information. It says nothing about screening embryos, so the practices covered in [[embryo-selection]] fall entirely outside it.

Enforcement has been thin. Regulators have brought a modest number of actions, most concerning employers collecting family medical history in pre-employment physicals or wellness programmes rather than sequencing anyone. The best-known US episode predates the statute: a railway company was found in 2001 to have secretly tested employees who had filed carpal-tunnel claims for a variant it believed predisposed them to the condition, and settled with the federal employment regulator the following year.

> [!caution] Absence of complaints is weak evidence
> Low case volume is often cited as showing that genetic discrimination is rare. It is equally consistent with the discrimination being undetectable: an applicant declined cover is rarely told which item in the file was decisive, and an underwriting model that weights a family history does not announce itself.

## Outside the United States

Canada enacted the strongest instrument, a 2017 statute making it a criminal offence to require a genetic test or the disclosure of results as a condition of a contract or service. Its constitutionality was challenged on federalism grounds — insurance is a provincial matter — and the Supreme Court of Canada upheld it as a valid exercise of the criminal law power in 2020.[^scc2020]

The United Kingdom relies on industry self-regulation. Under a code agreed between government and insurers, companies do not ask for predictive genetic test results, with a single exception for Huntington's disease on life cover above a high threshold. It has held for two decades and has no statutory force; the equivalent Australian arrangement was a voluntary industry moratorium, and in 2024 the Australian government announced it would legislate a ban on the use of genetic test results in life insurance underwriting.

The Council of Europe's Oviedo Convention prohibits discrimination on grounds of genetic heritage among its parties, and EU data protection law treats genetic data as a special category requiring an explicit lawful basis. Neither settles the underwriting question directly, and jurisdictions have reached opposite conclusions about whether an insurer may use information the applicant already holds. The same convention supplies the only binding international prohibition on heritable modification, which is why it recurs throughout [[governance-of-genome-editing]] and the debates over [[germline-editing]].

## The adverse-selection argument

Insurers make a real argument against prohibition. If applicants may know their genetic risk while insurers may not, the high-risk buy more cover, premiums rise, and the low-risk exit — the classic adverse-selection spiral. Where health cover is universal and publicly funded the argument does not arise, which is why the issue is sharpest in countries with private health and life markets.

The empirical picture is less alarming than the theory. Studies of markets operating under a genetic-testing moratorium have not found the predicted destabilisation, plausibly because the number of people holding a strongly predictive result remains small relative to the pool and because most policies are bought before anyone tests. The distributive stakes also run the other way from those in [[access-and-inequality]]: here the worry is not that a treatment is unaffordable but that knowing one may need it makes cover unobtainable. Whether that holds as polygenic risk scoring becomes routine is unknown, and it is the version of the question that matters for the next decade.

## Consumer genomics and the data itself

Direct-to-consumer testing put tens of millions of genotypes into private databases governed by terms of service rather than by health-privacy law. Three developments have made that arrangement look fragile.

Law enforcement access came first: the 2018 identification of a serial offender in California through a genealogy database showed that a relative's upload can expose a person who never tested.[^erlich2018] Forensic genetic genealogy is now routine in US investigations, and the individuals whose data enables it did not consent to that use.

Security followed. A large consumer testing company disclosed in 2023 that attackers using credentials stolen elsewhere had accessed millions of profiles through a relative-matching feature — an attack the architecture invited, since the product's purpose is to connect one account's data to another's.

Insolvency completed the set. The same company entered Chapter 11 bankruptcy protection in 2025, and its genetic database was treated as an asset in a court-supervised sale. Several US state attorneys general advised residents to delete their accounts. No US federal law establishes what happens to a genomic database when its custodian fails, and the terms of service under which the data was collected are contracts that a bankruptcy court may modify.

## Where it is heading

Three shifts are already straining the existing framework.

Polygenic scores turn discrimination from a question about rare variants into a question about everyone, since a score can be computed for any trait from a standard genotype array — the mechanism behind [[genetic-enhancement-of-intelligence]] and the screening applications marketed under the heading of [[designer-babies]]. Insurers do not need a diagnosis to price a risk; a distribution will do.

Population genomic programmes are placing sequence data into national health systems, where the governance question is not whether an insurer may see it but whether a state may.

And the arrival of heritable modification, if it ever occurs, creates a category the statutes do not contemplate: discrimination not on the basis of what a person's genome is but on the basis of what was done to it. Nothing in existing law reaches it, and the parallel debate over [[neurorights]] has begun to raise the same structural question for neural data. Whether the correct response is another category-specific statute, or a general prohibition on unjustified prediction-based discrimination of the kind [[disability-rights-and-enhancement]] scholarship has long argued for, remains unsettled in every jurisdiction that has legislated so far.

## See also

- [[mental-privacy]]
- [[neurorights]]
- [[polygenic-embryo-screening]]
- [[access-and-inequality]]
- [[governance-of-genome-editing]]
- [[embryo-selection]]
- [[disability-rights-and-enhancement]]
- [[designer-babies]]

## References

[^murray1997]: `book` Murray, T.H. "Genetic Exceptionalism and 'Future Diaries': Is Genetic Information Different from Other Medical Information?" In *Genetic Secrets*, Yale University Press, 1997. {The chapter names genetic exceptionalism in order to argue against it, and predates cheap whole-genome sequencing.}
[^scc2020]: `law` Supreme Court of Canada. *Reference re Genetic Non-Discrimination Act*, 2020 SCC 17. {The ruling settles which level of government may legislate, not whether the prohibition changes underwriting practice.}
[^erlich2018]: `paper` Erlich, Y., Shor, T., Pe'er, I., Carmi, S. "Identity inference of genomic data using long-range familial searches." *Science*, 2018. {A quantitative analysis of how much database coverage is needed to reach most people through a relative's record.}

==============================================================================
ARTICLE: genetic-enhancement-of-intelligence
TITLE:   Genetic enhancement of cognition
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/genetic-enhancement-of-intelligence
SOURCE:  https://futurehumanwiki.com/raw/genetic-enhancement-of-intelligence
==============================================================================

---
title: "Genetic enhancement of cognition"
slug: "genetic-enhancement-of-intelligence"
type: "concept"
status: "speculative"
horizon: "2040s"
categories: ["enhancement", "genetics"]
tags: ["cognition", "polygenic scores", "gwas", "enhancement", "genomics", "heritability"]
summary: "The proposal to raise human cognitive ability by selecting or editing genetic variants, constrained by the extreme polygenicity of the trait and by pleiotropy."
updated: "2026-07-27"
humanEvidence: "No human has been genetically enhanced for cognition; the human data are polygenic scores whose accuracy between siblings, the comparison an embryo choice actually faces, is well below their accuracy across strangers."
issues: ["The infobox heritability range of 50-80 per cent is given without a citation."]
---

```infobox
{
  "caption": "Proposed form of human enhancement",
  "rows": [
    { "label": "Target trait", "value": "Cognitive ability" },
    { "label": "Heritability in adults", "value": "Roughly 50–80%" },
    { "label": "Architecture", "value": "Thousands of tiny effects" },
    { "label": "Proposed routes", "value": "Selection, editing, gene therapy" },
    { "label": "Demonstrated in humans", "value": "No" },
    { "label": "Demonstrated in animals", "value": "Single-gene learning effects only" }
  ]
}
```

**Genetic enhancement of cognition** is the proposal to raise human cognitive ability by choosing or altering genetic variants, either by selecting among embryos or by editing them. Cognitive ability is substantially heritable, which is what makes the proposal seem tractable. It is also among the most polygenic traits ever mapped, which is what makes it intractable in practice. The distance between those two facts accounts for most of the disagreement about whether the idea is a near-term prospect or a category error.

```keyfacts
[
  { "value": "3M+", "label": "Participants in the largest educational-attainment GWAS", "note": "Okbay et al., 2022" },
  { "value": "12–16%", "label": "Variance explained between families", "note": "substantially less within families, where selection operates" },
  { "value": "0", "label": "Cognitive gene therapies attempted in humans", "note": "as of 2026" }
]
```

## What the genetics shows

Twin, adoption and family studies place the heritability of adult cognitive test performance somewhere in the range of one half to four fifths, rising through childhood into adulthood. Heritability estimated from measured common variants is considerably lower, on the order of a fifth to a quarter, and the polygenic scores built from those variants explain less again.

The largest relevant studies target educational attainment, a proxy that is cheap to collect at biobank scale. A study of about three million individuals produced a score explaining something in the region of twelve to sixteen per cent of the variance in years of schooling in independent samples.[^okbay2022] Direct genome-wide studies of measured intelligence are smaller and identify hundreds of loci.[^savage2018] In both cases the individual variants are minuscule: the largest common-variant effects correspond to a small fraction of a point on a standardised test.

Two asymmetries follow. First, there is no small set of high-effect "intelligence variants" to install. Second, the variants of large effect that do exist are almost uniformly harmful, because mutations that disrupt neurodevelopment produce intellectual disability. Nature offers a long list of ways to break cognition by changing one gene and essentially no examples of improving it that way.

> [!key] Polygenicity is the whole problem
> A trait built from ten thousand variants of a hundredth of a point each cannot be enhanced by a technique that acts on one locus at a time. Every proposed route has to solve the many-loci problem before anything else.

## Proposed routes

### Embryo selection

The nearest-term route is [[polygenic-embryo-screening]], ranking embryos by a cognitive or educational-attainment score. The expected gain is set by how much genetic variation exists among one couple's embryos, how accurate the score is between siblings, and how many embryos are available. Modelling by Karavani and colleagues put the expected shift at a couple of points at most when choosing the best of ten, with wide variance around that expectation, and less again for the number of embryos a real cycle produces.[^karavani2019]

Two proposals attempt to relax the binding constraint. [[in-vitro-gametogenesis]] would generate large numbers of embryos from a couple's cells, raising selection intensity. Iterated embryo selection, described by Carl Shulman and Nick Bostrom, would run several rounds of gamete derivation and selection in vitro to compound gains across simulated generations without waiting for real ones.[^shulman2014] Neither has been demonstrated in humans, and the second requires deriving gametes from embryonic stem cells at a scale nobody has approached.

### Editing

[[germline-editing]] could in principle install alleles that neither parent carries, escaping the sibling-variance limit entirely. Three obstacles stand in front of that, and only one of them concerns cutting DNA.

The first is that association studies do not hand over a list of variants to install. A genome-wide hit marks a stretch of chromosome correlated with the trait; the base actually responsible may be any of hundreds nearby. Statistical fine-mapping resolves a minority of loci to a plausible causal variant, and for a trait with thousands of contributing regions the causal set is mostly uncharacterised. Editing a marker rather than a cause changes nothing.

The second is scale. Effects measured in hundredths of a point mean that a shift worth having requires very many substitutions in one cell. The highest multiplex edit counts on record were achieved in cultured cell lines, which can be screened, expanded and thrown away; an embryo offers a single attempt, and confirming what happened to it means consuming cells it needs.

The third is damage. Every additional cut raises the probability of the large deletions and rearrangements documented in [[crispr-off-target-effects]], and a one-cell embryo has no spare copies. [[base-editing]] and [[prime-editing]] rewrite bases without severing both strands and are the only plausible chemistry for a high-multiplex attempt, but their behaviour across many loci at once in a human embryo has not been measured.

### Somatic intervention

Editing an adult brain is a different proposition. Most of the relevant biology acts during development, delivery to the central nervous system is the standing obstacle for [[aav-vectors]], and [[somatic-gene-therapy]] for a non-disease indication would face a regulatory bar no sponsor has tried to clear. Interventions aimed at adult cognition are, for now, pharmacological rather than genetic, and the evidence there is covered in [[nootropics]].

## Animal and biological evidence

The clearest single-gene demonstration remains a mouse. Overexpressing the NR2B subunit of the NMDA receptor in the forebrain produced animals that learned faster on several tasks.[^tang1999] The result held up, but so did a subsequent finding that the same animals showed heightened sensitivity to inflammatory pain, an early and instructive illustration of pleiotropy.

A second line concerns klotho, a protein better known in aging research. Human carriers of one copy of the KL-VS variant show slightly better performance on some cognitive measures, and administering klotho to aged rhesus monkeys improved memory performance in a controlled study.[^castner2023] This is a genuine cross-species signal, and it is also narrow: a single, modest, hormone-mediated effect, not a general method for moving a polygenic trait.

Comparative and evolutionary arguments are sometimes offered as evidence that large gains are available, on the grounds that human cognition changed rapidly in evolutionary time. That change involved coordinated modification of developmental programmes over hundreds of thousands of years, not the substitution of a few alleles.

## Why this is harder than it sounds

**Within-family attenuation.** A score's advertised accuracy is measured across unrelated strangers, and a substantial part of what it measures is not the child's own DNA. Some of it is the parents' behaviour, which correlates with their genotype and shapes the household. Some of it is ancestry and mating pattern, which align genotype with social position. Sibling comparisons strip those out, because siblings share parents, household and ancestry; what survives is the direct effect, and only the direct effect is available to an embryo choice. For educational attainment the surviving fraction is well below the headline figure, which matters because educational attainment is where nearly all the discovery power sits.

**Pleiotropy.** Genome-wide genetic correlations link educational attainment to a range of psychiatric outcomes, positively for some conditions and negatively for others.[^bulik2015] Selecting or editing on a cognitive score therefore moves risk for other traits in ways that are only partly characterised. The direction of some of these correlations is itself contested, since they are estimated from the same confounded population data.

**Gene–environment interplay.** Population mean scores on cognitive tests rose substantially across the twentieth century in many countries, a change no genetic mechanism explains. Whatever produced that shift was environmental, and it was larger than anything genetic selection currently offers.

**Measurement.** Test performance, educational attainment and cognition in any richer sense are not the same variable. Optimising the measurable proxy is not obviously optimising the thing of interest, a problem that recurs throughout [[human-enhancement]].

> [!caution] Contested claims
> Claims that cognitive enhancement by embryo selection is now practical rest on population-level score accuracy applied to a within-family decision. Mainstream statistical genetics regards that substitution as invalid, and no company offering the service has published outcome data on selected children.

## Ethics and politics

The arguments split along familiar lines. [[procreative-beneficence]] supplies the case that parents have reason to select for cognitive advantage as for any other expected benefit. Against it, [[disability-rights-and-enhancement]] raises the expressivist objection, and [[access-and-inequality]] raises the prospect that a cognitive advantage available only to the wealthy compounds existing stratification faster than any other enhancement would. Because cognitive advantage is substantially positional, the dynamics analysed in [[enhancement-arms-race]] apply with particular force: if the gain is competitive rather than absolute, universal adoption leaves everyone where they started while consuming real resources.

There is also a research-conduct problem. Cognitive genomics has a history entangled with eugenics, and much of the field's caution about enhancement framing reflects that history rather than the statistics alone. Several of the companies marketing cognitive scores for embryos have been criticised by the same researchers whose summary statistics they use.

## Outlook

The most likely near-term development is not a working enhancement but better evidence about how little the current tools deliver. Within-family genome-wide studies at biobank scale are the technically decisive experiment: they estimate direct genetic effects, and the size of those estimates sets a ceiling on what any selection or editing scheme could achieve. If within-family effects on cognition prove as attenuated as they are for educational attainment, the enhancement case rests on multiplex editing, which is not close.

The harder question is what a positive result would mean. A validated within-family score capable of shifting cognition by a few points would sit alongside interventions in nutrition, schooling and air quality that are known to move the same outcome by comparable amounts at a fraction of the cost, and would still be argued about far more.

Two things would change that calculation. The first is a mechanistic result rather than a statistical one: a variant whose effect on neural development is understood well enough to be engineered rather than merely selected. Nothing on the current maps looks like that, and the pleiotropy problem suggests few candidates will. The second is a shift in what is being enhanced. Most of the argument assumes general cognitive ability as measured by existing tests, which is the trait with the largest and least tractable polygenic architecture. Narrower targets — working-memory capacity, resistance to age-related cognitive decline, the specific deficits of a diagnosed condition — are smaller problems and land closer to therapy than to enhancement.

That last point is where the debate is most likely to actually go. The line between preventing decline and raising a baseline is not sharp, and interventions developed for [[proteostasis|neurodegeneration]] will be the first to test how societies police it.

## See also

- [[polygenic-embryo-screening]]
- [[designer-babies]]
- [[germline-editing]]
- [[in-vitro-gametogenesis]]
- [[nootropics]]
- [[intelligence-amplification]]
- [[bioethics-of-enhancement]]
- [[enhancement-arms-race]]

## References

[^okbay2022]: `paper` Okbay, A. et al. "Polygenic prediction of educational attainment within and between families from genome-wide association analyses in 3 million individuals." *Nature Genetics*, 2022. {The sample is of European ancestry, and scores built from it predict substantially worse in other ancestry groups.}
[^karavani2019]: `paper` Karavani, E. et al. "Screening Human Embryos for Polygenic Traits Has Limited Utility." *Cell*, 2019. {The gains are simulated from real sibling genotypes; no embryo was selected or transferred in the study.}
[^savage2018]: `paper` Savage, J. E. et al. "Genome-wide association meta-analysis in 269,867 individuals identifies new genetic and functional links to intelligence." *Nature Genetics*, 2018.
[^shulman2014]: `paper` Shulman, C. and Bostrom, N. "Embryo Selection for Cognitive Enhancement: Curiosity or Game-changer?" *Global Policy*, 2014.
[^tang1999]: `paper` Tang, Y.-P. et al. "Genetic enhancement of learning and memory in mice." *Nature*, 1999.
[^castner2023]: `paper` Castner, S. A. et al. "Longevity factor klotho enhances cognition in aged nonhuman primates." *Nature Aging*, 2023. {Klotho was injected as a protein into aged rhesus monkeys rather than delivered as a gene; the equivalent experiment has not been reported in people.}
[^bulik2015]: `paper` Bulik-Sullivan, B. et al. "An atlas of genetic correlations across human diseases and traits." *Nature Genetics*, 2015.

==============================================================================
ARTICLE: george-church
TITLE:   George Church
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/george-church
SOURCE:  https://futurehumanwiki.com/raw/george-church
==============================================================================

---
title: "George Church"
slug: "george-church"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "genetics"]
tags: ["genomics", "synthetic biology", "gene editing", "de-extinction", "xenotransplantation", "sequencing"]
summary: "American geneticist who helped invent direct genomic sequencing, founded the Personal Genome Project, and co-founded dozens of biotechnology companies."
updated: "2026-07-27"
issues: ["The Epstein funding, the 2019 dating-app proposal, and the dire wolf criticism are uncited."]
---

```infobox
{
  "caption": "Geneticist and synthetic biologist",
  "rows": [
    { "label": "Born", "value": "28 August 1954, Florida" },
    { "label": "Nationality", "value": "American" },
    { "label": "Education", "value": "PhD, Harvard University, 1984" },
    { "label": "Known for", "value": "Genomic sequencing; genome recoding", "link": "/wiki/recoded-organisms" },
    { "label": "Field", "value": "Genomics; synthetic biology" },
    { "label": "Affiliation", "value": "Harvard Medical School; Wyss Institute" },
    { "label": "Companies co-founded", "value": "Dozens, including eGenesis and Colossal", "link": "/wiki/colossal-biosciences" }
  ]
}
```

**George Church** is an American geneticist whose laboratory at Harvard Medical School has produced
foundational methods in DNA sequencing, multiplexed genome engineering and genome recoding, and
whose alumni and spin-out companies occupy much of contemporary synthetic biology. He is
unusual among senior scientists in treating publicly provocative projects — reversing extinction,
editing pigs for human transplant, writing data into living genomes — as legitimate research
programmes rather than thought experiments.

## Career

Church was dismissed from a doctoral programme at Duke in the 1970s after neglecting coursework in
favour of crystallography, and completed his PhD at Harvard in 1984 with Walter Gilbert. His thesis
work introduced direct genomic sequencing and multiplexed approaches that anticipated later
high-throughput methods.[^church1984] He attended the 1984 meeting in Alta, Utah, that first
seriously discussed sequencing the human genome, and has been involved in genome-scale projects
ever since. He is professor of genetics at Harvard Medical School and a core faculty member of the
Wyss Institute.

He has spoken publicly about narcolepsy and dyslexia, and has credited both with shaping how he
works — the first by making conventional deadlines unworkable, the second by pushing him toward
spatial and structural reasoning.

## Key contributions

### Sequencing and reading genomes

Church's laboratory developed and commercialized methods that fed into the second generation of
sequencing platforms, and its cost curve work helped drive per-genome prices down by orders of
magnitude across the 2000s. In 2005 he launched the Personal Genome Project, which asked
participants to publish their genomes and phenotypes under open consent rather than promised
anonymity — a deliberate argument that genomic privacy in the conventional sense is not achievable
and that informed volunteers should be allowed to say so. The project's premise remains contested;
see [[genetic-discrimination]].

### Writing and rewriting genomes

Multiplex automated genome engineering, published in 2009, made many simultaneous small edits to a
bacterial genome and selected across the resulting population, treating genome engineering as a
directed-evolution problem rather than a series of single edits.[^wang2009] His group then
systematically replaced every instance of one stop codon in *Escherichia coli*, producing an
organism whose genetic code differed from the natural one and which was resistant to viruses that
depend on the standard code.[^lajoie2013] That line of work underpins [[recoded-organisms]] and the
biocontainment strategies built on synthetic auxotrophy, and connects to the broader
[[synthetic-genomes]] programme.

In 2013 his laboratory published one of the first demonstrations of RNA-guided editing in human
cells, appearing alongside a report from Feng Zhang's group. His lab has also encoded books and
image sequences in DNA, both as a demonstration of storage density and as a test of writing
capacity.

### Xenotransplantation and de-extinction

Church's group used CRISPR to inactivate dozens of porcine endogenous retrovirus sequences in pig
cells, and later produced live PERV-inactivated piglets — removing one of the classical objections
to [[xenotransplantation]].[^niu2017] The company eGenesis was founded on that work. He
co-founded [[colossal-biosciences]] in 2021 around [[de-extinction]] programmes for the woolly
mammoth, thylacine and dodo, an effort he has consistently described in terms of engineering
cold-tolerant Asian elephants rather than recreating an extinct species.

> [!caution] What "de-extinction" delivers
> Editing a small number of trait-associated genes into a living relative produces a proxy, not a
> resurrected species. Church has said this himself; press coverage of Colossal's announcements
> frequently has not, and the 2025 dire wolf claim drew sharp criticism from paleogeneticists on
> exactly this point.

### Aging

Church's aging work runs through Rejuvenate Bio, which pursues combination gene therapy delivered
by [[aav-vectors]] — constructs such as [[myostatin-inhibition|follistatin]], soluble TGF-beta
receptor and klotho, tested in mice and dogs. He has argued that [[gene-therapy-for-aging]] is more
tractable than small-molecule geroscience because a single delivery event can produce durable
expression, and that combinations matter more than any single gene. The results so far are animal
results; no human trial has demonstrated an effect on aging.

## Reception

Church is widely regarded as one of the most generative laboratory heads in biology, measured by
methods adopted and by former trainees running their own programmes — Kevin Esvelt's
[[gene-drive]] work and Luhan Yang's xenotransplantation company among them. He is also the most
frequent target of the charge that provocative announcements outrun data. A 2013 interview about
whether a Neanderthal genome could in principle be brought to term was widely misreported as a
recruitment call for a surrogate. A 2019 proposal for a genetic-compatibility dating application
drew accusations of eugenic framing, which he disputed. He accepted research funding from Jeffrey
Epstein and apologized publicly for the association in 2019.

His scientific critics tend to make a narrower point: the distance between a demonstration in a
cell line or a mouse and a usable human therapy is systematically longer than his public statements
imply, and his simultaneous roles as investigator, founder and adviser make his timelines difficult
to read as neutral estimates. Church's answer has generally been that stating an ambitious goal is
how funding and talent get organized, and that he separates what has been done from what is
proposed.

## Legacy

The methods are the durable part: direct genomic sequencing, multiplexed editing, code expansion,
PERV-free pigs. The companies are a second kind of legacy, and a more ambiguous one — a
substantial fraction of the commercial genome-editing sector traces to his laboratory, which
concentrates both capability and conflict of interest in an unusual way.

The open question his programme poses most sharply is where recoding stops. An organism with a
compressed genetic code is virus-resistant and biologically contained, which is attractive; the same
techniques applied to human cells would produce something whose relationship to the species is
unclear, and there is currently no framework for evaluating it. See
[[governance-of-genome-editing]] and [[mirror-life]] for the two nearest attempts.

## See also

- [[recoded-organisms]]
- [[synthetic-genomes]]
- [[xenotransplantation]]
- [[colossal-biosciences]]
- [[de-extinction]]
- [[gene-drive]]
- [[jennifer-doudna]]
- [[crispr-cas9]]

## References

[^church1984]: `paper` Church, G. M. and Gilbert, W. "Genomic sequencing." *Proceedings of the National Academy of Sciences*, 1984. {A methods paper: it reads sequence directly from genomic DNA without cloning, and reports no genome.}
[^wang2009]: `paper` Wang, H. H. et al. "Programming cells by multiplex genome engineering and accelerated evolution." *Nature*, 2009. {The method works in Escherichia coli and depends on oligonucleotide recombineering, which has no direct equivalent in human cells.}
[^lajoie2013]: `paper` Lajoie, M. J. et al. "Genomically recoded organisms expand biological functions." *Science*, 2013.
[^niu2017]: `paper` Niu, D. et al. "Inactivation of porcine endogenous retrovirus in pigs using CRISPR-Cas9." *Science*, 2017. {The paper reports live piglets with the retroviral sequences removed; it reports no transplant into a primate or a person.}

==============================================================================
ARTICLE: geroscience-hypothesis
TITLE:   Geroscience hypothesis
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/geroscience-hypothesis
SOURCE:  https://futurehumanwiki.com/raw/geroscience-hypothesis
==============================================================================

---
title: "Geroscience hypothesis"
slug: "geroscience-hypothesis"
type: "concept"
status: "emerging"
horizon: "2030s"
categories: ["longevity", "foundations"]
tags: ["aging", "chronic disease", "clinical trials", "regulation", "biogerontology"]
summary: "The proposition that aging is the shared upstream driver of most chronic disease, so that slowing it would delay many conditions at once rather than one at a time."
updated: "2026-07-27"
humanEvidence: "No randomised trial has shown that any intervention slows human aging; the human support is observational, from long-lived families and from an Ecuadorian Laron cohort with less cancer and diabetes but no longer life."
issues: ["The three-year competing-risk figure is attributed to standard estimates without a citation."]
---

```infobox
{
  "caption": "Organizing hypothesis in biogerontology",
  "rows": [
    { "label": "Field", "value": "Geroscience" },
    { "label": "Core claim", "value": "Aging drives chronic disease" },
    { "label": "Formalised", "value": "c. 2013–2014" },
    { "label": "Key framework", "value": "Seven pillars of aging" },
    { "label": "Test case", "value": "TAME trial", "link": "/wiki/metformin" },
    { "label": "Status", "value": "Strong animal support, no human proof" }
  ]
}
```

**Geroscience hypothesis** is the proposition that biological aging is the principal upstream driver of most chronic disease, and that an intervention slowing aging would therefore delay cancer, cardiovascular disease, dementia, diabetes and frailty together. It reframes the target of medicine from individual pathologies to the process that makes all of them likely. The animal evidence for it is substantial. The human evidence is almost entirely absent.

```keyfacts
[
  { "value": "7–8 years", "label": "Doubling time of adult mortality risk", "note": "the Gompertz relation, the field's founding observation" },
  { "value": "7", "label": "Pillars of aging", "note": "Kennedy and colleagues, 2014" },
  { "value": "~3 years", "label": "Life expectancy at birth gained by eliminating all cancer deaths", "note": "standard competing-risk estimates; the argument for a different target" }
]
```

## The argument

Adult mortality risk rises roughly exponentially with age, doubling every seven to eight years, and the incidence curves for most chronic diseases follow it closely. Someone at seventy is not merely more likely to have cancer than someone at thirty; they are also more likely to have heart disease, dementia, kidney disease and osteoporosis, and to have several at once. The geroscience argument reads this covariance as evidence of a common cause.

The competing-risk arithmetic supplies the practical motivation. Because an older person who avoids one fatal disease usually dies of another, eliminating any single disease adds surprisingly little to population life expectancy. Removing all cancer deaths would add roughly three years at birth by standard estimates. An intervention that shifted the entire aging trajectory, by contrast, would push every curve to the right simultaneously.

> [!key] Why the framing changes what gets funded
> Under the disease-by-disease model, a drug that reduced frailty, improved immune response to vaccination and delayed dementia by a modest amount in one trial would have no regulatory home. Under the geroscience framing, that combination is the point.

## Origins

The intellectual precursors run through Gompertz's 1825 mortality law, through James Fries's 1980 argument for [[compression-of-morbidity]], and through the demonstration in the 1980s and 1990s that single-gene mutations could extend lifespan in nematodes and mice while delaying multiple age-related pathologies, of which
[[cynthia-kenyon|the 1993 daf-2 result in nematodes]] is the best known.

The modern formulation dates to the early 2010s. A trans-NIH Geroscience Interest Group formed in 2012 and convened a summit the following year, and a 2014 paper by Brian Kennedy and colleagues set out seven "pillars of aging" as a shared conceptual base: adaptation to stress, epigenetics, inflammation, macromolecular damage, metabolism, proteostasis, and stem cells and regeneration.[^kennedy2014] The pillars overlap heavily with the [[hallmarks-of-aging]] and were framed more explicitly around translation, grouping mechanisms such as [[proteostasis]] failure, [[inflammaging|chronic inflammation]] and [[stem-cell-exhaustion]] as shared substrates rather than as separate research areas. In parallel, S. Jay Olshansky and colleagues argued the economic case that became the [[longevity-dividend]].[^olshansky2006] Institutions built around the framing, notably the [[buck-institute]], predate the label by more than a decade.

## Evidence in animals

The strongest support comes from interventions that extend lifespan in mammals and delay multiple pathologies at once. [[caloric-restriction]] does this across many species. [[rapamycin]] extends lifespan in genetically heterogeneous mice even when started late in life,[^harrison2009] and treated mice show delayed decline in immune and cognitive function. Growth-hormone-pathway mutants such as Ames dwarf and growth hormone receptor knockout mice live longer and develop tumours later. [[senolytics]] improve function across several tissues in aged mice.

Crucially, these interventions do not act on one disease. A rapamycin-fed mouse is not protected against cancer alone; decline is delayed in several tissues at once, although not uniformly, and some age-related pathologies proceed unchanged. That multi-system pattern is what the hypothesis predicts, and it is why the animal literature is treated as strong evidence.

## Evidence in humans

Human support is indirect. People who reach extreme old age tend to compress their period of illness rather than extend it, developing major disease later and dying after a shorter terminal decline than people who die in their seventies.[^andersen2012] Long-lived families show delayed onset across several unrelated conditions. Common variants associated with longevity are few and of small effect, but the familial clustering is real.

A revealing natural experiment is Laron syndrome, in which growth hormone receptor signalling is absent. An Ecuadorian cohort followed for decades showed near-absence of cancer and of type 2 diabetes, matching the mouse mutants, without a corresponding increase in lifespan; deaths came instead from accidents, alcohol-related causes and cardiovascular disease.[^guevara2011] The result supports the disease-delay half of the hypothesis and complicates the lifespan half.

What does not exist is a randomised human trial showing that any intervention slows aging. [[exercise-and-aging|Exercise]] is the best-evidenced practice for delaying multiple age-related outcomes, and it is not a drug. [[metformin]], [[nad-precursors]] and other candidates have not demonstrated the effect in controlled human studies.

## Objections

**The framework may be a taxonomy, not a theory.** Listing pillars or hallmarks does not establish that they share a driver, and interventions that hit one hallmark do not reliably move the others. Clearing senescent cells does not repair [[mitochondrial-dysfunction|mitochondrial damage]]; restoring [[autophagy|autophagic]] flux does not lengthen [[telomeres-and-telomerase|telomeres]]. Critics argue that "aging" names a correlated bundle of processes rather than a single manipulable variable, and that the appeal of the framing outruns its explanatory content.

**Mouse results translate badly.** The mammalian aging literature has produced few candidates that survive contact with human trials. Senolytic drugs are the clearest current example: dramatic mouse data followed by human results that have been modest or null.

**Trade-offs are underweighted.** Rapamycin is an immunosuppressant at transplant doses. Growth-hormone pathway suppression costs stature and, in mice, some cognitive and wound-healing capacity. An intervention that slows aging by slowing everything may not be one people want.

**Success might expand morbidity rather than compress it.** The empirical record in high-income countries over recent decades is closer to expansion of the years lived with disease than to compression, which is a point against the optimistic reading of what delaying aging would deliver.

## Regulatory implications

No major regulator recognises aging as an indication, so a drug cannot be approved to treat it. The exception is veterinary: the US Food and Drug Administration's animal-drug centre will consider lifespan extension itself as an indication, which is the route [[loyal]] is taking in dogs, and nothing about that pathway carries over to human approval. The workaround pursued by [[nir-barzilai]] and colleagues is TAME, designed less as a test of the drug itself than as a template: a trial whose primary endpoint is time to the first occurrence of any of several age-related diseases or death, which if accepted would establish that a multi-disease endpoint can support a label. It has never been fully funded, in part because a generic drug offers no sponsor a return, and it has not enrolled.

The alternative route is a validated surrogate. Without one, geroscience trials are long and expensive, which is the problem set out under [[aging-biomarkers]]. The field's difficulty is that its two paths to the clinic each depend on the other being easy.

## What would settle it

A clean test would take a single intervention, apply it to a population in a randomised trial, and show delayed onset across several unrelated age-related conditions relative to placebo, with functional measures moving in the same direction. Nothing of that design has reported. Intermediate evidence may come from companion dogs, whose lifespans are short enough to observe and whose environment resembles their owners'; work of this kind has faced funding instability. Until such a result exists, geroscience remains a well-motivated hypothesis with a strong animal literature and no human demonstration, and claims about [[longevity-escape-velocity]] or [[maximum-human-lifespan]] that assume it has been proven are assuming the conclusion.

## See also

- [[hallmarks-of-aging]]
- [[healthspan]]
- [[compression-of-morbidity]]
- [[longevity-dividend]]
- [[aging-biomarkers]]
- [[metformin]]
- [[rapamycin]]
- [[xprize-healthspan]]

## References

[^kennedy2014]: `paper` Kennedy, B.K. et al. "Geroscience: linking aging to chronic disease." *Cell*, 2014.
[^olshansky2006]: `paper` Olshansky, S.J., Perry, D., Miller, R.A., Butler, R.N. "In pursuit of the longevity dividend." *The Scientist*, 2006. {An essay in a science magazine making an economic case for funding aging research, not an empirical study.}
[^harrison2009]: `paper` Harrison, D.E. et al. "Rapamycin fed late in life extends lifespan in genetically heterogeneous mice." *Nature*, 2009. {Treatment began at 20 months of age, and the median lifespan gain was larger in females than in males.}
[^andersen2012]: `paper` Andersen, S.L. et al. "Health span approximates life span among many supercentenarians: compression of morbidity at the approximate limit of life span." *The Journals of Gerontology: Series A*, 2012. {The subjects are people who already reached extreme age, a group selected on the outcome, so the pattern need not generalise.}
[^guevara2011]: `paper` Guevara-Aguirre, J. et al. "Growth hormone receptor deficiency is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans." *Science Translational Medicine*, 2011.

==============================================================================
ARTICLE: glp-1-receptor-agonists
TITLE:   GLP-1 receptor agonists
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/glp-1-receptor-agonists
SOURCE:  https://futurehumanwiki.com/raw/glp-1-receptor-agonists
==============================================================================

---
title: "GLP-1 receptor agonists"
slug: "glp-1-receptor-agonists"
type: "intervention"
status: "established"
horizon: "present"
trl: 9
categories: ["longevity", "society"]
tags: ["obesity", "incretins", "diabetes", "clinical trials", "metabolic health", "drug access"]
summary: "Drugs that mimic the gut hormone GLP-1 to lower blood glucose and food intake, producing large weight loss that persists only while treatment continues."
updated: "2026-07-27"
humanEvidence: "Randomized trials in tens of thousands of adults show sustained weight loss and fewer cardiovascular events in people with overweight or obesity and established heart disease; no human trial has tested lifespan."
access: "Approved and widely prescribed in most high-income countries for diabetes and weight management; US list prices remain on the order of a thousand dollars a month and payer coverage is heavily restricted."
reversibility: "reversible"
issues: ["Lean-mass loss is described qualitatively; a sourced body-composition figure would be better.", "Pricing and coverage detail is US-centric and moves quickly."]
---

```infobox
{
  "caption": "Class of metabolic drugs",
  "rows": [
    { "label": "Drug class", "value": "Incretin mimetics" },
    { "label": "Molecular target", "value": "GLP-1 receptor" },
    { "label": "First approval", "value": "2005 (exenatide)" },
    { "label": "Main agents", "value": "Semaglutide, liraglutide, dulaglutide" },
    { "label": "Related", "value": "Tirzepatide (GIP/GLP-1 dual)" },
    { "label": "Approved for aging", "value": "No" },
    { "label": "Readiness", "value": "TRL 9" }
  ]
}
```

**GLP-1 receptor agonists** are drugs that activate the receptor for glucagon-like peptide-1, a
hormone released by the gut after a meal; most are engineered peptides rebuilt to survive in the
bloodstream for days rather than minutes. They lower blood glucose, slow gastric emptying, and act on
brainstem and hypothalamic circuits to reduce food intake, and in adults with obesity the newer
agents produce weight loss that no earlier drug approached. Whether that makes them a longevity
intervention is the argument this article is about: they demonstrably treat conditions that shorten
life, and nothing shows that they change the rate at which anyone ages.

```keyfacts
[
  { "value": "2005", "label": "First GLP-1 agonist approved", "note": "exenatide, derived from Gila monster venom peptide" },
  { "value": "~15%", "label": "Mean weight loss at 68 weeks", "note": "semaglutide in the STEP 1 trial; roughly 2% on placebo" },
  { "value": "0", "label": "Trials with a lifespan endpoint", "note": "as of mid-2026; mortality appears only as a secondary outcome in disease trials" }
]
```

## How they work

GLP-1 is secreted by enteroendocrine cells lining the distal small intestine and colon when nutrients
arrive. It amplifies insulin release in a glucose-dependent way, which is why these drugs rarely
cause hypoglycaemia on their own; it suppresses glucagon; it slows gastric emptying; and it binds
receptors in the hypothalamus and the area postrema, where the signal registers as
satiety.[^drucker2018] The native peptide is destroyed within a few minutes by the enzyme DPP-4, so
the therapeutic problem was never finding the hormone but making it last.

Two solutions dominate. The first came from a lizard: exendin-4, isolated from Gila monster venom, is
naturally resistant to DPP-4 and activates the human receptor, and became exenatide.[^eng1992] The
second is chemical modification of the human sequence, combining amino-acid substitutions with a
fatty-acid chain that binds circulating albumin and slows both enzymatic degradation and renal
clearance. That combination makes liraglutide a once-daily injection and semaglutide a once-weekly
one.

Most of the weight loss follows from eating less. That resemblance to a caloric deficit is what first
drew attention from biogerontology, though the analogy to [[caloric-restriction]] is loose: dietary
restriction experiments hold lean animals below their normal intake, while these drugs mainly remove
excess adiposity from people who have it. Whether any of the cardiovascular or renal benefit is
independent of weight loss is actively researched and unsettled.

> [!note] Terminology
> GLP-1 and GIP are the two incretin hormones. Semaglutide, made by Novo Nordisk and sold as Ozempic
> for type 2 diabetes and Wegovy for weight management, acts at the GLP-1 receptor only, as do
> liraglutide, dulaglutide, and exenatide. Tirzepatide, Eli Lilly's Mounjaro and Zepbound, is a
> **dual** GIP and GLP-1 receptor agonist and is not a GLP-1 agonist in the strict sense;
> retatrutide adds a glucagon receptor arm. Press coverage collapses all of these into "GLP-1s",
> which obscures real pharmacological differences.

## Development history

```timeline
[
  { "year": "1987", "title": "Incretin action established", "text": "Work in several laboratories shows that the GLP-1 fragment released after eating stimulates insulin secretion, identifying it as a physiological incretin." },
  { "year": "1992", "title": "Exendin-4 isolated", "text": "John Eng and colleagues characterize a DPP-4-resistant GLP-1 receptor agonist in the venom of the Gila monster, the peptide that becomes exenatide." },
  { "year": "2005", "title": "First approval", "text": "Exenatide is approved in the United States for type 2 diabetes, twice-daily by injection." },
  { "year": "2010", "title": "Liraglutide reaches market", "text": "Novo Nordisk's daily analogue is approved for type 2 diabetes; a higher-dose version is approved for weight management in 2014." },
  { "year": "2016", "title": "Cardiovascular benefit in diabetes", "text": "The LEADER trial reports fewer major cardiovascular events with liraglutide in people with type 2 diabetes at high cardiovascular risk." },
  { "year": "2017", "title": "Semaglutide approved", "text": "The once-weekly analogue is approved for type 2 diabetes; an oral tablet formulation follows in 2019." },
  { "year": "2021", "title": "Obesity result and shortage", "text": "STEP 1 reports roughly 15 percent mean weight loss over 68 weeks and semaglutide is approved for weight management; demand outstrips manufacturing capacity for the next several years." },
  { "year": "2022", "title": "Dual agonism", "text": "Tirzepatide, targeting both the GIP and GLP-1 receptors, is approved for type 2 diabetes; SURMOUNT-1 reports mean weight loss around a fifth of body weight at the highest dose." },
  { "year": "2023", "title": "Outcomes without diabetes", "text": "SELECT reports fewer major adverse cardiovascular events with semaglutide in adults with overweight or obesity and established cardiovascular disease but no diabetes; tirzepatide is approved for obesity." },
  { "year": "2024", "title": "Beyond weight and glucose", "text": "The FLOW trial, halted early for efficacy the previous year, reports slower kidney-disease progression with semaglutide in type 2 diabetes. Labels expand toward cardiovascular risk reduction and, for tirzepatide, obstructive sleep apnoea." }
]
```

## What the trials show

The weight results exceed those of any previously approved weight-loss drug. In STEP 1, adults with
overweight or obesity and without diabetes lost roughly 15 percent of body weight on average over 68
weeks on semaglutide, against about 2 percent on placebo.[^wilding2021] Tirzepatide, at its highest dose in
SURMOUNT-1, averaged around a fifth.[^jastreboff2022] Means conceal a wide spread: a substantial
minority lose very little, and both trials gave every participant diet and activity counselling
alongside the injection.

The strongest hard-outcome evidence is the SELECT cardiovascular outcomes trial, which enrolled
adults with overweight or obesity and established cardiovascular disease but without diabetes, and
reported a relative reduction of roughly
a fifth in a composite of cardiovascular death, non-fatal heart attack, and non-fatal stroke over
about three years.[^lincoff2023] The absolute difference was a small number of percentage points, in
a population already at high risk. Deaths from any cause were fewer in the treated group, a
difference the trial's statistical hierarchy did not establish as a confirmed result. Regulators
subsequently expanded semaglutide's label to cardiovascular risk reduction.

Other organ systems have followed. Semaglutide slowed progression of kidney disease in people with
type 2 diabetes and chronic kidney disease.[^perkovic2024] Trials have reported improved symptoms in
heart failure with preserved ejection fraction, reduced severity of
[[sleep-and-longevity|obstructive sleep apnoea]] with tirzepatide, and reduced knee osteoarthritis
pain. The failures matter as much: Novo Nordisk reported
in late 2025 that two large trials of oral semaglutide in early Alzheimer's disease did not slow
progression of the disease, the most direct test so far of the claim that incretin drugs act broadly
on age-related disease rather than through metabolism.[^novo2025]

## The geroscience argument

The case for treating this class as a longevity intervention is a population argument rather than a
biological one. Obesity and type 2 diabetes are among the largest modifiable contributors to
premature death, and they accelerate several of the [[hallmarks-of-aging]]: chronic low-grade
inflammation of the kind described under [[inflammaging]], accumulation of senescent cells in adipose
tissue as covered under [[cellular-senescence]], vascular and renal damage. A drug that reverses
those at scale could move population life expectancy further than any candidate geroprotector now in
trials has any prospect of doing, which is the [[longevity-dividend]] argument arriving from an
unexpected direction. Proponents add that one agent delaying several apparently unrelated
age-related conditions is the shape of result the [[geroscience-hypothesis]] predicts, and the shape
[[metformin|the TAME trial]] was designed to produce.

The case against is that this is disease treatment doing what disease treatment does. Restoring
someone with obesity toward the metabolic state of someone who never had it returns them to a
baseline; it does not push past it. Nothing in the record shows a change in the underlying rate of
aging, and the animal work, overwhelmingly in rodents, has measured metabolic and body-weight
endpoints rather than lifespan. The distinction is exactly the one that separates this class from
[[rapamycin]] and the other candidate
geroprotectors, whose whole claim is that they act on aging upstream of any particular disease. On
the evidence to date, GLP-1 agonists remove an accelerator; they do not slow the clock.

> [!debate] Removing an accelerator or slowing the clock
> If the whole benefit runs through weight and glucose, these are excellent drugs for two diseases
> and say nothing about aging. If part of it is direct receptor signalling in vessels, kidney, and
> brain, they belong in the geroprotector conversation. The Alzheimer's failure is evidence for the
> first reading. Nobody has run the trial that would settle it, and movement in an
> [[epigenetic-clock]] reading or another [[aging-biomarkers|composite aging measure]] would not
> settle it either, because a change in a prediction is not a change in an outcome.

## Limitations

Stopping reverses almost everything. The extension of STEP 1 followed participants after withdrawal
and found most of the lost weight returning within about a year, with blood pressure and lipid
improvements reverting alongside it.[^wilding2022] These are therefore chronic medications on the
model of antihypertensives, not courses of treatment, and real-world analyses report that a large
share of patients stop within the first year because of cost, side effects, or both. An intervention
that works only while it is being paid for has a different relationship to
[[healthspan|healthy lifespan]] than one that produces a durable change.

A substantial share of the weight lost is lean tissue rather than fat. Body-composition substudies
consistently report this, and the proportion is broadly what a caloric deficit of similar size
produces by any route, but it is a sharper concern here because the drugs are increasingly used by
older adults in whom muscle loss is itself a driver of frailty. Trials pairing incretin drugs with
antibodies against the activin and [[myostatin-inhibition|myostatin]] pathways are underway, and
resistance training is the countermeasure with the longest evidence base — one more place where
[[exercise-and-aging|the exercise literature]] is the comparator any drug has to beat.

## Risks

Gastrointestinal effects dominate the adverse-event profile: nausea, vomiting, diarrhoea, and
constipation are common, largely dose-related, and the main reason people discontinue. Gallbladder
disease is more frequent with rapid weight loss, and pancreatitis is rare and contested. Delayed
gastric emptying has prompted anaesthesiology societies to issue guidance on aspiration risk before
elective procedures. Signals raised in observational data are resolved case by case: a rare
optic-nerve condition prompted a European safety review that ended in a label change, while other
reported associations remain unconfirmed.

> [!caution] The rodent thyroid finding
> Long-term dosing produced thyroid C-cell tumours in rats and mice, which is the basis for a boxed
> warning and a contraindication in people with a personal or family history of medullary thyroid
> carcinoma. Rodent C-cells express the GLP-1 receptor far more abundantly than human ones do, and
> no corresponding human effect has been demonstrated. The warning stands because the question is
> open, not because the finding has been shown to transfer.

## Access and cost

Supply constrained the market for several years after 2021, and in the United States that shortage
legally permitted compounding pharmacies to sell copies, an unregulated parallel supply that
contracted once the shortages were declared resolved. Price is the binding constraint now. US list
prices sit on the order of a thousand dollars a month before rebates, several times what the same
products cost in other high-income countries; manufacturers have opened self-pay channels in the
mid-hundreds; payers have restricted eligibility or withdrawn coverage for weight indications on
budget grounds. Patents on the earlier agents have begun to lapse in some markets, the usual route to
a price collapse.

Obesity prevalence is rising fastest in middle-income countries, where these drugs are furthest out
of reach, making this a clean current example of the pattern described under
[[access-and-inequality]]: an effective intervention distributed inversely to need. Cosmetic use by
people at or near typical weight raises the treatment–enhancement boundary that
[[bioethics-of-enhancement|enhancement ethics]] has argued over for decades, with the added feature
that supply diverted to it is supply unavailable to someone with diabetes. The drugs also anchor a
consumer metabolic-monitoring market, [[continuous-glucose-monitoring]] and
[[wearable-health-sensors]] among it, selling the same premise of managed metabolism, while
displacing a weight-loss [[dietary-supplements|supplement]] trade built on far weaker evidence.

## Outlook

The pipeline is moving toward oral small molecules that could be manufactured and distributed like
ordinary tablets, toward multi-agonists combining GLP-1 with GIP, glucagon, or amylin receptor
activity, and toward combinations designed to spare muscle. If oral agents at generic prices arrive,
the access argument changes shape and the question becomes what happens when a large fraction of a
population takes an appetite-modifying drug indefinitely, a natural experiment with no precedent and
no control group.

What remains unanswered is whether any of this shows up where it would count. Population life
expectancy responds to metabolic disease with a lag of decades and is confounded by everything else
in those decades, so the claim that incretin drugs bought a country years of life will be hard to
establish and correspondingly easy to assert. The stronger and more testable claim is narrower: that
they compress the years of illness at the end of a life without moving its end, which is
[[compression-of-morbidity]] rather than any change in [[maximum-human-lifespan]]. That is a large
achievement, and it is not the one the popular framing has settled on.

## See also

- [[geroscience-hypothesis]]
- [[metformin]]
- [[caloric-restriction]]
- [[exercise-and-aging]]
- [[healthspan]]
- [[compression-of-morbidity]]
- [[access-and-inequality]]
- [[continuous-glucose-monitoring]]

## References

[^drucker2018]: `paper` Drucker, D.J. "Mechanisms of Action and Therapeutic Application of Glucagon-like Peptide-1." *Cell Metabolism*, 2018.
[^eng1992]: `paper` Eng, J. et al. "Isolation and characterization of exendin-4, an exendin-3 analogue, from Heloderma suspectum venom." *Journal of Biological Chemistry*, 1992. {The peptide came from Gila monster venom, screened because reptile venoms are rich in protease-resistant hormone analogues.}
[^wilding2021]: `paper` Wilding, J.P.H. et al. "Once-Weekly Semaglutide in Adults with Overweight or Obesity." *New England Journal of Medicine*, 2021.
[^jastreboff2022]: `paper` Jastreboff, A.M. et al. "Tirzepatide Once Weekly for the Treatment of Obesity." *New England Journal of Medicine*, 2022. {Tirzepatide is a dual GIP and GLP-1 receptor agonist, so its results do not transfer to the pure GLP-1 agonists.}
[^lincoff2023]: `paper` Lincoff, A.M. et al. "Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes." *New England Journal of Medicine*, 2023. {Participants had established cardiovascular disease, so the trial says nothing about primary prevention in otherwise healthy people with obesity.}
[^wilding2022]: `paper` Wilding, J.P.H. et al. "Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension." *Diabetes, Obesity and Metabolism*, 2022.
[^perkovic2024]: `paper` Perkovic, V. et al. "Effects of Semaglutide on Chronic Kidney Disease in Patients with Type 2 Diabetes." *New England Journal of Medicine*, 2024.
[^novo2025]: `statement` Novo Nordisk. Headline results announcement for the evoke and evoke+ trials of oral semaglutide in early Alzheimer's disease, 2025. {A sponsor announcement of a negative result, reported before the full trial publication.}

==============================================================================
ARTICLE: governance-of-genome-editing
TITLE:   Governance of human genome editing
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/governance-of-genome-editing
SOURCE:  https://futurehumanwiki.com/raw/governance-of-genome-editing
==============================================================================

---
title: "Governance of human genome editing"
slug: "governance-of-genome-editing"
type: "concept"
status: "emerging"
horizon: "present"
categories: ["society", "genetics"]
tags: ["governance", "regulation", "bioethics", "crispr", "law", "who"]
summary: "The patchwork of national law, professional norms, funding rules, and international soft law that determines what may be done to a human genome and by whom."
updated: "2026-07-27"
issues: ["The claim that three edited children exist is stated without a source.", "Penalties in the compare block, including the seven-year figure for China, are uncited."]
---

```infobox
{
  "caption": "Regulatory and policy domain",
  "rows": [
    { "label": "Scope", "value": "Somatic and heritable editing" },
    { "label": "Binding instruments", "value": "National law only" },
    { "label": "Main treaty", "value": "Oviedo Convention, 1997" },
    { "label": "International body", "value": "World Health Organization" },
    { "label": "Registry", "value": "WHO, established 2019" },
    { "label": "Heritable use", "value": "Prohibited in most countries" },
    { "label": "Enforcement", "value": "Domestic courts and licensing" }
  ]
}
```

**Governance of human genome editing** refers to the assembled rules — statutes, licensing regimes, funding conditions, professional statements, and international declarations — that determine which genetic interventions in humans may be attempted, by whom, and under what oversight. There is no global authority and no binding international instrument of general application. What exists is a patchwork whose strongest layer is domestic criminal and medical law, whose most visible layer is professional self-regulation, and whose weakest point is anyone unwilling to be bound by either.

## The layers

Four distinct mechanisms operate, and they are frequently confused with one another.

**Product regulation** governs [[somatic-gene-therapy]] exactly as it governs any other medicine. In the United States a gene therapy is a biological product requiring a licence; in the European Union it is an advanced therapy medicinal product. Trials require authorisation, manufacturing requires inspection, and approved products such as [[casgevy]] carry long-term follow-up obligations, in some cases running to fifteen years. This layer is mature and works about as well as drug regulation generally does.

**Funding conditions** shape what research happens without prohibiting anything. The US Dickey-Wicker Amendment, in force since 1996, bars federal money from research in which human embryos are created for research purposes or destroyed. Since embryo editing research destroys embryos, it proceeds in the United States only with private money — which means the work is unfunded rather than unlawful, and correspondingly less visible.

**Prohibition** is where heritable modification sits. Most countries that address the question at all forbid transferring an edited embryo to a womb, though the instruments vary from criminal statutes to clauses in assisted-reproduction acts drafted before editing was possible.

**Professional norms** — summit statements, academy reports, journal policies, institutional review boards — do most of the day-to-day work and have no legal force. Their power comes from controlling access to publication, funding, and employment, which is considerable inside the academy and nil outside it. Legal scholars have noted that the resulting body of policy was mostly assembled before genome editing existed, out of instruments written for cloning, embryo research, and assisted reproduction, and fits the technology only approximately.[^isasi2016]

## International instruments

The Council of Europe's Convention on Human Rights and Biomedicine, opened for signature at Oviedo in 1997, is the only binding treaty that directly addresses the question. Article 13 permits interventions on the human genome only for preventive, diagnostic, or therapeutic purposes, and only where the aim is not to introduce a modification in the genome of any descendants. It binds the states that have ratified it — around thirty — and several major research nations, including the United Kingdom and Germany, are not among them.

UNESCO's Universal Declaration on the Human Genome and Human Rights, also from 1997, names germline intervention among practices that could be contrary to human dignity. It is a declaration, not a treaty, and creates no obligation.

The three international summits on human genome editing, held in 2015, 2018, and 2023, produced statements rather than rules. The 2020 report of an international commission convened by the US National Academies and the UK Royal Society went further, setting out preconditions and a narrow translational pathway for any future heritable use, on the explicit basis that no such use was then justified.[^ihgge2020] Its criteria are now the reference point in most technical discussions of [[germline-editing]].

## National law

```compare
{
  "columns": ["United States", "United Kingdom", "China"],
  "rows": [
    { "label": "Embryo research", "values": ["Lawful, no federal funding", "Lawful under HFEA licence, 14-day limit", "Lawful under guidelines, 14-day limit"] },
    { "label": "Heritable clinical use", "values": ["FDA barred from reviewing applications", "Prohibited by statute", "Criminal offence since 2021"] },
    { "label": "Instrument", "values": ["Annual appropriations rider", "Human Fertilisation and Embryology Act", "Criminal Law Amendment (XI)"] },
    { "label": "Penalty for the act itself", "values": ["No specific offence", "Criminal offence under the Act", "Up to seven years in serious cases"] },
    { "label": "Exception created", "values": ["None", "Mitochondrial donation, 2015", "None"] }
  ]
}
```

The American arrangement is unusually indirect. There is no federal statute banning heritable editing. Instead a rider attached to appropriations legislation each year since 2016 forbids the Food and Drug Administration from acknowledging an application involving a heritable modification of an embryo. Since clinical use without FDA authorisation would be unlawful marketing of an unapproved biologic, the effect is a prohibition — but one that expires annually and could lapse through inattention rather than decision.

The United Kingdom regulates through licensing. Embryo research requires a licence from the Human Fertilisation and Embryology Authority, transfer of a genetically altered embryo to a woman is unlawful, and the one exception was created by an act of Parliament following public consultation: the 2015 regulations permitting [[mitochondrial-replacement-therapy]], under which the first births were reported a decade later. Whatever one thinks of the outcome, it demonstrates that a democratic society can decide such questions deliberately rather than by default.

China's regime was reconstructed after 2018. Ministry guidelines that carried no penalty were replaced by a specific criminal offence covering implantation of gene-edited or cloned embryos, provisions in the Civil Code on research involving human genes and embryos, and tightened ethical review procedures for human-subjects research. The [[he-jiankui-affair]] was prosecuted under a general statute on illegal medical practice because nothing more specific existed at the time.

Beyond these, a survey of national policies published in 2020 found that a large majority of the countries examined prohibited heritable modification in some form, though many did so through ambiguous instruments never tested in court, and many others have no policy at all.[^baylis2020]

## The WHO framework and registry

The World Health Organization convened an expert advisory committee after 2018 and published its output in 2021: a governance framework, a set of recommendations, and a position paper.[^who2021] The recommendations are institutional rather than substantive. They call for a registry of human genome-editing research, mechanisms for reporting illegitimate work including whistleblower protection, international collaboration on oversight of cross-border activity, attention to intellectual property as a barrier to access, and education of scientists in governance. The committee stated plainly that no country should permit clinical heritable editing at present.

> [!key] What the WHO framework is for
> It does not regulate anything. Its purpose is to give national regulators a common vocabulary and a shared set of expectations, so that a country writing its first genome-editing law does not have to invent one from nothing. Judged as a coordination device it is useful; judged as a constraint it is not one.

The registry has the same character. Entries are voluntary, and the actors most likely to matter are those least likely to register.

## Enforcement reality

Every consequence that has actually befallen anyone for editing a human genome came from a domestic court applying an ordinary law. There is no international inspectorate, no reporting obligation between states, and no mechanism by which one country can act on conduct in another.

The structural problem is that the required capability is widely distributed. Assisted reproduction is a global industry of thousands of clinics operating under highly variable regulation; editing reagents are inexpensive and shipped commercially; the expertise is taught in graduate courses. Jurisdictional arbitrage is already routine in reproductive medicine, where patients travel for procedures unavailable at home, and the same route is available for anything a clinic is willing to attempt.

Professional self-regulation, the model inherited from the [[asilomar-conference]], assumes that the relevant actors care about standing among peers. That assumption held for decades because the only people who could do the work were academics. Privately funded ventures announced in 2025 that they intended to pursue heritable editing outside the university system, which converts the question from one about norms into one about jurisdiction: which regulator, in which country, has authority over a company that never applies for a licence and never files for approval. Related enforcement gaps recur wherever the technology escapes institutional settings — in [[biohacking]] and self-experimentation, in [[gene-doping]], and in the environmental release questions raised by [[gene-drive|gene drives]] under the Convention on Biological Diversity.

## Open problems

**The line between therapy and enhancement is not codified anywhere useful.** Most statutes speak of "serious disease" without defining it, leaving the boundary to be drawn case by case by regulators — the substantive difficulty examined in [[bioethics-of-enhancement]].

**Selection is barely governed at all.** [[polygenic-embryo-screening]] is commercially available in some jurisdictions with essentially no regulation, achieves some of what people imagine editing would achieve, and involves no genome modification, so it falls outside every instrument described above. If [[in-vitro-gametogenesis]] makes large numbers of embryos available, this gap becomes the important one.

**Access is a governance failure in its own right.** The 2023 summit devoted much of its attention to the fact that approved somatic therapies reach almost none of the people who need them, arguing that a governance regime concerned only with restraint, and not with delivery, has answered half the question. That argument connects directly to [[access-and-inequality]], and to the pricing dynamics that have made several licensed cures commercially unviable.

**Nobody is responsible for edited people.** No framework assigns the duty to monitor the health of a person born from an edited embryo, or to pay for that monitoring, or to reconcile it with their privacy. Three such people exist. The number of institutions with an obligation to them is zero.

## See also

- [[germline-editing]]
- [[he-jiankui-affair]]
- [[asilomar-conference]]
- [[somatic-gene-therapy]]
- [[access-and-inequality]]
- [[polygenic-embryo-screening]]
- [[dual-use-research]]
- [[precautionary-principle]]

## References

[^ihgge2020]: `report` International Commission on the Clinical Use of Human Germline Genome Editing. *Heritable Human Genome Editing*. US National Academy of Medicine, National Academy of Sciences, and the Royal Society, 2020. {Convened by academies rather than by a regulator, so its preconditions bind nobody and carry no licensing force.}
[^who2021]: `report` World Health Organization. *Human Genome Editing: A Framework for Governance* and *Human Genome Editing: Recommendations*. WHO, 2021.
[^baylis2020]: `paper` Baylis, F., Darnovsky, M., Hasson, K., Krahn, T.M. "Human Germline and Heritable Genome Editing: The Global Policy Landscape." *The CRISPR Journal*, 2020.
[^isasi2016]: `paper` Isasi, R., Kleiderman, E., Knoppers, B.M. "Editing policy to fit the genome?" *Science*, 2016. {A short policy-forum piece arguing that the instruments predate the technology, rather than a systematic legal survey.}

==============================================================================
ARTICLE: grey-goo
TITLE:   Grey goo
PORTAL:  Nanomedicine
URL:     https://futurehumanwiki.com/wiki/grey-goo
SOURCE:  https://futurehumanwiki.com/raw/grey-goo
==============================================================================

---
title: "Grey goo"
slug: "grey-goo"
type: "risk"
status: "speculative"
horizon: "indefinite"
categories: ["nanomedicine", "society"]
tags: ["nanotechnology", "existential risk", "self-replication", "governance", "drexler", "biosafety"]
summary: "The scenario in which self-replicating nanomachines consume the biosphere, an idea introduced by Eric Drexler in 1986 and later disowned by him."
updated: "2026-07-28"
humanEvidence: "Nothing to observe in people: no molecular assembler or self-replicating nanomachine has been built, and the nearest documented harm from engineered nanomaterials is fibre toxicity reported in mice."
issues: ["Smalley's objections to mechanosynthesis are discussed twice with no citation.", "The closing reference to 1992 is unexplained; name Drexler's Nanosystems or drop the date."]
---

```infobox
{
  "caption": "Hypothetical technological risk",
  "rows": [
    { "label": "Term coined", "value": "1986, Engines of Creation" },
    { "label": "Coined by", "value": "Eric Drexler", "link": "/wiki/eric-drexler" },
    { "label": "Mechanism", "value": "Uncontrolled molecular self-replication" },
    { "label": "Prerequisite", "value": "Working molecular assemblers" },
    { "label": "Assessed by", "value": "Royal Society, 2004" },
    { "label": "Mainstream view", "value": "Not a foreseeable risk" },
    { "label": "Author's position", "value": "Retracted the emphasis" }
  ]
}
```

**Grey goo** is a hypothetical catastrophe in which self-replicating nanoscale machines escape control, consume available matter to make copies of themselves, and destroy the biosphere in the process. The scenario appeared in [[eric-drexler]]'s *Engines of Creation* in 1986 as a warning attached to his proposal for [[molecular-assembler|molecular assemblers]]. It became the dominant public image of nanotechnology risk for two decades, shaped early regulatory debate, and was subsequently disowned by the person who named it.

## Overview

Drexler's argument was structural. If assemblers can build atomically precise machinery, and if the fastest route to macroscopic quantities of product is machines that build copies of themselves, then the technology's core capability is self-replication — and self-replication with a mutation or a design error is a runaway process. He devoted a chapter of *Engines of Creation* to the hazard, alongside the misuse of the same technology by states.[^drexler1986]

The scenario is a genuine argument rather than a throwaway image, and it has an obvious biological analogue. What separates it from an ordinary invasive-species problem is the claimed generality of the eating: a machine that can disassemble arbitrary matter into feedstock is not restricted to a niche.

The link to medicine is direct. The therapeutic devices catalogued in [[medical-nanorobots]], from the [[respirocytes|respirocyte]] onward, were to be manufactured by the same machinery; the hazard and the benefit share a prerequisite, which is why the two literatures cannot be separated.

## The mechanism

The scenario requires four things in sequence. A device must be able to build a functional copy of itself from materials it finds in the environment. It must have access to those materials in a usable chemical form. It must obtain energy. And it must survive outside a controlled setting long enough to complete each cycle.

Exponential growth does the rest, in principle. A replicator with an hour-long doubling time reaches macroscopic scale within days and planetary scale within weeks, on the assumption that nothing limits it.

Robert Freitas examined those limits in a detailed 2000 analysis of what he called global ecophagy. His conclusion was that waste heat, not materials, is the binding constraint: converting biomass at the rates the scenario requires releases energy that would raise local temperature far beyond what any machinery could tolerate, and would produce a thermal signature detectable well before the process approached completion. He treated this as an argument for monitoring rather than complacency, and proposed design restrictions on any replicating system.[^freitas2000]

> [!caution] Why the prerequisite matters
> Grey goo cannot happen without molecular assemblers, and no assembler has been built or shown to be buildable. Every assessment of the scenario is therefore an assessment conditional on a technology whose feasibility is itself disputed. This makes the risk unusual: the debate about its plausibility is downstream of a debate about chemistry that has never been settled experimentally.

## Plausibility

Mainstream scientific opinion treats the scenario as very improbable, for several reasons that operate independently.

The prerequisite is missing. Richard Smalley's objections to mechanosynthesis, if correct, rule out the assembler and therefore the goo. Proponents dispute the objections, but no one has demonstrated the capability.

Free-ranging replication is not required by the goal. Molecular manufacturing could be organised as fixed factory machinery that produces products, with replication confined to a controlled production line — the arrangement Drexler and Chris Phoenix argued for in 2004, on the grounds that it is both safer and easier to engineer than autonomous replicators.

Biology sets a hard precedent. Life has been optimising self-replication in exactly this environment for billions of years, under selection pressure no design process can match, and has not produced a universal disassembler. Real organisms are specialists constrained by feedstock chemistry, temperature, pH, and predation. A synthetic replicator would face the same constraints while lacking the accumulated adaptations.

Engineering a machine that both operates at atomic precision and tolerates uncontrolled environments is close to a contradiction. Precision machinery requires clean conditions; the outdoors supplies dust, water, salt, ultraviolet light, and organisms that eat things.

The technologies that actually produce nanoscale machinery reinforce the point. Structures built by [[dna-nanotechnology]] copy themselves only with enzymes supplied by an experimenter and degrade within hours in serum. The untethered devices surveyed in [[microrobots-in-medicine]] draw their energy from external fields and stop when the field is switched off. The [[xenobots]] built from frog cells come closest, having been reported to gather dissociated cells in a dish into piles that mature into further xenobots — replication only for as long as an experimenter keeps supplying the cells, and consuming nothing that was not already living tissue. None of these is a step toward an autonomous replicator, and none was designed to be.

## How the scenario entered policy

The idea spread far beyond the technical community. Bill Joy's 2000 essay in *Wired* placed self-replicating nanotechnology alongside genetics and robotics as technologies that might make humans obsolete, and argued for relinquishing some lines of research.[^joy2000] Michael Crichton's novel *Prey* (2002) dramatised a swarm scenario. Public figures in the United Kingdom raised the question, the ETC Group called for a moratorium on nanomaterials, and by 2003 nanotechnology policy discussions were conducted partly in the vocabulary of runaway machines.

The Royal Society and Royal Academy of Engineering examined the field in 2004 at the British government's request. Their report concluded that self-replicating nanomachines were not a foreseeable prospect and that regulatory attention belonged on the actual hazard: free engineered nanoparticles, whose toxicology and environmental behaviour were poorly characterised.[^royalsociety2004] That judgement has held.

Leaders of the US National Nanotechnology Initiative, Smalley among them, publicly rejected the scenario, in part because a federal materials-science programme did not want to be associated with science fiction. The effect was a peculiar alignment: the field's critics and its funders both had reasons to discuss molecular manufacturing as fantasy, which foreclosed the experimental work that might have settled the underlying question.

## Drexler's retraction

Drexler came to regard the emphasis as a mistake. He has said he regrets introducing the image, and his later writing reframes the field as atomically precise manufacturing carried out by fixed, factory-style systems that never replicate autonomously.[^drexler2013] The 2004 paper with Phoenix made the technical version of the argument: replication is a design choice, not a requirement, and a manufacturing system built without it is not subject to the failure mode.

The retraction is partial in an important way. Drexler continues to argue that atomically precise manufacturing would be strategically destabilising — cheap, rapid production of arbitrary hardware, including weapons — and that this is the serious risk. That claim is much harder to dismiss than grey goo and receives far less attention.

## What displaced it

The self-replication risk did not disappear; it moved to biology, where the replicators already exist and are programmable. Engineered pathogens are the leading biological candidate for catastrophic misuse, as discussed in [[dual-use-research]] and in the taxonomies of [[existential-risk]] that [[nick-bostrom]] set out in 2002, which listed nanotechnological accident and engineered disease side by side among the technological hazards. The ability to write [[synthetic-genomes]] to order, unavailable when *Engines of Creation* appeared, is the specific capability that shifted the balance. Self-propagating genetic elements raise a real version of the containment problem in [[gene-drive]], where the technology works and the reversibility question is live. [[mirror-life]] represents the closest existing analogue to grey goo taken seriously by working scientists: a chirally inverted organism that existing immune systems and enzymes could not recognise, which prompted a public call in 2024 from a large group of researchers to halt work toward it.

The nanomaterial risks the Royal Society flagged also proved real. Long, thin carbon nanotubes introduced into mice produced inflammatory responses resembling those caused by asbestos fibres, an early signal that particle geometry matters for toxicity in ways bulk-material safety data do not capture.[^poland2008]

## Governance

There is no international instrument specific to self-replicating nanomachines, and no obvious need for one while the enabling technology does not exist. What exists instead is a set of proposals — design rules requiring dependence on artificial feedstocks unavailable in nature, mandatory encryption of replication instructions, geographic and physical containment — most of which were developed by proponents of molecular manufacturing rather than by regulators.

The more useful legacy of the debate is methodological. Grey goo is the standard case study in how a vivid low-probability scenario can crowd out attention to the mundane risks of the same technology, a pattern relevant to how the [[precautionary-principle]] is applied and to arguments for [[differential-technological-development]]. It also illustrates a failure mode particular to speculative fields: a warning issued to encourage careful development instead became the reason the development was never seriously attempted, leaving the underlying question about atomic-scale manufacturing exactly where it was in 1992.

## See also

- [[molecular-assembler]]
- [[eric-drexler]]
- [[medical-nanorobots]]
- [[mirror-life]]
- [[existential-risk]]
- [[dual-use-research]]
- [[precautionary-principle]]
- [[dna-nanotechnology]]

## References

[^drexler1986]: `book` Drexler, K.E. *Engines of Creation: The Coming Era of Nanotechnology.* Anchor Books, 1986.
[^freitas2000]: `report` Freitas, R.A. "Some Limits to Global Ecophagy by Biovorous Nanoreplicators, with Public Policy Recommendations." Institute for Molecular Manufacturing, 2000. {Written from inside the molecular manufacturing community; it assumes assemblers are feasible and asks only what would limit them.}
[^joy2000]: `news` Joy, B. "Why the Future Doesn't Need Us." *Wired*, April 2000. {An essay arguing for relinquishment of whole lines of research, not reporting or technical assessment.}
[^royalsociety2004]: `report` The Royal Society and the Royal Academy of Engineering. *Nanoscience and Nanotechnologies: Opportunities and Uncertainties.* 2004.
[^drexler2013]: `book` Drexler, K.E. *Radical Abundance: How a Revolution in Nanotechnology Will Change Civilization.* PublicAffairs, 2013.
[^poland2008]: `paper` Poland, C.A. et al. "Carbon nanotubes introduced into the abdominal cavity of mice show asbestos-like pathogenicity in a pilot study." *Nature Nanotechnology*, 2008. {An injection model chosen to test mesothelial response to fibre shape; it does not establish that inhaled nanotubes reach the same tissue.}

==============================================================================
ARTICLE: hallmarks-of-aging
TITLE:   Hallmarks of aging
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/hallmarks-of-aging
SOURCE:  https://futurehumanwiki.com/raw/hallmarks-of-aging
==============================================================================

---
title: "Hallmarks of aging"
slug: "hallmarks-of-aging"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity", "foundations"]
tags: ["aging", "biogerontology", "mechanisms", "damage", "geroscience"]
summary: "The consensus taxonomy of cellular and molecular processes held to drive mammalian aging, set out as nine mechanisms in 2013 and expanded to twelve in 2023."
updated: "2026-07-27"
humanEvidence: "No hallmark has been shown to satisfy the framework's third criterion in humans; the amelioration results come from worms, flies and mice, and the human data are observational and correlational."
issues: ["The mortality-doubling-time figure in the keyfacts block carries no source.", "The Outlook refers to prizes and consortia defining outcome measures without naming any."]
---

```infobox
{
  "caption": "Framework in biogerontology",
  "rows": [
    { "label": "Proposed", "value": "2013" },
    { "label": "Revised", "value": "2023" },
    { "label": "Authors", "value": "López-Otín, Blasco, Partridge, Serrano, Kroemer" },
    { "label": "Count", "value": "9, later 12" },
    { "label": "Tiers", "value": "Primary, antagonistic, integrative" },
    { "label": "Model precedent", "value": "Hallmarks of Cancer" },
    { "label": "Field", "value": "Biogerontology" }
  ]
}
```

**Hallmarks of aging** are the cellular and molecular processes that, in the dominant framework of contemporary biogerontology, jointly account for the progressive functional decline of mammals over time. The list was set out as nine hallmarks in a 2013 review in *Cell* and expanded to twelve by the same five authors in 2023.[^lopez2013][^lopez2023] Each candidate is supposed to satisfy three tests: it manifests during normal aging, experimentally worsening it accelerates aging, and experimentally easing it slows aging and extends healthy life.

```keyfacts
[
  { "value": "9 → 12", "label": "Hallmarks listed", "note": "2013 review; 2023 revision added three" },
  { "value": "3", "label": "Criteria for inclusion", "note": "manifestation, aggravation, amelioration" },
  { "value": "~8 yr", "label": "Mortality doubling time in adults", "note": "the Gompertz regularity the framework tries to explain" }
]
```

## Overview

Human mortality rises close to exponentially with adult age; the annual probability of death roughly doubles every eight years, a regularity described by Benjamin Gompertz in 1825. Beneath that curve sit thousands of measurable changes — shorter telomeres, stiffer arteries, fewer functional stem cells, higher circulating interleukin-6, drifting DNA methylation. The hallmarks framework reduces that inventory to a set of processes that are general across tissues, conserved across mammals, and at least arguably actionable.

The framework does two jobs. It supplies a common vocabulary for a field that had spent decades arguing between rival single-cause theories. It also supplies a target list: most academic programmes and companies in the [[geroscience-hypothesis|geroscience]] sector map their work onto one or more hallmarks, and reviewers treat hallmark membership as prima facie evidence that a mechanism matters. That adoption is itself a fact about the field. The hallmarks are less a discovery than a coordination device that worked.

> [!note] Terminology
> "Hallmark" is borrowed from oncology, where Hanahan and Weinberg used it for capabilities a cell must acquire to become malignant.[^hanahan2000] The analogy is imperfect: cancer hallmarks describe what a tumour must do, while aging hallmarks describe what happens to an organism whether or not anything is required of it.

## Origins

Before 2013 the biology of aging was a collection of competing theories, each with a constituency: the free-radical theory, somatic mutation accumulation, error catastrophe, the disposable soma, antagonistic pleiotropy, and telomere shortening. Most were framed as explanations of aging in the singular. Experiments that undercut one theory left the others untouched, and the literature fragmented.

```timeline
[
  { "year": "2000", "title": "Hallmarks of Cancer published", "text": "Hanahan and Weinberg condense oncology into six acquired capabilities, demonstrating that a well-chosen list can organize a field." },
  { "year": "2013", "title": "Nine hallmarks proposed", "text": "López-Otín, Blasco, Partridge, Serrano and Kroemer publish 'The Hallmarks of Aging' in Cell, grouping nine processes into primary, antagonistic and integrative tiers." },
  { "year": "2021", "title": "Paradigm critique", "text": "Gems and de Magalhães argue the framework repackages damage-maintenance thinking and does not distinguish cause from consequence." },
  { "year": "2022", "title": "Candidate hallmarks multiply", "text": "A Copenhagen meeting summary proposes further candidates including compromised autophagy, altered mechanical properties and splicing dysregulation." },
  { "year": "2023", "title": "Expanded to twelve", "text": "The original authors add disabled macroautophagy, chronic inflammation and dysbiosis, and emphasize the interconnection of the hallmarks." }
]
```

## What counts as a hallmark

The 2013 paper set three criteria. A hallmark should manifest during normal aging; its experimental aggravation should accelerate aging; and its experimental amelioration should retard aging and extend healthy lifespan. The first criterion is easy — a great many things change with age. The second is harder but tractable, since almost any biological system can be broken in a way that produces a short-lived, sickly animal. The third carries nearly all the weight, and it is where the evidence thins.

Nearly every amelioration result comes from short-lived laboratory organisms: *Caenorhabditis elegans*, *Drosophila*, and inbred or genetically heterogeneous mice. No hallmark has been shown to satisfy the third criterion in humans, for the simple reason that no intervention has been demonstrated to slow human aging. Establishing that would require either a multi-decade mortality trial or a surrogate endpoint that regulators accept, which is the problem the field of [[aging-biomarkers|aging biomarkers]] exists to solve.

A second difficulty is that accelerating aging and causing disease look alike from outside. Progeroid syndromes — Werner syndrome from *WRN* mutations, Hutchinson–Gilford progeria from the *LMNA* variant that produces progerin — recapitulate some features of aging and omit others. Patients with progeria develop severe atherosclerosis but not the cancer or dementia risk of ordinary old age. Whether such syndromes are accelerated aging or a distinct pathology that mimics it remains disputed.

## The twelve hallmarks

The 2013 paper sorted the hallmarks into three tiers by causal position, and the 2023 revision kept the scheme.

### Primary hallmarks: damage

These are unambiguously deleterious processes that accumulate. **Genomic instability** covers nuclear point mutations, structural variants, and somatic mosaicism, together with damage to mitochondrial DNA. **Telomere attrition** is treated separately because the [[telomeres-and-telomerase|end-replication problem]] gives it a distinct mechanism and a distinct relationship to cancer. **Epigenetic alterations** include DNA methylation drift, histone modification changes, and loss of heterochromatin, the changes that the [[epigenetic-clock|epigenetic clocks]] built by [[steve-horvath]] and others measure. **Loss of proteostasis** is the failure of the chaperone, proteasome and lysosomal machinery that keeps the [[proteostasis|proteome]] folded and turned over. **Disabled macroautophagy**, added in 2023, separates the [[autophagy|autophagic]] arm of that system into its own hallmark on the grounds that it declines with age across tissues and that genetically raising it extends lifespan in several model organisms — an epistatic and gain-of-function argument in animals, not a demonstration that autophagy can be restored in an aged human.

### Antagonistic hallmarks: responses that turn harmful

These begin as protective and become damaging when chronic or excessive. **Deregulated nutrient sensing** covers the insulin/IGF-1 axis that [[cynthia-kenyon|Cynthia Kenyon's]] worm genetics opened up, along with mTOR, AMPK and the sirtuins — signalling that reports abundance and, when persistently activated, suppresses maintenance. It is the node that [[caloric-restriction|dietary restriction]], [[rapamycin]] and [[nad-precursors|NAD⁺ precursors]] all address from different directions. **[[mitochondrial-dysfunction|Mitochondrial dysfunction]]** describes falling respiratory capacity, altered dynamics and impaired quality control. **[[cellular-senescence|Cellular senescence]]** is a stable proliferative arrest that suppresses tumours in youth and, through the secretory phenotype of accumulated senescent cells, degrades tissue in age.

### Integrative hallmarks: the phenotype

These are the level at which the organism visibly ages. **[[stem-cell-exhaustion|Stem cell exhaustion]]** is the decline of regenerative capacity in blood, muscle, gut and skin. **Altered intercellular communication** covers endocrine, neuronal and immune signalling changes, including the systemic factors studied in parabiosis. **Chronic inflammation**, added in 2023, formalizes [[inflammaging]] as its own hallmark rather than a subtype of communication failure. **Dysbiosis**, also added in 2023, is the least settled of the twelve; the human microbiome changes with age, but whether those changes cause anything is largely untested outside germ-free and gnotobiotic mice.

## Evidence and use

The strongest support for the framework is indirect: interventions aimed at individual hallmarks extend lifespan in mice, and several do so in genetically heterogeneous mice tested in parallel at multiple independent sites through the National Institute on Aging's Interventions Testing Program. That programme was designed to filter out the single-laboratory, single-strain results that had littered the earlier literature, and its positive findings — rapamycin most robustly, along with acarbose, 17-α-estradiol and canagliflozin, the latter two with effects reported only in males — act mostly on nutrient sensing and metabolic signalling. [[senolytics|Senolytic]] clearance of senescent cells improves physical function in aged mice, and transient expression of reprogramming factors reverses several molecular markers of age in mouse tissues, the approach pursued at scale by [[epigenetic-reprogramming|reprogramming]] companies.

None of this has yet produced a human result showing slowed aging rather than improved disease-specific endpoints. The human evidence is almost entirely observational and correlational: methylation-based age estimates predict mortality after adjustment for chronological age, senescent-cell markers rise in aged tissue, and inflammatory markers track frailty. Correlations of that kind are consistent with the hallmarks being causes and equally consistent with their being downstream readouts of something else. The framework is also used prescriptively in a way its authors did not license, as a checklist against which consumer testing panels and [[dietary-supplements|supplement formulations]] are marketed.

> [!caution] Contested
> Hallmark membership is not a causal weighting. The framework says which processes belong on the list; it does not say which contribute most to mortality, in what order they act, or which would be worth intervening on first. Several authors argue this is the framework's central omission rather than a detail to be filled in later.

## Criticism

The most sustained critique is that the hallmarks are a taxonomy presented as a theory. Gems and de Magalhães argue that the framework restates the damage-and-maintenance view of aging in new packaging while remaining agnostic on whether each hallmark is a cause, a consequence, or an adaptive response, and that this agnosticism makes it hard to falsify.[^gems2021] Because the hallmarks are heavily interdependent — senescent cells drive inflammation, inflammation impairs stem cell function, impaired stem cell function increases senescent burden — assigning causal priority from the list alone is not possible.

A related objection concerns list inflation. The inclusion criteria are permissive enough that many age-associated processes qualify, and candidate hallmarks have accumulated in the literature faster than the canonical list has been revised.[^schmauck2022] If the list grows whenever a new mechanism is characterized, it functions as an index of active research areas rather than a constraint on theory.

Alternative frameworks reject the damage-centric premise. Blagosklonny's hyperfunction theory holds that aging is not damage accumulation but the continued, unregulated running of developmental growth programmes after they are needed, with mTOR as the principal driver.[^blago2006] On that account senescence and hypertrophy are the disease, and damage is downstream. A complex-systems view argues that aging emerges from the loss of homeostatic regulation across networks and is not decomposable into discrete mechanisms at all.[^cohen2022] [[sens-research-foundation|SENS]], the damage-repair programme associated with [[aubrey-de-grey]], predates the hallmarks and organizes aging into seven categories of accumulated damage on the argument that repair is more tractable than metabolic modulation.

Finally, the framework is mouse-centric in ways that may not transfer. Laboratory mice have long telomeres and constitutive telomerase in many tissues, so telomere attrition means something different in mice than in humans. Species that violate the expected pattern — the [[negligible-senescence|negligibly senescent]] animals such as naked mole-rats and bowhead whales — are not well explained by any tiering of the twelve.

## Outlook

The open problem is converting a list into a ranking. Doing that requires measurements in humans that track the processes rather than their correlates, and interventions selective enough to move one hallmark without moving the others. Both are hard: most candidate drugs hit several hallmarks at once, and most candidate biomarkers are correlational. Prizes and consortia have begun to define outcome measures in terms of functional restoration across muscle, cognition and immunity rather than a molecular readout, precisely because the molecular readouts remain unvalidated as surrogates for [[healthspan]] or [[maximum-human-lifespan|lifespan]]. Until an intervention aimed at a named hallmark changes a hard human endpoint, the framework's central claim — that these processes are the causes of aging rather than its accompaniments — remains an organizing assumption rather than a demonstrated result. Arguments about [[biological-age]] reversal and [[longevity-escape-velocity]] are built on that assumption, and inherit its weakness.

## See also

- [[geroscience-hypothesis]]
- [[cellular-senescence]]
- [[aging-biomarkers]]
- [[epigenetic-clock]]
- [[senolytics]]
- [[autophagy]]
- [[caloric-restriction]]
- [[negligible-senescence]]
- [[biological-age]]

## References

[^lopez2013]: `paper` López-Otín, C., Blasco, M. A., Partridge, L., Serrano, M., Kroemer, G. "The Hallmarks of Aging." *Cell*, 2013.
[^lopez2023]: `paper` López-Otín, C., Blasco, M. A., Partridge, L., Serrano, M., Kroemer, G. "Hallmarks of aging: An expanding universe." *Cell*, 2023. {A review by the same five authors as the 2013 paper; the expansion to twelve reflects their judgement rather than any formal consensus process.}
[^hanahan2000]: `paper` Hanahan, D., Weinberg, R. A. "The Hallmarks of Cancer." *Cell*, 2000.
[^gems2021]: `paper` Gems, D., de Magalhães, J. P. "The hoverfly and the wasp: A critique of the hallmarks of ageing as a paradigm." *Ageing Research Reviews*, 2021.
[^schmauck2022]: `paper` Schmauck-Medina, T. et al. "New hallmarks of ageing: a 2022 Copenhagen ageing meeting summary." *Aging*, 2022. {A meeting summary, so the additional hallmarks it lists are proposals from the participants rather than an agreed revision of the canonical list.}
[^blago2006]: `paper` Blagosklonny, M. V. "Aging and immortality: quasi-programmed senescence and its pharmacologic inhibition." *Cell Cycle*, 2006. {A theoretical proposal argued from the existing mTOR literature; no experiment was designed to test the hyperfunction account against the damage account.}
[^cohen2022]: `paper` Cohen, A. A. et al. "A complex systems approach to aging biology." *Nature Aging*, 2022.

==============================================================================
ARTICLE: he-jiankui-affair
TITLE:   He Jiankui affair
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/he-jiankui-affair
SOURCE:  https://futurehumanwiki.com/raw/he-jiankui-affair
==============================================================================

---
title: "He Jiankui affair"
slug: "he-jiankui-affair"
type: "event"
status: "historical"
horizon: "historical"
categories: ["genetics", "society"]
tags: ["crispr", "embryos", "heritable editing", "bioethics", "governance", "china"]
summary: "The 2018 announcement that twin girls had been born from CRISPR-edited embryos in Shenzhen, and the scientific condemnation and criminal prosecution that followed."
updated: "2026-07-27"
issues: ["The birth of the third child in 2019 is stated without a source.", "The 2003 Chinese ministry guidelines are described but not cited."]
---

```infobox
{
  "caption": "Scientific and legal scandal",
  "rows": [
    { "label": "Date announced", "value": "25 November 2018" },
    { "label": "Location", "value": "Shenzhen, China" },
    { "label": "Principal", "value": "He Jiankui" },
    { "label": "Gene targeted", "value": "CCR5" },
    { "label": "Children born", "value": "Three, 2018–2019" },
    { "label": "Sentence", "value": "Three years, illegal medical practice" },
    { "label": "Released", "value": "April 2022" }
  ]
}
```

The **He Jiankui affair** was the first and so far only case in which human embryos edited with CRISPR were transferred to a womb and carried to term. On 25 November 2018 the Shenzhen-based biophysicist He Jiankui announced through online videos and an exclusive news interview that twin girls, given the pseudonyms Lulu and Nana, had been born from embryos in which he had disrupted the CCR5 gene, intending to make them resistant to HIV. The reaction from the international scientific community was immediate and close to unanimous condemnation, and He was convicted of illegal medical practice thirteen months later.

## Background

He trained in physics in China, took a doctorate at Rice University, and did postdoctoral work at Stanford in single-molecule sequencing before returning to China in 2012 as an associate professor at the Southern University of Science and Technology in Shenzhen.[^davies2020] He founded sequencing companies and was, by 2017, working privately on embryo editing while formally on unpaid leave from the university.

The scientific setting mattered. [[crispr-cas9]] had made embryo editing technically accessible; Chinese groups had published the first reports of editing non-viable human embryos in 2015, finding low efficiency and pervasive mosaicism. An international summit in Washington that December had declared clinical germline use irresponsible until safety questions were resolved and there was broad societal consensus. Those were statements, not laws. China's rules in 2018 consisted of ministry guidelines from 2003 barring implantation of edited embryos, without criminal penalty attached.

## What happened

```timeline
[
  { "year": "2015", "title": "The technique and the norm", "text": "Chinese groups publish the first CRISPR experiments in non-viable human embryos; an international summit in Washington declares clinical germline use irresponsible pending safety and societal consensus." },
  { "year": "2017", "title": "Recruitment", "text": "He recruits couples in which the father is HIV-positive through a Beijing advocacy organisation, using consent documents framed around AIDS vaccine development." },
  { "year": "25 Nov 2018", "title": "Announcement", "text": "Online videos and a news exclusive reveal that twin girls have been born from embryos edited at the CCR5 locus." },
  { "year": "28 Nov 2018", "title": "Hong Kong summit", "text": "He presents the work at the Second International Summit on Human Genome Editing and discloses a second pregnancy. The organising committee condemns the experiment." },
  { "year": "Jan 2019", "title": "Investigation", "text": "A Guangdong provincial inquiry concludes that he evaded supervision and used forged ethical review documents." },
  { "year": "Dec 2019", "title": "Conviction", "text": "A Shenzhen district court sentences He to three years' imprisonment and a fine for illegal medical practice." },
  { "year": "2021", "title": "New criminal law", "text": "China creates a specific offence covering implantation of gene-edited or cloned embryos into humans or animals." },
  { "year": "Apr 2022", "title": "Release", "text": "He is released from prison and resumes public activity, announcing further research projects." }
]
```

He recruited couples through a Beijing HIV/AIDS advocacy organisation. In each, the prospective father was HIV-positive with suppressed viral load and the mother was HIV-negative. Embryos were created by intracytoplasmic sperm injection with the editing reagents co-injected, cultured, biopsied, and transferred.

The stated rationale was protection against HIV. It does not survive inspection. Sperm washing, a routine procedure, already reduces the risk of paternal transmission to close to zero, so the children were never at meaningful risk of acquiring HIV from their fathers. The edit was aimed at a hypothetical future infection — prophylaxis at best, and closer to the enhancement category discussed under [[human-enhancement]] than to treatment of a disease. Nor was there any condition to screen for: [[embryo-selection]] had nothing to select against, which removed the one situation in which editing is sometimes argued to add capability.

He announced the births days before he was scheduled to speak at the Second International Summit on Human Genome Editing in Hong Kong. He presented there on 28 November to a packed hall, defended the work, and revealed that a second pregnancy was under way. A third child was born from that pregnancy in 2019.

## What went wrong

The technical failures were as serious as the ethical ones.

**Neither twin received the variant he was aiming at.** The natural CCR5-Δ32 deletion, common in northern European populations and known to confer resistance to the most common HIV strains, is a specific 32-base-pair loss. Cas9 repairs by error-prone end joining produce whatever indels the cell happens to make. The twins carried novel mutations at the locus with no characterised function, not the protective allele. Installing a defined sequence rather than a random disruption is precisely what [[base-editing]] and [[prime-editing]] were developed to do: base editing was barely two years old when the embryos were made, and prime editing was not described until the year after the births.

**At least one twin was mosaic.** Different cells carried different edits, so no single genotype describes her. Excerpts from He's unpublished manuscript, released with expert commentary in 2019, showed that the biopsy data he relied on did not support the claims he made from them.[^regalado2019]

**Verification was impossible in principle.** A trophectoderm biopsy samples cells destined for the placenta, and the embryo that is transferred is never itself sequenced — the structural obstacle to safe embryo editing described in [[germline-editing]]. Reported [[crispr-off-target-effects|off-target analysis]] was thin, and later work in human embryos revealed classes of damage at the target site that his methods would not have detected at all.

**The consent process misrepresented the study.** Documents described the project in the vocabulary of AIDS vaccine development and ran to more than twenty pages of technical material presented to couples with a strong incentive to accept. Ethics approval was attributed to a Shenzhen hospital that denied any involvement and said the signatures on the document were not authentic.

**CCR5 loss is not free.** People without functional CCR5 have elevated susceptibility to West Nile virus and other flavivirus infections. Trading a well-managed risk for a poorly characterised one is a bad exchange even if the edit had worked as intended.

> [!key] Why the target choice damaged the case further
> Disabling CCR5 is one of very few edits with an obvious protective rationale, which is exactly why it appealed. That the most defensible available target still produced an indefensible experiment is the affair's most useful lesson about the gap between a plausible idea and a justified intervention.

## Reaction

The summit's organising committee, chaired by David Baltimore, called the work irresponsible and said it had failed to conform to international norms.[^cyranoski2018] More than a hundred Chinese scientists signed an open letter condemning it. [[jennifer-doudna]], who had met He shortly before his presentation, described the work as shocking and used the episode to press for enforceable oversight. Journals declined the manuscript. His university dismissed him; his former doctoral adviser in the United States was investigated by his institution and subsequently left it. Legal scholars who reconstructed the sequence of decisions concluded that the failure was not of any single rule but of a system in which no rule had teeth.[^greely2019]

Dissent from the consensus was rare and narrow. [[george-church]] argued publicly that HIV prevention was a defensible target and that the pile-on was excessive, while criticising the execution. Almost no one defended the conduct of the experiment itself.

## Consequences

In December 2019 the Nanshan District People's Court in Shenzhen convicted He and two collaborators of illegal medical practice. He received three years' imprisonment and a fine of three million yuan; the others received shorter terms.[^normile2019] The prosecution rested on practising medicine without a licence and forging approval documents, not on any statute about genome editing, because none then existed.

The verdict therefore turned on paperwork and licensure rather than on what had been done to the embryos, which is why the case is a poor precedent for anything except itself. Legislators noticed the gap: within three years China had written implantation of an edited embryo into its criminal code as an offence in its own right, and the wider rebuilding of Chinese oversight that followed is described in [[governance-of-genome-editing]]. Internationally the episode hardened an argument that was already under way, between those who wanted a defined translational pathway on the grounds that a licensed route is safer than an unpoliced one, and those who read pathway-building as tacit endorsement — the same split that runs through the self-regulation model of the [[asilomar-conference]].

The children's situation remains the least examined part of the story. Their health status is not public. No institution has acknowledged responsibility for the lifelong monitoring that an edited person plainly requires, and any such monitoring conflicts with their privacy and that of their families. They are the only human beings alive with deliberately edited nuclear genomes, a status they did not choose and cannot renounce.

## Aftermath

He was released in April 2022 and returned to public activity, announcing a series of projects — gene therapy for Duchenne muscular dystrophy, then embryo-editing research aimed at installing an Alzheimer's-protective variant — and cultivating attention on social media. He has published little that has passed peer review since, and his announcements are treated with widespread scepticism. Travel plans and academic invitations have repeatedly been withdrawn under pressure.

The affair is cited most often as evidence that scientific self-governance failed. A more precise reading is that it never applied. He was outside the academic system's incentives: privately funded, uninterested in publication, indifferent to reputation among peers. The mechanisms that constrain researchers — grants, journals, tenure, institutional review — had no purchase on him, and the mechanisms that eventually did were an ordinary criminal court and a medical licensing statute. That distinction matters increasingly as privately capitalised ventures announce similar ambitions, and it complicates the assumption behind both [[precautionary-principle|precautionary]] and pathway-based approaches that the relevant actors are members of a professional community in the first place.

## See also

- [[germline-editing]]
- [[governance-of-genome-editing]]
- [[crispr-cas9]]
- [[asilomar-conference]]
- [[designer-babies]]
- [[embryo-selection]]
- [[bioethics-of-enhancement]]
- [[crispr-off-target-effects]]

## References

[^cyranoski2018]: `paper` Cyranoski, D. and Ledford, H. "Genome-edited baby claim provokes international outcry." *Nature*, 2018.
[^greely2019]: `paper` Greely, H.T. "CRISPR'd babies: human germline genome editing in the 'He Jiankui affair'." *Journal of Law and the Biosciences*, 2019.
[^regalado2019]: `news` Regalado, A. "China's CRISPR babies: Read exclusive excerpts from the unseen original research." *MIT Technology Review*, 2019. {The excerpts came from a manuscript journals declined and that has never been published, so the underlying data have never been reviewed.}
[^normile2019]: `paper` Normile, D. "Chinese scientist who produced genetically altered babies sentenced to 3 years in jail." *Science*, 2019.
[^davies2020]: `book` Davies, K. *Editing Humanity: The CRISPR Revolution and the New Era of Genome Editing*. Pegasus Books, 2020.

==============================================================================
ARTICLE: head-transplant
TITLE:   Head transplantation
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/head-transplant
SOURCE:  https://futurehumanwiki.com/raw/head-transplant
==============================================================================

---
title: "Head transplantation"
slug: "head-transplant"
type: "concept"
status: "contested"
horizon: "indefinite"
categories: ["bodies"]
tags: ["transplantation", "spinal cord", "neurosurgery", "medical ethics", "identity", "fringe science"]
summary: "The proposed transfer of a living human head onto a donor body, blocked chiefly by the inability to reconnect a severed spinal cord and rejected by mainstream neurosurgery."
updated: "2026-07-27"
humanEvidence: "No human attempt has been made; living-subject work stops at Demikhov's 1950s dogs and White's 1970 rhesus monkeys, whose cords were never joined, and the 2017 human procedure was a rehearsal on two cadavers."
issues: ["The spinal-cord-injury population figure in the closing section is unsourced.", "The 2023 whole-eye transplant at NYU Langone needs a source."]
---

```infobox
{
  "caption": "Proposed surgical procedure",
  "rows": [
    { "label": "Also called", "value": "Cephalosomatic anastomosis" },
    { "label": "First animal attempts", "value": "1950s, dogs" },
    { "label": "Primate experiment", "value": "Robert White, 1970" },
    { "label": "Modern proposal", "value": "Sergio Canavero, 2013" },
    { "label": "Human attempts", "value": "None" },
    { "label": "Core unsolved problem", "value": "Spinal cord reconnection" },
    { "label": "Mainstream assessment", "value": "Not feasible; ethically rejected" }
  ]
}
```

**Head transplantation** is the proposed surgical transfer of a living head onto a body from a different individual, so that the recipient's brain is supplied by a new circulatory and musculoskeletal system. The procedure is more accurately called a body transplant, since the person who would survive is the one attached to the head. No human attempt has been made. The obstacle that makes it implausible is not surgical technique but the fact that a completely severed spinal cord cannot be repaired, and every proposal for performing the operation depends on solving that problem first.

## Animal precedents

The experiments usually cited are older than the current debate and more limited than their reputation suggests.

In the 1950s Vladimir Demikhov, a Soviet surgeon who did foundational work on cardiac and pulmonary transplantation, created dogs bearing a second head and forelimbs grafted onto the neck of an adult host. The grafted heads were conscious and could lap fluid. The animals survived days, dying of rejection.

Robert White, a neurosurgeon at Case Western Reserve University, performed the most technically advanced version in 1970, transferring the head of one rhesus monkey onto the body of another with vascular anastomosis and hypothermic protection. The transplanted head regained consciousness, tracked with its eyes, and responded to stimuli; the animals survived for periods reported as hours to a few days before succumbing to rejection or complications.[^white1971] Crucially, White made no attempt to join the spinal cords. The animals were quadriplegic and ventilated, and White himself described the result as a demonstration that the brain could be maintained on a foreign circulation, not as a functional transplant.

## The modern proposal

In 2013 the Italian neurosurgeon Sergio Canavero published an outline for a human procedure, which he called HEAVEN, with an associated protocol for spinal cord fusion he called GEMINI.[^canavero2013] Its elements were a very sharp blade to sever the cord with minimal crush injury, deep hypothermia to extend the brain's tolerance of interrupted perfusion, immediate application of polyethylene glycol as a fusogen to promote membrane fusion between cut axons, and a period of spinal cord stimulation and immobilisation during recovery.

Canavero announced a Russian volunteer with spinal muscular atrophy, who later withdrew. He collaborated with Xiaoping Ren at Harbin Medical University, and in 2017 the pair announced what press coverage described as the first human head transplant. It was a rehearsal on two cadavers: the vessels, cord and tissues of two deceased donors were joined over many hours. Nothing was restored to life, and no functional claim could be tested. Reports of recovery after cord transection and polyethylene glycol treatment in rodents have come almost entirely from the same collaboration and have not been independently replicated.

The reaction from neuroscience and neurosurgery was uniformly negative. The bioethicist Arthur Caplan described the programme as unsupported by science; leaders of professional neurosurgical bodies stated they would not wish the outcome on any patient; spinal cord researchers, including some who had worked with White, characterised the fusion claims as unfounded. In 2024 a promotional concept video for a robotic head-transplant system circulated widely under the name BrainBridge; it was an animation produced outside any research institution and presented no experimental work at all.

> [!caution] What was reported and what was done
> The 2017 announcement is a case study in how a procedural rehearsal becomes a claimed achievement. Joining tissues between two cadavers demonstrates that the anatomy can be dissected and sutured. It says nothing about circulation, consciousness, immune tolerance or neurological function, which are the entire content of the proposal.

## The spinal cord problem

Central nervous system axons do not regenerate across a lesion in adult mammals. The reasons are well characterised: inhibitory molecules in myelin debris, a glial scar rich in chondroitin sulfate proteoglycans, an unfavourable growth state in the injured neuron, and the absence of guidance cues that existed only during development.[^silver2004] Decades of work on chondroitinase treatment, neural stem cell grafts, growth-factor scaffolds and epidural electrical stimulation have produced meaningful partial recovery in *incomplete* injuries, where surviving fibres can be strengthened. A completely transected cord reconnected end to end is a different problem, and no laboratory has restored voluntary movement after one in a large animal. Peripheral nerves regenerate slowly but genuinely; central tracts do not, which is why the difficulty here is closer to that of [[limb-regeneration]] than to that of nerve repair.

Restoring function would require far more than membrane fusion. Millions of axons would have to reconnect, each to a target of the appropriate class, in a topographically correct arrangement, and the resulting circuit would have to be usable by a brain whose motor map was built for a different body. Polyethylene glycol can fuse the membranes of adjacent cut axons in laboratory preparations; it does not perform routing.

The scale of the difficulty was underlined by a 2023 transplant at NYU Langone in which a patient received a partial face transplant that included an entire eye. The globe survived with perfused retina, which was itself a first. Vision was not restored, because the optic nerve is a central tract and could not be reconnected — a single nerve, under ideal circumstances, with the world's attention on it. Restoring sight after optic nerve loss currently requires bypassing the nerve altogether, which is what a cortical visual prosthesis attempts and why the devices in [[retinal-implant]] exist.

## Other barriers

**Ischaemia.** The brain tolerates warm interruption of blood flow for a few minutes. Hypothermia extends this substantially, and cardiac surgery routinely uses circulatory arrest under deep cooling for limited periods, but the head would need continuous or near-continuous perfusion through a procedure lasting many hours, requiring cross-circulation arrangements of considerable complexity.

**Immunology.** The immune system resides in the body: bone marrow, thymus, spleen, lymph nodes. A transplanted head is therefore foreign tissue confronting an intact donor immune system, and would face rejection of brain, skull, skin and sensory organs simultaneously, requiring immunosuppression at levels beyond those used in face and hand transplantation. Vascularized composite allotransplantation already has high complication rates from immunosuppression alone, and none of the tolerance strategies developed for [[xenotransplantation]] and conventional grafting has been tested on a target this large. The one established transplant with a planned exit is [[uterus-transplantation]], where the graft is removed after a pregnancy so that immunosuppression can be stopped; nothing comparable is available here.

**Autonomic control.** Sympathetic outflow to the heart and vasculature descends through the cervical cord. A severed cord produces neurogenic shock with profound hypotension, and if the anastomosis is above the phrenic nerve roots, permanent ventilator dependence. Temperature regulation, bladder and bowel function, and blood pressure control are all lost.

**Timing and consent.** The procedure requires a beating-heart donor body and a recipient prepared simultaneously in an adjacent theatre, with the donor's family consenting to a use of the body unlike any existing donation category. That body would otherwise supply several transplantable organs to people on a waiting list, so the procedure consumes rather than relieves the deficit described in [[organ-shortage]].

## Ethics and identity

Even granting technical success, the procedure raises questions the surgical literature does not usually face. The resulting person would carry another individual's genome in every cell below the neck, including the gametes, so any children conceived would be genetically the donor's. Legal identity, in most systems, would follow the head, but no statute anywhere addresses the case. The psychological consequences of waking in an unfamiliar body are unknown; limb transplant recipients have occasionally requested removal of a functioning graft, and a whole body is not comparable.

The philosophical question is thinner than it looks. Almost every theory of personal identity locates the person with the brain, so a head transplant is a body transplant and the identity puzzle is mild compared with the cases discussed in [[personal-identity-and-continuity]] and [[teleportation-problem]]. The serious ethical objection is not metaphysical but clinical: performing an irreversible operation whose central step has never worked in any animal, on a patient who is typically disabled rather than dying, and whose realistic outcome is death or a state worse than the one they began in. Appeals to [[morphological-freedom]] do not settle this, since a right to modify one's own body is generally held to presuppose that the modification is technically possible and the consent adequately informed.

## Where the idea stands

Head transplantation occupies a peculiar position: it is not disproved, merely dependent on a prerequisite that nobody can supply. If complete spinal cord transection became repairable, the achievement would transform the treatment of the hundreds of thousands of people living with spinal cord injury, and that population — not head transplant candidates — is where such a technique would be used first and where its success would be measured.

The motivations offered for the procedure are better served by other approaches. Spinal muscular atrophy, the condition of the volunteer who was announced and then withdrew, now has approved gene therapies of the kind described in [[somatic-gene-therapy]], which arrived within a few years of the head transplant proposal and changed the prognosis for the disease it was meant to address. For other degenerative conditions the realistic paths remain gene and cell therapy and assistive technology, including [[brain-computer-interface]] control of external devices and [[exoskeleton|powered exoskeletons]]. For the ambition of escaping a failing body altogether, the relevant articles are [[cryonics]], [[whole-brain-emulation]], [[mind-uploading]] and [[neuroprosthetics]], each of which faces its own unsolved problems but none of which requires reconnecting a spinal cord this year.

## See also

- [[neuroprosthetics]]
- [[personal-identity-and-continuity]]
- [[organ-shortage]]
- [[xenotransplantation]]
- [[limb-regeneration]]
- [[cryonics]]
- [[brain-computer-interface]]
- [[morphological-freedom]]

## References

[^white1971]: `paper` White, R. J. et al. "Cephalic exchange transplantation in the monkey." *Surgery*, 1971. {The spinal cords were never joined, so the animals were quadriplegic and ventilated, and survival was hours to a few days.}
[^canavero2013]: `paper` Canavero, S. "HEAVEN: The head anastomosis venture Project outline for the first human head transplantation with spinal linkage (GEMINI)." *Surgical Neurology International*, 2013. {A protocol proposal rather than a report of experiments; it presents no new data of its own.}
[^silver2004]: `paper` Silver, J. and Miller, J. H. "Regeneration beyond the glial scar." *Nature Reviews Neuroscience*, 2004.

==============================================================================
ARTICLE: healthspan
TITLE:   Healthspan
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/healthspan
SOURCE:  https://futurehumanwiki.com/raw/healthspan
==============================================================================

---
title: "Healthspan"
slug: "healthspan"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity", "foundations"]
tags: ["aging", "healthspan", "frailty", "public health", "endpoints", "demography"]
summary: "The portion of life spent in good health and function, as distinct from total lifespan; widely used as a goal but not standardized as a measure."
updated: "2026-07-27"
issues: ["HALE and intrinsic capacity are described without a WHO source."]
---

```infobox
{
  "caption": "Concept in gerontology and public health",
  "rows": [
    { "label": "Definition", "value": "Years lived in good function" },
    { "label": "Contrast with", "value": "Lifespan (total years)" },
    { "label": "Standard measure", "value": "None agreed" },
    { "label": "Common proxies", "value": "HALE, DFLE, frailty index" },
    { "label": "Typical gap to lifespan", "value": "About a decade" },
    { "label": "Status", "value": "In routine use, unstandardized" }
  ]
}
```

**Healthspan** is the period of a person's life spent free of significant chronic disease and functional impairment. It is the quantity most longevity research says it is trying to extend, and the one it measures least consistently. Unlike lifespan, which has a single unambiguous endpoint, healthspan requires a decision about what counts as ill health — and that decision changes the answer by years.

## Why the definition matters

Lifespan is settled by a death certificate. Healthspan requires drawing a line somewhere on a continuum of decline, and the plausible lines are far apart. A definition based on the first diagnosis of any chronic condition puts the end of healthspan in the fifties for most people in high-income countries, since hypertension and osteoarthritis are near-universal. A definition based on loss of independence puts it in the late seventies or eighties. A definition based on self-rated health tracks neither reliably.

This is not a pedantic problem. It determines whether a given intervention appears to extend healthspan, compress it, or leave it unchanged, and it makes results from different studies non-comparable. The same cohort can show a decade of difference depending on which operationalization the analyst chose.

> [!note] Terminology
> "Healthspan" has no agreed technical definition and no standard measurement instrument. Papers using the word should be read for how they operationalized it; the choice is usually buried in the methods section and usually drives the headline result.

## Measures in practice

Several formal constructs are used as stand-ins.

**Healthy life expectancy (HALE)** is the World Health Organization's measure: life expectancy adjusted downward by time lived with disability, weighted by severity, using disability weights from the Global Burden of Disease programme. It is a population statistic computed from life tables and prevalence data, generally by the Sullivan method, and cannot be applied to an individual.

**Disability-free life expectancy (DFLE)** is the simpler binary version — years expected before the onset of a defined disability — and is more sensitive to where the disability threshold is set.

**Frailty indices** count accumulated deficits across dozens of clinical and functional items and express them as a proportion. The cumulative-deficit approach of Rockwood and colleagues and the phenotypic definition of Fried and colleagues, based on weight loss, exhaustion, weakness, slowness and low activity, are the two dominant families.[^fried2001] Frailty indices predict mortality and hospitalization well and are among the most portable measures across species.

**Intrinsic capacity**, developed by the WHO for its integrated care framework, aggregates locomotion, vitality, cognition, psychological state and sensory function. It is designed to be measured in primary care rather than in a research cohort.

None of these measures healthspan directly. Each measures something correlated with it, and the correlations among them are imperfect.

## The healthspan–lifespan gap

Across countries, people spend the last stretch of life in impaired health, and that stretch has not shrunk as lifespan has grown. A 2024 analysis of WHO member-state data reported a mean gap between life expectancy and healthy life expectancy of roughly a decade, wider in women than in men, wider in high-income countries than in low-income ones, and largest in the United States.[^garmany2024]

The direction of change matters more than the level. If added years of life arrive mainly as added years of disability, the gain is smaller than the mortality statistics suggest — the expansion-of-morbidity scenario. The opposing possibility, in which illness is pushed into a shorter interval before death, is the subject of [[compression-of-morbidity]], and the evidence for it in national data has been mixed at best.

> [!key] Why the gap widens rather than closes
> Reducing mortality from a disease is not the same as reducing its prevalence. Countries that have been most successful against cardiovascular disease and cancer have largely converted fatal events into survivable chronic conditions, which lengthens lifespan and prevalence together. A therapy that prevented the disease outright would move the two measures in opposite directions.

## Healthspan in animal research

The distinction between lifespan and healthspan is sharper in model organisms, where both can be measured on a practical timescale. Mouse studies use grip strength, rotarod performance, gait speed, treadmill endurance, cognitive tasks, and standardized frailty indices adapted from clinical scales.

The results are a caution against assuming the two move together. Work in *Caenorhabditis elegans* found that several long-lived mutants spend a *greater* proportion of life in a frail, low-function state than wild-type worms — lifespan extended, healthspan proportionally not.[^bansal2015] That revives the objection first raised against [[cynthia-kenyon|Cynthia Kenyon's]] long-lived insulin-signalling mutants, that an animal may be living longer by living less. Similar dissociations appear in mice for some interventions. Several established geroprotectors do appear to extend both: [[caloric-restriction]] and [[rapamycin]] improve multiple functional measures in rodents alongside lifespan, and [[senolytics]] were developed specifically around functional endpoints such as gait speed and cardiac function rather than survival curves.

Whether any of this transfers is unresolved. The human evidence for [[metformin]] and for NAD+ precursors addresses metabolic markers, not function; the trials of [[heat-and-cold-exposure|heat and cold exposure]] are short and mechanistic; the only intervention with strong human evidence for preserving function is [[exercise-and-aging|physical activity]].

## As an endpoint

Healthspan's ambiguity is a practical obstacle for the field. Regulators approve drugs against defined indications with measurable endpoints, and "extends healthspan" is neither. This is one reason the [[geroscience-hypothesis]] programme has concentrated on composite outcomes — the incidence of multiple age-related diseases in a single trial — rather than on healthspan as such, and why [[aging-biomarkers|biomarker validation]] and [[epigenetic-clock|clock-based]] measures attract disproportionate attention.

Prize competitions have taken the opposite approach by specifying function directly. [[xprize-healthspan]] defines its target as the restoration of muscle, cognitive and immune function in older adults, measured against a person's own baseline, over a fixed treatment window. Specifying the domains in advance sidesteps the definitional argument at the cost of ignoring everything outside them.

The commercial market has adopted the word almost entirely without measurement. Longevity clinics and [[dietary-supplements|supplement vendors]] advertise healthspan extension on the basis of [[biological-age|biological age]] readouts, which measure neither health nor span.

## Open problems

The central unresolved question is whether healthspan and lifespan are separable in humans at all, or whether extending one necessarily drags the other along. The [[hallmarks-of-aging]] framework implies they should be coupled, since the same damage processes drive both mortality and dysfunction. The animal data show they can come apart. If they can come apart in humans, an intervention could produce more disability rather than less — a possibility that longevity research discusses less than it should.

A second problem is that healthspan cannot be measured prospectively in an individual. A person's healthspan is known only in retrospect, which makes it useless as a clinical decision variable and pushes practice back onto proxies: frailty scores, functional tests, and the [[inflammaging|inflammatory]] and [[stem-cell-exhaustion|regenerative]] markers that correlate with them. Whether any proxy is good enough to run a trial on is the question the field has to answer before it can claim to be extending healthspan at all.

## See also

- [[compression-of-morbidity]]
- [[maximum-human-lifespan]]
- [[longevity-escape-velocity]]
- [[geroscience-hypothesis]]
- [[aging-biomarkers]]
- [[longevity-dividend]]
- [[xprize-healthspan]]

## References

[^fried2001]: `paper` Fried, L.P. et al. "Frailty in older adults: evidence for a phenotype." *The Journals of Gerontology: Series A*, 2001.
[^bansal2015]: `paper` Bansal, A., Zhu, L.J., Yen, K. and Tissenbaum, H.A. "Uncoupling lifespan and healthspan in *Caenorhabditis elegans* longevity mutants." *PNAS*, 2015. {The uncoupling was measured in nematodes; whether long-lived mammals spend a similar proportion of life in a frail state is not established.}
[^garmany2024]: `paper` Garmany, A. and Terzic, A. "Global Healthspan-Lifespan Gaps Among 183 World Health Organization Member States." *JAMA Network Open*, 2024. {An analysis of country-level WHO estimates rather than individual data, and healthy life expectancy is itself modelled from disability weights.}

==============================================================================
ARTICLE: heat-and-cold-exposure
TITLE:   Heat and cold exposure
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/heat-and-cold-exposure
SOURCE:  https://futurehumanwiki.com/raw/heat-and-cold-exposure
==============================================================================

---
title: "Heat and cold exposure"
slug: "heat-and-cold-exposure"
type: "intervention"
status: "contested"
horizon: "present"
trl: 5
categories: ["longevity", "enhancement"]
tags: ["sauna", "cold exposure", "hormesis", "brown adipose tissue", "heat shock proteins", "recovery", "observational epidemiology"]
summary: "Deliberate exposure to heat or cold, from sauna bathing to cold-water immersion, as a health practice built on a hormetic mechanism whose effect on human lifespan is unmeasured."
updated: "2026-07-27"
humanEvidence: "Randomized trials show short-term changes in blood pressure, vascular function and mood; the mortality findings come from an observational Finnish cohort of men, and no trial has tested lifespan or healthspan."
access: "Unregulated and widely available: public and gym saunas, home units, cold-plunge studios and open water, at costs from a swim ticket to four-figure equipment."
reversibility: "reversible"
issues: ["Cold-immersion mood literature is summarized qualitatively without a cited meta-analysis", "Heat acclimation for athletic performance is uncovered and may warrant its own section"]
---

```infobox
{
  "caption": "Thermal exposure practice",
  "rows": [
    { "label": "Common forms", "value": "Sauna, hot bath, cold plunge, winter swimming" },
    { "label": "Proposed mechanism", "value": "Hormesis" },
    { "label": "Heat effectors", "value": "Heat-shock proteins, cardiovascular strain" },
    { "label": "Cold effectors", "value": "Brown fat, catecholamines" },
    { "label": "Strongest human data", "value": "Observational, sauna" },
    { "label": "Lifespan trial", "value": "None" },
    { "label": "Regulation", "value": "Not regulated as medicine" },
    { "label": "Readiness", "value": "TRL 5" }
  ]
}
```

**Heat and cold exposure** is the deliberate use of thermal stress, in forms such as sauna bathing,
hot-water immersion, cold plunging and winter swimming, as a general health practice rather than as
treatment for a named disease. The case for it rests on hormesis: a stressor too mild to injure
provokes a defensive response that leaves the organism better off than it was. The cellular
mechanisms invoked are real and well characterized. The human outcome evidence is much thinner than
the public conversation implies, and the strongest piece of it is a cohort study that cannot
establish cause.

```keyfacts
[
  { "value": "2009", "label": "Brown fat confirmed in adult humans", "note": "three imaging studies in one journal issue" },
  { "value": "2015", "label": "Finnish sauna cohort published", "note": "observational, middle-aged men" },
  { "value": "0", "label": "Randomized trials with a mortality endpoint", "note": "for heat or cold, as of 2026" }
]
```

## How it works

Hormesis names a dose–response curve in which a small amount of a stressor produces the opposite
effect to a large amount, and heat and cold are among its textbook cases. A rise in core temperature
triggers the heat-shock response: chaperone proteins are transcribed rapidly and set about refolding
or disposing of damaged protein, a direct intervention in the
[[proteostasis|protein quality-control system]] that fails with age and appears in most catalogues of
the [[hallmarks-of-aging]]. In nematodes, the heat-shock transcription factor HSF-1 is
required for the long life of the insulin-signalling mutants that [[cynthia-kenyon|Cynthia Kenyon's]]
laboratory made famous, and raising its activity extends life in the same species.

Heat also imposes a circulatory load. Skin blood flow rises sharply, heart rate climbs, plasma volume
expands with acclimation, and the pattern of cardiac output resembles a bout of moderate
[[exercise-and-aging|aerobic exercise]] without the muscular work. Repeated passive heating improves
endothelial function and lowers arterial stiffness in small controlled studies of sedentary
adults.[^brunt2016] This is the least speculative mechanism on offer, and also the one that most
obviously invites the question of why a person would not simply exercise.

Cold works through different machinery. Falling skin temperature drives a surge of noradrenaline and
recruits brown adipose tissue, a mitochondria-dense fat depot that generates heat by uncoupling
respiration from ATP synthesis through the protein UCP1. Adult humans were long assumed to lack
functional brown fat; imaging studies published in 2009 showed otherwise.[^cypess2009] Days of mild
cold acclimation increase brown-fat activity in healthy volunteers, and a small study in people with
type 2 diabetes reported improved insulin sensitivity afterwards, an effect its authors linked to
skeletal-muscle glucose uptake rather than to brown fat.[^hanssen2015] Cooling separately induces
cold-shock proteins; the best studied, RBM3, has been tied to preserved synapses in cooled mice and
has not been shown to do anything comparable in people. Both extremes
converge on stress-response pathways shared with [[autophagy]] and with the nutrient-sensing axis
that [[caloric-restriction]] engages.

> [!note] What hormesis does and does not license
> Hormesis is well documented for particular stressors at particular doses against particular
> endpoints. It is not a general rule that discomfort is good for an organism. The dose at which
> heat or cold crosses from hormetic to harmful in humans has not been mapped, and no exposure has a
> validated dose–response curve for any aging outcome.

## Development history

```timeline
[
  { "year": "1962", "title": "The heat-shock response is discovered", "text": "Ferruccio Ritossa notices a distinctive puffing pattern in Drosophila salivary-gland chromosomes after a temperature rise, the observation that leads to the heat-shock proteins." },
  { "year": "2003", "title": "Heat-shock factor tied to lifespan", "text": "Work in C. elegans shows that the heat-shock transcription factor HSF-1 is required for the extended life of insulin-signalling mutants, linking a thermal defence system to aging biology." },
  { "year": "2006", "title": "Cooler mice live longer", "text": "Transgenic mice engineered to run a slightly lower core body temperature show longer median lifespan without eating less, dissociating body temperature from caloric intake." },
  { "year": "2009", "title": "Brown fat confirmed in adults", "text": "Three groups publish PET-imaging evidence in a single journal issue that adult humans retain cold-activated brown adipose tissue, overturning the view that it disappears after infancy." },
  { "year": "2013", "title": "A cold-sensing longevity pathway", "text": "A thermosensitive TRP channel is shown to mediate lifespan extension at low temperature in C. elegans, indicating a regulated program rather than merely slowed chemistry." },
  { "year": "2015", "title": "Finnish sauna cohort published", "text": "An analysis of the Kuopio Ischaemic Heart Disease cohort associates frequent sauna bathing with lower cardiovascular and all-cause mortality in middle-aged men." },
  { "year": "2015", "title": "Cold immersion blunts strength gains", "text": "A controlled study reports that regular post-exercise cold-water immersion reduces long-term muscle hypertrophy and strength adaptation compared with active recovery." },
  { "year": "2016", "title": "Hyperthermia tested against depression", "text": "A small sham-controlled trial reports that a single session of whole-body hyperthermia reduces depressive symptoms for several weeks." },
  { "year": "2020", "title": "Sauna culture recognized by UNESCO", "text": "Finnish sauna culture is added to the Representative List of the Intangible Cultural Heritage of Humanity, as sauna and cold plunging spread through commercial fitness elsewhere." }
]
```

## The sauna evidence

The most-cited human finding comes from the Kuopio Ischaemic Heart Disease Risk Factor Study, a
prospective cohort of middle-aged Finnish men. Participants who reported bathing four to seven times
a week had lower rates of sudden cardiac death, fatal coronary disease and all-cause mortality than
those bathing once a week, with a graded relationship across frequency.[^laukkanen2015] Later
analyses of the same cohort reported inverse associations with other outcomes, including dementia.

The design is observational, and the obvious objection is severe. Sauna frequency is chosen, not
assigned, and the ability to sit in a hot room several times a week is itself a measure of health.
People with angina, heart failure, orthostatic intolerance or poor exercise capacity self-select out.
Frequency plausibly tracks income, employment and social contact as well. Statistical
adjustment cannot remove confounding by a variable measured worse than the exposure, and reverse
causation by subclinical disease is the failure mode this design handles worst. The finding is a
good reason to run a trial and a poor reason to assert an effect.

Randomized work in heat exists but is small and short. Beyond the vascular studies noted above, a
sham-controlled trial of a single whole-body hyperthermia session reported reduced depressive
symptoms lasting weeks in adults with major depression, a result worth replicating at scale and too
small to treat as established.[^janssen2016] None of this touches mortality, and none has run long
enough to speak to [[healthspan]].

## The cold evidence

Cold-water immersion has far less behind it. There is no cohort study of cold exposure and mortality
comparable to the Finnish sauna data, and no randomized trial with a clinical endpoint.

What exists is mechanistic and short-term: brown-fat recruitment, improved insulin sensitivity in
small acclimation studies, and an acute rise in circulating catecholamines that plausibly explains
the alertness regular plungers describe. A widely discussed experiment found that volunteers trained
in a method combining cold exposure with cyclic hyperventilation showed a blunted inflammatory
response to injected endotoxin; the breathing component appears to carry much of that effect, which
makes it weak evidence about cold as such.

The mood literature is the weakest part of the case and the most confidently reported. Studies are
mostly small, uncontrolled and impossible to blind, in a practice whose participants have paid for it
and expect it to work. The same expectancy problem constrains inference in [[psychedelic-therapy]].

> [!caution] Cold exposure is not free
> Repeated cold-water immersion immediately after resistance training reduces long-term gains in
> muscle size relative to active recovery, an effect reported across several controlled studies and
> pooled analyses; the strength penalty is found less consistently than the hypertrophy
> one.[^roberts2015] The likely reason is that cold suppresses the inflammatory and anabolic
> signalling the training stimulus depends on. An intervention that blunts an adaptive response
> cannot be assumed to be adaptive itself, which matters wherever recovery practice meets
> [[enhancement-in-sport|athletic preparation]].

> [!debate] Which way does temperature run?
> Many poikilotherms live longer in the cold, a cold-sensing channel mediates the effect in
> nematodes, and transgenic mice running a slightly lower core temperature outlived controls without
> eating less.[^conti2006] Sauna epidemiology points the other way. These cannot both be general laws
> about body temperature, and the field has not reconciled them. The obvious reconciliation is that a
> transient thermal stress and a sustained thermoregulatory set point are different variables, but no
> human data distinguish them.

## Limitations

Neither exposure can be blinded. A participant always knows whether they are in a hot room or a cold
tub, so every trial is at best single-blind against an active comparator, and subjective endpoints
carry an expectancy component that cannot be subtracted.

The dose is undefined. Temperature, duration, frequency, humidity, immersion depth and the interval
before or after exercise all vary between studies and venues, and no protocol has been shown to be
the right one. Popular guidance routinely converts a cohort study's exposure categories into a
prescription the study cannot support.

Endpoint substitution is the recurring problem across geroscience. A shift in an
[[epigenetic-clock]] reading or a composite [[biological-age|age estimate]] after a few weeks of
sauna use moves a predictor, not a demonstrated outcome, the point argued under [[aging-biomarkers]].
No regulator accepts any of these as a surrogate, and the honest summary of the lifespan work on
heat-shock proteins and [[mitochondrial-dysfunction|mitochondrial uncoupling]] is that it is animal
work.

## Risks

Heat carries real hazards. Prolonged exposure produces hyperthermia, dehydration and orthostatic
hypotension, and syncope on standing from a hot bath is a common mechanism of injury. Alcohol
consumption alongside sauna use is a documented contributor to sauna-related deaths in Finland.
The risk profile differs clearly for people with unstable cardiovascular disease and for pregnant
women in early gestation, since sustained elevation of core temperature in early pregnancy is
associated with developmental harm.

Cold's dangers are more acute. Sudden immersion triggers the cold shock response: an involuntary
gasp, uncontrolled hyperventilation, tachycardia and a blood-pressure surge, with drowning the
leading cause of death in cold water and the first minute the most dangerous.[^tipton2017]
Simultaneous sympathetic and parasympathetic activation can provoke arrhythmia in susceptible people.
Longer immersions bring progressive loss of manual dexterity and swim failure before core temperature
falls far. These are ordinary risks of an ordinary activity, but they are systematically absent from
the marketing.

## Outlook

Unlike [[rapamycin]] or a senolytic, these exposures are cheap, legal and already practised by
millions, so the barrier is evidence rather than access. A cardiovascular outcomes trial of regular
sauna use in a high-risk population is feasible in principle and has not been run. Researchers in the
field generally attribute the gap to incentives: nobody owns the intervention, so nobody funds the
trial, the structural problem that also stalls generic-drug work under the
[[geroscience-hypothesis]].

Pharmacological brown-fat activation has been pursued as a metabolic target for years without
producing an approved agent, and effective [[glp-1-receptor-agonists|incretin drugs]] have given
industry a shorter route to the same metabolic endpoints. Consumer instrumentation is meanwhile making the practice
measurable: heart-rate variability and temperature estimates from
[[wearable-health-sensors|wearables]], logged by [[quantified-self|self-quantifiers]] and
[[biohacking|biohackers]], generate large uncontrolled datasets whose value is hypothesis generation.

The defensible position as of 2026 is that these practices are plausible, mostly safe, poorly
characterized and oversold. They sit with [[dietary-supplements]] and
[[blue-zones|longevity folklore]] rather than with interventions carrying outcome data, and the
comparison that matters is not against doing nothing but against the better-evidenced alternatives:
exercise above all, then the treatable causes of disturbed [[sleep-and-longevity|sleep]] and of the
chronic [[inflammaging|low-grade inflammation]] that a cold tub does not address.

## See also

- [[exercise-and-aging]]
- [[caloric-restriction]]
- [[proteostasis]]
- [[human-hibernation]]
- [[aging-biomarkers]]
- [[healthspan]]
- [[biohacking]]
- [[geroscience-hypothesis]]

## References

[^laukkanen2015]: `paper` Laukkanen, T., Khan, H., Zaccardi, F., Laukkanen, J.A. "Association Between Sauna Bathing and Fatal Cardiovascular and All-Cause Mortality Events." *JAMA Internal Medicine*, 2015. {Prospective cohort of middle-aged Finnish men. Sauna frequency was self-reported and not assigned, so tolerance of heat is itself a marker of health.}
[^brunt2016]: `paper` Brunt, V.E. et al. "Passive heat therapy improves endothelial function, arterial stiffness and blood pressure in sedentary humans." *The Journal of Physiology*, 2016. {Small, several weeks long, hot-water immersion rather than sauna, and endpoints are vascular measures rather than events.}
[^janssen2016]: `paper` Janssen, C.W. et al. "Whole-Body Hyperthermia for the Treatment of Major Depressive Disorder: A Randomized Clinical Trial." *JAMA Psychiatry*, 2016. {A few dozen participants and a single session; the sham condition is difficult to make convincing when the active arm raises core temperature.}
[^cypess2009]: `paper` Cypess, A.M. et al. "Identification and Importance of Brown Adipose Tissue in Adult Humans." *New England Journal of Medicine*, 2009. {One of three independent imaging reports published in the same issue, which is why the result was accepted quickly.}
[^hanssen2015]: `paper` Hanssen, M.J.W. et al. "Short-term cold acclimation improves insulin sensitivity in patients with type 2 diabetes mellitus." *Nature Medicine*, 2015. {Small metabolic study over days, in a patient population; it does not test whether the change persists or affects outcomes.}
[^roberts2015]: `paper` Roberts, L.A. et al. "Post-exercise cold water immersion attenuates acute anabolic signalling and long-term adaptations in muscle to strength training." *The Journal of Physiology*, 2015. {The comparator was active recovery, not rest, which makes the blunting harder to attribute to anything but the cold.}
[^conti2006]: `paper` Conti, B. et al. "Transgenic mice with a reduced core body temperature have an increased life span." *Science*, 2006. {The temperature drop was engineered centrally rather than imposed environmentally, so it is evidence about set point, not about cold plunging.}
[^tipton2017]: `paper` Tipton, M.J., Collier, N., Massey, H., Corbett, J., Harper, M. "Cold water immersion: kill or cure?" *Experimental Physiology*, 2017.

==============================================================================
ARTICLE: human-cloning
TITLE:   Human cloning
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/human-cloning
SOURCE:  https://futurehumanwiki.com/raw/human-cloning
==============================================================================

---
title: "Human cloning"
slug: "human-cloning"
type: "technology"
status: "contested"
horizon: "indefinite"
trl: 4
categories: ["reproduction", "genetics"]
tags: ["cloning", "reproduction", "stem cells", "bioethics", "regulation", "identity"]
summary: "Producing a human genetically identical to an existing person by somatic cell nuclear transfer, demonstrated to the blastocyst stage and never taken to a pregnancy."
updated: "2026-07-27"
humanEvidence: "Nuclear transfer has produced human blastocysts and embryonic stem cell lines in culture; no human pregnancy or birth from a cloned embryo has been documented, and all live-birth data are from other mammals."
access: "Not obtainable: reproductive cloning is prohibited by statute in most countries with a biotechnology sector and no clinic offers it, while nuclear transfer for research is done in a handful of laboratories."
reversibility: "irreversible"
issues: ["The 1998 Council of Europe protocol and the 2005 UN declaration are described without sources."]
---

```infobox
{
  "caption": "Reproductive and cell-biological technique",
  "rows": [
    { "label": "Method", "value": "Somatic cell nuclear transfer" },
    { "label": "First cloned mammal", "value": "Dolly the sheep, 1996" },
    { "label": "First cloned primates", "value": "2018, macaques" },
    { "label": "Human embryos", "value": "Blastocysts and stem cell lines, 2013" },
    { "label": "Human births", "value": "None reported" },
    { "label": "Legal status", "value": "Banned in most countries" }
  ]
}
```

**Human cloning** is the production of a person genetically identical to an existing individual, almost always understood to mean somatic cell nuclear transfer into an egg. Nuclear transfer has produced live offspring in more than twenty mammalian species, including primates, and has produced human embryos and stem cell lines. No human pregnancy from it has been reported anywhere, and the reasons are as much about a collapse in motivation as about law or technique.

## Two things called cloning

*Reproductive cloning* aims at a live birth. *Therapeutic cloning*, more precisely called nuclear transfer for stem cell derivation, aims only at a blastocyst from which embryonic stem cells genetically matched to a patient can be taken. The two share every step up to embryo transfer and diverge completely after it.

The distinction mattered enormously between about 1998 and 2007, when patient-matched stem cells were the main scientific argument for permitting nuclear transfer at all. It stopped mattering when [[induced-pluripotent-stem-cells]] arrived. Reprogramming a skin cell with the [[yamanaka-factors]] produces patient-matched pluripotent cells without an egg, without an embryo, and without the ethical objection that dominated a decade of legislation. Therapeutic cloning was rendered largely obsolete within a few years of being legalised in the places that legalised it.

## How nuclear transfer works

An unfertilised egg is held with a pipette and its own chromosomes removed. A somatic cell such as a fibroblast, a cumulus cell, or a lymphocyte is placed under the zona pellucida and fused with the egg by an electrical pulse, or its nucleus is injected directly. The reconstructed cell is activated chemically or electrically to mimic fertilisation, cultured to the blastocyst stage, and either transferred to a surrogate or used to derive stem cells.

The hard part is neither the micromanipulation nor the activation. It is that the egg cytoplasm must, within hours, erase the epigenetic configuration of a differentiated cell and impose that of a totipotent zygote. This is the same reprogramming that underlies [[epigenetic-reprogramming]] more generally, but performed by unidentified maternal factors on a schedule nobody controls. It fails most of the time. Embryos that survive frequently carry aberrant methylation at imprinted loci and abnormal placental development, producing the pattern known in livestock as large offspring syndrome.

```timeline
[
  { "year": "1996", "title": "Dolly", "text": "Ian Wilmut and Keith Campbell's team at the Roslin Institute produces a lamb from an adult mammary cell nucleus. Hundreds of reconstructed embryos yield a single live birth." },
  { "year": "1998–2005", "title": "Species accumulate", "text": "Mice, cattle, pigs, cats, and in 2005 the first dog are cloned. Commercial pet and livestock cloning follows." },
  { "year": "2004–2006", "title": "The Hwang fraud", "text": "Claims of human stem cell lines derived by nuclear transfer are published and then retracted as fabricated; the cloned dog from the same laboratory proves genuine." },
  { "year": "2013", "title": "Human stem cell lines", "text": "Shoukhrat Mitalipov's group derives verified human embryonic stem cell lines by nuclear transfer, using caffeine to stabilise the spindle." },
  { "year": "2018", "title": "Primates", "text": "A Shanghai team produces two long-tailed macaques from fetal fibroblasts, aided by a histone demethylase and a deacetylase inhibitor to relieve epigenetic barriers." },
  { "year": "2020s", "title": "Conservation cloning", "text": "A black-footed ferret and a Przewalski's horse are cloned from decades-old frozen cell lines to restore lost genetic diversity." }
]
```

## Dolly and what she showed

The 1997 report of Dolly's birth overturned a settled assumption that differentiation was irreversible in mammals.[^wilmut1997] The efficiency was poor: a few hundred reconstructed embryos produced one lamb. That ratio has improved but not transformed in the decades since.

Dolly was euthanised at six after developing a lung tumour caused by a retrovirus endemic in housed sheep, and her arthritis prompted speculation that clones age prematurely. Her telomeres were shorter than expected for her age. A later study of cloned sheep derived from the same cell line found them ageing normally into old age, which weakened the premature-ageing claim without settling the underlying question of how nuclear transfer interacts with the mechanisms described in [[telomeres-and-telomerase]].[^sinclair2016]

## Primates, and why humans are different

Primate cloning resisted for two decades. Part of the reason was technical: in primate eggs, spindle-associated proteins are removed along with the chromosomes during enucleation, disabling the reconstructed cell. Modified enucleation methods and chemical stabilisation of the spindle addressed this, and in 2013 human blastocysts and stem cell lines were produced.[^tachibana2013]

Live primate births required more. The 2018 macaque work added treatments that relieve epigenetic barriers, including messenger RNA for a histone demethylase and an inhibitor of histone deacetylases.[^liu2018] Even then, only embryos made from fetal fibroblasts produced surviving offspring; attempts using adult cells failed. Two live macaques resulted from a large number of transferred embryos and surrogate pregnancies.

That last fact is the honest state of the technology. Cloning a human from an adult cell would mean many surrogate pregnancies with a high rate of loss, malformation, and neonatal death, in a species where each of those events involves a person who has consented to something and a child who has not. No reputable body regards this as an acceptable risk profile, and the argument does not turn on symbolism.

> [!caution] Claims and their status
> Several groups have announced human cloning attempts or successes since 2001, including a religious organisation that claimed a birth in 2002. None provided verifiable evidence. As of 2026 no cloned human birth has been documented.

## A clone is not a copy

Public discussion has consistently overstated the relationship. A clone is a delayed monozygotic twin, and one raised in a different uterus, a different family, and a different decade. Identical twins share a prenatal environment and a birth cohort and are therefore more alike than a clone and its progenitor would be.

Nor is the genome identical in the strict sense. Mitochondrial DNA comes from the egg donor rather than the nuclear donor, so a clone made with a stranger's egg has a different mitochondrial genotype — the same asymmetry that makes [[mitochondrial-replacement-therapy]] a heritable modification. Somatic mutations accumulated in the donor cell are carried forward. Epigenetic marks, X-inactivation patterns, immune receptor repertoires and neural wiring all differ.

The identity question this raises is philosophically thinner than it looks. Nothing in [[personal-identity-and-continuity]] suggests that genetic identity constitutes personal identity; a clone is a distinct person with a distinct psychological history from the first moment. The real ethical concerns are about the expectations placed on a child created to resemble someone, which is a familiar problem in an unfamiliar form.

## Law and ethics

Reproductive human cloning is prohibited by statute in most countries with an established biotechnology sector. The Council of Europe adopted a protocol banning it in 1998; a United Nations declaration in 2005 called on states to prohibit cloning practices "incompatible with human dignity", passing by a divided vote and carrying no binding force. The United States has no federal statute; the Food and Drug Administration asserts jurisdiction over the procedure and a number of states ban it outright, an arrangement that resembles the gap analysed in [[governance-of-genome-editing]].

The ethical literature on cloning is unusually well developed because it preceded the technology's plausibility. [[leon-kass|Leon Kass]]'s argument that widespread repugnance can carry moral information was made about cloning first, and remains the most cited statement of the position surveyed in [[bioconservatism]].[^kass1997] Against it, critics argued that repugnance had also attended IVF and organ transplantation, and that the specific harms alleged, namely confused lineage, instrumentalisation, and loss of genetic uniqueness, either did not follow from the technique or already occurred in accepted practices.

What ultimately blocked reproductive cloning was neither argument nor law but the absence of a use case. For infertility, [[embryo-selection]] and donor gametes work better; for genetic disease, [[germline-editing]] and selection address the cause; for replacing a dead child, the technique does not do what the bereaved parent wants. The one domain where nuclear transfer remains actively useful is animal work, including livestock, the conservation cloning of species with collapsed genetic diversity, and the edited pigs used in [[xenotransplantation]] and by companies pursuing [[de-extinction]].

If a human is ever cloned, the most likely route is not a fertility clinic but a well-funded private effort in a permissive jurisdiction, and the most likely reason will be that someone wanted it done. Whether the efficiency and epigenetic problems can be reduced to a level that would make the attempt survivable is an empirical question that nobody has an ethical way to answer.

## See also

- [[induced-pluripotent-stem-cells]]
- [[epigenetic-reprogramming]]
- [[mitochondrial-replacement-therapy]]
- [[germline-editing]]
- [[de-extinction]]
- [[personal-identity-and-continuity]]
- [[bioconservatism]]
- [[in-vitro-gametogenesis]]

## References

[^wilmut1997]: `paper` Wilmut, I. et al. "Viable offspring derived from fetal and adult mammalian cells." *Nature*, 1997.
[^sinclair2016]: `paper` Sinclair, K. D. et al. "Healthy ageing of cloned sheep." *Nature Communications*, 2016.
[^tachibana2013]: `paper` Tachibana, M. et al. "Human embryonic stem cells derived by somatic cell nuclear transfer." *Cell*, 2013.
[^liu2018]: `paper` Liu, Z. et al. "Cloning of Macaque Monkeys by Somatic Cell Nuclear Transfer." *Cell*, 2018. {Only embryos made from fetal fibroblasts produced surviving offspring; attempts using adult donor cells, the case relevant to cloning an adult, failed.}
[^kass1997]: `paper` Kass, L. "The Wisdom of Repugnance." *The New Republic*, 1997. {An essay in a general-interest magazine written in the year after Dolly; it argues a position and reports no analysis of the technique.}

==============================================================================
ARTICLE: human-digital-twins
TITLE:   Human digital twins
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/human-digital-twins
SOURCE:  https://futurehumanwiki.com/raw/human-digital-twins
==============================================================================

---
title: "Human digital twins"
slug: "human-digital-twins"
type: "technology"
status: "experimental"
horizon: "2040s"
trl: 4
categories: ["bodies", "foundations"]
tags: ["simulation", "in silico trials", "personalized medicine", "computational modelling", "medical devices", "cardiology"]
summary: "Computational models of an individual person's physiology, built from that person's own data and run to predict what an intervention would do before it is attempted."
updated: "2026-07-27"
humanEvidence: "In patients, simulation is routine only for radiotherapy dose and CT-derived coronary physiology; virtual-heart ablation planning rests on small studies. No whole-body twin exists for any person."
access: "Organ-specific models are sold as regulated software and reimbursed in some health systems, notably CT-derived coronary flow analysis. Products marketed as whole-person twins are usually dashboards."
reversibility: "reversible"
issues: ["Needs a source for the 2025 FDA announcement on animal-testing alternatives", "Thin on oncology and immunology modelling"]
---

```infobox
{
  "caption": "Patient-specific computational modelling",
  "rows": [
    { "label": "Type", "value": "Simulation software" },
    { "label": "Term origin", "value": "Aerospace and manufacturing" },
    { "label": "Personalised from", "value": "Imaging, physiology, sensors" },
    { "label": "Deployed uses", "value": "Coronary flow, ablation planning, dose" },
    { "label": "Whole-body twin", "value": "None demonstrated" },
    { "label": "Credibility standard", "value": "ASME V&V 40 (2018)" },
    { "label": "Regulatory frame", "value": "Context of use" },
    { "label": "Readiness", "value": "TRL 4" }
  ]
}
```

**Human digital twins** are computational models of an individual person's physiology, built from that person's own imaging, measurements and sensor data and run to predict what an intervention would do before it is attempted. The idea is imported from aerospace and manufacturing, where a simulation of one specific engine is kept synchronised with the engine itself and used to decide when to service it. Medicine has no whole-body equivalent and is not near one. What exists is a set of organ- and device-specific models, each narrow, each personalised in a few respects and generic in the rest.

```keyfacts
[
  { "value": "0", "label": "Whole-body human twins in clinical use", "note": "as of 2026; every deployed model is organ- or device-specific" },
  { "value": "2014", "label": "Living Heart Project launched", "note": "Dassault Systèmes, with the FDA as a research partner" },
  { "value": "2018", "label": "ASME V&V 40 published", "note": "the credibility framework regulators now cite for medical models" }
]
```

## What the term covers

Three different objects travel under the name, and the difference between them is most of the subject.

The first is a mechanistic simulation whose geometry is taken from a patient's scan and whose equations come from physics and physiology: blood flow through that person's coronary tree, electrical propagation around the scar in that person's ventricle, the field a stimulation lead produces in that person's brain. The second is a statistical or machine-learned predictor fitted to population data and evaluated on one person's record, closer in kind to [[biological-age|a biological-age estimate]] than to a simulation. The third is a longitudinal data display: a record assembled from [[wearable-health-sensors]], [[continuous-glucose-monitoring]] and the laboratory, rendered next to a picture of a body.

Only the first can be intervened on, which is what the word twin promises. Vendors apply it to all three, and hospital software sold as a digital twin platform is usually the third. [[quantified-self|Self-tracking]] products have inherited the vocabulary as well. The distinction worth holding onto is between a model personalised with an individual's own measurements and a model carrying population parameters that merely displays that individual's numbers.

> [!note] Not an emulation
> A digital twin is a predictive instrument, judged by whether its outputs match measurements taken from the body it models. [[whole-brain-emulation]] proposes something categorically different: a simulation detailed enough that it might itself be a bearer of the original's mental life, which is the premise [[mind-uploading]] rests on. Treating the two as points on one scale imports the whole of [[personal-identity-and-continuity]] into what is otherwise an engineering question about model error.

## How one is built

A patient-specific model is assembled in four steps, and the third decides whether the object deserves the name.

Geometry comes first. A CT or MRI volume is segmented into a mesh of the relevant anatomy, a step now largely automated and genuinely individual: the mesh is that patient's artery, that patient's infarct. Physics comes second, as constitutive equations selected from the literature — a fluid model for blood, an ionic model for a cardiac cell, a compartmental model for a drug distributing through tissue.

Third is personalisation, meaning estimation of the model's parameters for this person. A cardiac electromechanics model carries tens to hundreds of parameters, far more than the independent measurements obtainable from a living patient. Most parameters are therefore fixed at population values while a small identifiable subset is tuned until the model reproduces something that was actually measured. The commercially deployed case makes the pattern plain: in CT-derived coronary flow analysis the arterial geometry belongs to the patient, but microvascular resistance and resting flow are assigned from scaling relationships between vessel size, myocardial mass and population physiology, because they cannot be measured non-invasively in that patient.

Fourth is the counterfactual run and its uncertainty. The model is re-solved with a stent in place, a lesion ablated, a dose changed. A prediction without an uncertainty interval cannot support a decision, and quantifying that interval is harder than the simulation itself, because it requires knowing how wrong the fixed population parameters might be.

A twin in the engineering sense is also supposed to stay synchronised, ingesting new data and re-estimating as the system changes. Almost no medical model does this. Most are built once from one scan, used for one decision, and then discarded.

## Development history

The name arrived from aerospace, where pairing a physical vehicle with a virtual model continuously updated from its sensors was proposed as a way to manage airframe fatigue over a lifetime of flights.[^glaessgen2012]

```timeline
[
  { "year": "Late 1990s", "title": "The Physiome Project", "text": "The International Union of Physiological Sciences launches a coordinated effort to build quantitative, multiscale models of human physiology and the file standards needed to exchange them." },
  { "year": "2002–2012", "title": "Digital twin named", "text": "Michael Grieves formulates the pairing of a physical artefact with a virtual counterpart updated from its data; NASA and US Air Force researchers give the idea its name and its first serious application." },
  { "year": "2007–2013", "title": "Virtual Physiological Human", "text": "European framework programmes fund a coordinated push toward integrated models of human physiology, followed by a standing institute to maintain the effort." },
  { "year": "2008", "title": "A simulator accepted instead of animals", "text": "The FDA accepts a metabolic simulator of type 1 diabetes as a substitute for animal testing in the preclinical evaluation of closed-loop insulin controllers." },
  { "year": "2014", "title": "Living Heart Project", "text": "Dassault Systèmes launches a collaborative high-fidelity model of the heart, with the FDA joining as a research partner for device evaluation." },
  { "year": "2016–2018", "title": "Regulators write the rules", "text": "FDA guidance sets out how computational modelling studies should be reported in device submissions; ASME publishes V&V 40, grading the credibility a model must show against the decision it informs." },
  { "year": "2018", "title": "An imaging trial run in software", "text": "FDA researchers replicate a breast-imaging comparison end to end on synthetic anatomy, with simulated detectors and algorithmic readers." },
  { "year": "2022", "title": "Animal testing no longer mandatory", "text": "The FDA Modernization Act 2.0 removes the blanket statutory requirement that new drugs be tested in animals and names computer modelling among acceptable alternatives." },
  { "year": "2023–2026", "title": "Marketing outruns evidence", "text": "Digital twin becomes standard vocabulary in hospital and device software while validated clinical uses stay confined to a handful of organ-scale models." }
]
```

## Where it works today

The most widely deployed patient-specific simulation in medicine is decades old and is rarely described as a twin. Radiotherapy planning computes the dose a beam arrangement will deposit throughout that patient's own CT volume, and treatment is prescribed from the model's output.

In cardiology, CT-derived fractional flow reserve estimates how far pressure falls across a coronary narrowing by solving flow equations on the reconstructed artery, sparing some patients a catheter. It is cleared for marketing, reimbursed in several health systems, and was evaluated against invasively measured pressure in patients already scheduled for catheterisation.[^norgaard2014] Separately, models of arrhythmia built from contrast-enhanced MRI of a patient's scar have been used to propose ablation targets, the simulation pacing the virtual heart to find where reentrant circuits can be induced.[^prakosa2018] That work has been tested in small studies, not in a large randomised trial.

In-silico testing of devices and drugs is further along than in-silico treatment of individuals. A metabolic simulator of type 1 diabetes was accepted as a substitute for animal testing in the preclinical evaluation of closed-loop insulin controllers, and it was used in developing control algorithms that went on to reach clinical use.[^kovatchev2009] FDA researchers later ran an entire imaging comparison in software, generating a population of synthetic breasts, simulating both modalities and reading the images with model observers; the result agreed with the direction of the patient trials.[^badano2018] Physiologically based pharmacokinetic models are routinely submitted to regulators to support dosing in groups never studied directly and to justify not running certain drug-interaction studies; related compartmental models are used to predict where a [[targeted-drug-delivery]] carrier ends up.

Elsewhere, field models estimate the tissue volume a [[deep-brain-stimulation]] lead activates in a specific patient's anatomy and are used to guide programming; musculoskeletal simulation scaled to a subject's dimensions and driven by their recorded motion informs orthopaedic surgery and [[exoskeleton]] design;[^delp2007] flow models support the evaluation of circulatory devices including the [[artificial-heart]]. Patient-derived [[organoids]] fill the same niche in wet biology, with the advantage of containing the person's actual cells rather than an estimate of their parameters.

## Regulatory standing

Regulators have moved further than clinicians. The FDA published guidance in 2016 on how computational modelling studies should be reported in device submissions,[^fda2016] and recognises a consensus standard, ASME V&V 40, that scales the credibility a model must demonstrate to the risk of the decision it informs.[^asme2018] The organising idea is context of use: a model is never validated in general, only for a stated question at a stated level of influence. A simulation used to screen candidate designs and a simulation used to replace a clinical trial are, under that framework, different objects with different burdens.

Law has moved the same way. The FDA Modernization Act 2.0, enacted at the end of 2022, removed the blanket requirement for animal testing of new drugs; in 2025 the agency announced a plan to reduce animal testing for certain classes of biologics, again naming computational methods among the alternatives. Neither change validates any particular model. Both raise the value of building one. Judged on [[technology-readiness-level]], the regulatory pathway is more mature than the science travelling down it.

## What is actually hard

The binding constraint is not compute. It is that a body is not a mechanism with known parameters, and the parameters that would individualise a model are mostly not measurable in a living person.

**Identifiability.** A model with more free parameters than independent measurements has many parameter sets that fit the data equally well and disagree about the intervention. Fixing the surplus at population values does not remove the problem; it hides it, and turns a personalised model into a population model wearing one patient's anatomy.

**No counterfactual to check against.** If the model recommends ablating one site and the clinician ablates it, the outcome of the alternative is never observed. Validating a predictive model of an individual therefore requires randomising against ordinary care, which is expensive and rarely done. A related shortage of checkable ground truth slows the validation of [[aging-biomarkers]].

**Scope.** A model represents what its builders chose to represent. Cardiac mechanics and drug distribution are tractable because the physics is well characterised and the geometry does much of the work. Immune response, metabolism, wound healing and the brain are not in that position, and a whole-body twin needs all of them at once, coupled.

**Drift.** A person is not a fixed system. Anatomy remodels, drugs change physiology, disease progresses. A model validated at one moment silently expires, and there is usually no signal telling anyone when.

> [!caution] The twin nobody has built
> No published work has produced a model of a whole human body, personalised to an individual, that predicts responses across organ systems. Roadmaps that show one arriving within a decade or two extrapolate from organ-scale successes, and those successes are concentrated exactly where the physics is simple and the anatomy carries the prediction. Cardiology and pharmacokinetics are not a representative sample of physiology.

> [!debate] Mechanism or learning
> One camp argues that only mechanistic models can answer interventional questions, because a model fitted to observed care cannot say what would happen under care that was never given. The other argues that mechanistic models will always be underdetermined in an individual, and that large learned models trained across millions of records will predict better in practice even if they explain nothing. Hybrids are common in the literature and are not yet common in anything a regulator has cleared.

## Risks and governance

A model dense enough to be useful is also a compact encoding of a person's health state, and one that can be queried for conditions the subject has not been told about. Existing protections were written for records and samples, not for parameterised models derived from them; the discrimination problem is the one already mapped by [[genetic-discrimination]], with the added wrinkle that a twin can be re-run against future questions. Models of neural tissue raise the concerns collected under [[mental-privacy]].

Overtrust is the nearer hazard. A simulation produces a specific number with a rendered picture attached, which is a persuasive combination regardless of the model's credibility grade, and clinicians are not generally equipped to audit the assumptions behind it. Liability is unsettled: a wrong prediction from software the clinician cannot inspect sits awkwardly between device failure and clinical judgment. Cost and data availability point the same way as everything else in [[access-and-inequality]], since the models are built from imaging and continuous monitoring that many health systems do not provide.

## Outlook

The direction with the most support behind it is not personal twins at all but virtual patient populations: cohorts of synthetic individuals, sampled to span real anatomical and physiological variation, used to test devices and dosing regimens across conditions no trial could enrol. That work has regulatory precedent, a credibility standard, and results. Proponents of the personal version argue it follows once enough individual data is routinely collected, and roadmaps published by research consortia describe a staged path from organ models to coupled ones. Sceptics point out that the missing ingredient is not integration but individual measurement, and that no amount of coupling fixes parameters nobody can observe.

Specific applications keep the idea alive. Long-duration missions are one, where a crew without a physician might carry a model of each member; [[space-medicine]] research treats this as attractive and unproven, and an astronaut's physiology changes in ways that are still being characterised, which is precisely the regime where personalisation fails. Learned components are another, with methods from [[ai-drug-discovery]] and structure prediction of the kind used in [[ai-protein-design]] proposed as a source of the parameters mechanistic models cannot measure; the proposal runs ahead of the evidence, since structure prediction returns coordinates rather than the binding affinities and conformational changes a physiological model would need. The benchmark that would settle the field's claims is a randomised comparison in which a personalised model predicts an individual's response to a therapy better than a clinician working from standard information. Trials of that design are rare, and none has been reported for a model spanning more than one organ system. Until that changes, the phrase describes an ambition and a set of organ models, not a technology anyone possesses.

## See also

- [[whole-brain-emulation]]
- [[organoids]]
- [[ai-drug-discovery]]
- [[wearable-health-sensors]]
- [[deep-brain-stimulation]]
- [[aging-biomarkers]]
- [[technology-readiness-level]]

## References

[^glaessgen2012]: `paper` Glaessgen, E. and Stargel, D. "The Digital Twin Paradigm for Future NASA and U.S. Air Force Vehicles." *53rd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference*, 2012. {The paper is about airframes; its framing of a twin as continuously updated from sensor data is the sense medicine borrowed and rarely implements.}
[^norgaard2014]: `paper` Nørgaard, B.L. et al. "Diagnostic Performance of Noninvasive Fractional Flow Reserve Derived from Coronary Computed Tomography Angiography in Suspected Coronary Artery Disease." *Journal of the American College of Cardiology*, 2014. {Diagnostic accuracy against invasive pressure measurement; the study does not show that acting on the model improves outcomes.}
[^prakosa2018]: `paper` Prakosa, A. et al. "Personalized virtual-heart technology for guiding the ablation of infarct-related ventricular tachycardia." *Nature Biomedical Engineering*, 2018. {Simulated targets were compared with the sites clinicians had actually ablated, in a small retrospective cohort.}
[^kovatchev2009]: `paper` Kovatchev, B.P. et al. "In Silico Preclinical Trials: A Proof of Concept in Closed-Loop Control of Type 1 Diabetes." *Journal of Diabetes Science and Technology*, 2009. {The simulator replaced animal testing only for control-algorithm evaluation; the devices themselves still required human trials.}
[^badano2018]: `paper` Badano, A. et al. "Evaluation of Digital Breast Tomosynthesis as Replacement of Full-Field Digital Mammography Using an In Silico Imaging Trial." *JAMA Network Open*, 2018. {Synthetic anatomy sampled from population statistics, not models of individual patients, and a comparison of imaging methods rather than of treatments.}
[^fda2016]: `regulator` U.S. Food and Drug Administration. *Reporting of Computational Modeling Studies in Medical Device Submissions: Guidance for Industry and Food and Drug Administration Staff*, 2016.
[^asme2018]: `report` American Society of Mechanical Engineers. *V&V 40-2018: Assessing Credibility of Computational Modeling Through Verification and Validation: Application to Medical Devices*, 2018. {Credibility is graded against the decision a model informs, so the same model can be adequate for one question and inadequate for another.}
[^delp2007]: `paper` Delp, S.L. et al. "OpenSim: Open-Source Software to Create and Analyze Dynamic Simulations of Movement." *IEEE Transactions on Biomedical Engineering*, 2007.

==============================================================================
ARTICLE: human-enhancement
TITLE:   Human enhancement
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/human-enhancement
SOURCE:  https://futurehumanwiki.com/raw/human-enhancement
==============================================================================

---
title: "Human enhancement"
slug: "human-enhancement"
type: "concept"
status: "contested"
horizon: "present"
categories: ["enhancement", "foundations"]
tags: ["enhancement", "bioethics", "therapy", "cognition", "fairness", "autonomy", "regulation"]
summary: "The deliberate use of biomedical or technical means to extend human capacities beyond what is needed to restore or maintain ordinary health."
updated: "2026-07-27"
issues: ["The claim about military performance-sustainment programmes names no state or programme.", "The Open problems claim about small effect sizes summarises a literature with no citation."]
---

```infobox
{
  "caption": "Concept in bioethics and applied philosophy",
  "rows": [
    { "label": "Field", "value": "Bioethics, philosophy of medicine" },
    { "label": "Core distinction", "value": "Therapy versus enhancement" },
    { "label": "Main modalities", "value": "Pharmacological, genetic, cybernetic" },
    { "label": "Leading proponents", "value": "Harris, Savulescu, Bostrom", "link": "/wiki/julian-savulescu" },
    { "label": "Leading critics", "value": "Kass, Sandel, Habermas", "link": "/wiki/leon-kass" },
    { "label": "Status", "value": "Partly routine, largely disputed" }
  ]
}
```

**Human enhancement** is the deliberate use of biomedical, pharmacological or technical means to raise a human capacity above the level required for ordinary health, rather than to restore a capacity lost to disease or injury. The category is defined by contrast: an intervention counts as enhancement when it is not therapy. That contrast turns out to be hard to draw cleanly, which is why most of the philosophical work on the subject is about the boundary itself rather than about any particular technique.

## The therapy–enhancement boundary

The standard way of drawing the line appeals to species-typical functioning. On Christopher Boorse's biostatistical account, health is the statistically normal contribution of each organ and process to survival and reproduction for an organism of a given age and sex; disease is subnormal functioning, and medicine's job is to restore the norm.[^boorse1977] Norman Daniels built a theory of just health care on a similar footing: society owes its members treatment that protects the normal opportunity range, and enhancement beyond it is a private matter.

The account has clear uses. It explains why insurers pay for insulin and not for a cosmetic procedure, and why growth hormone for a child with a pituitary tumour is treated differently from growth hormone for a healthy short child. It also fails in predictable places. Vaccination raises immune competence above the untreated norm and is universally classed as medicine. Contraception disables a functioning system. Reading glasses correct a deficit that is statistically normal after 45. Caffeine is a cognitive enhancer taken daily by most adults in industrialised countries and is not regulated as one.

```compare
{
  "columns": ["Therapy", "Enhancement"],
  "rows": [
    { "label": "Reference point", "values": ["Species-typical function", "Whatever the person wants"] },
    { "label": "Typical justification", "values": ["Restoring normal opportunity", "Autonomy, welfare, competitive gain"] },
    { "label": "Who usually pays", "values": ["Insurer or state", "The individual"] },
    { "label": "Regulatory pathway", "values": ["Approval against a disease indication", "Off-label, supplement, or unregulated"] },
    { "label": "Evidence usually available", "values": ["Randomised trials in patients", "Small studies in healthy volunteers"] }
  ]
}
```

Two repairs are commonly proposed. One replaces the species norm with the individual's own prior baseline, which handles spectacles and vaccination but makes enhancement depend on when the measurement is taken and cannot classify anything done before birth. The other abandons the boundary in favour of a direct welfare test: an intervention is justified if it makes the person's life go better, whatever the statistical norm. That test is what proponents of enhancement generally endorse, and its weakness is measurement, since well-being over a lifetime is exactly the quantity no trial is powered to detect.

Buchanan, Brock, Daniels and Wikler concluded in *From Chance to Choice* that the distinction cannot bear the moral weight often placed on it, but that it remains useful as a rough guide to what a health system is obliged to provide.[^bdw2000] That is roughly the mainstream position twenty-five years later: the line is administratively real and philosophically porous.

> [!note] Terminology
> "Enhancement" is used in at least three ways in the literature: exceeding species-typical function, exceeding an individual's own prior baseline, and improving a capacity by any means including training. Arguments frequently trade on the ambiguity, since almost nobody objects to the third sense.

## Modalities

Enhancement is not one technology but a set of routes that differ in reversibility, cost and who bears the risk.

**Pharmacological.** Stimulants, wakefulness-promoting agents and putative cognitive drugs are the most widely used route, discussed in [[nootropics]]. A 2008 informal poll of *Nature* readers found that about one in five respondents reported using prescription drugs for non-medical cognitive purposes, a figure often quoted and rarely replicated in representative samples.[^maher2008] Serotonergic psychedelics are often counted here as well, but the trials collected under [[psychedelic-therapy]] enrol patients with a psychiatric diagnosis rather than healthy volunteers, so they establish nothing about enhancement. Anabolic agents and blood-oxygen manipulation dominate [[enhancement-in-sport]].

**Genetic.** Selection is available now and editing is not. [[polygenic-embryo-screening]] ranks existing embryos on statistical predictors; [[germline-editing]] would introduce heritable changes and remains prohibited for clinical use in most jurisdictions. The limits of both for complex traits are set out in [[genetic-enhancement-of-intelligence]] and [[designer-babies]]: the traits people most want to enhance are influenced by thousands of variants of tiny effect, which caps what selection can deliver and makes editing impractical.

**Cybernetic and sensory.** [[brain-computer-interface|Brain–computer interfaces]] are being developed as assistive devices for paralysis, with enhancement of healthy users a stated long-term aim of some companies and a distant one on current bandwidth. [[sensory-augmentation]] adds channels rather than restoring them, and [[non-invasive-neuromodulation]] is the modality with the largest gap between consumer marketing and replicated evidence.

**Physical and metabolic.** [[myostatin-inhibition]], [[exoskeleton|powered exoskeletons]], and interventions aimed at [[healthspan]] rather than disease sit at the boundary where geroscience meets enhancement, since delaying [[hallmarks-of-aging|biological aging]] in a healthy person is by definition not treating a diagnosed illness. Use of [[glp-1-receptor-agonists|incretin drugs]] by people at or near typical weight is the current mass-market instance of the same boundary.

**Moral and affective.** [[moral-enhancement]] proposes to alter dispositions rather than capacities, and raises objections that do not arise for the others.

## The case for

The affirmative arguments come in three broad families.

The welfarist argument holds that if an intervention makes a life go better and harms no one else, the burden of proof lies with those who would prohibit it. John Harris pushes this to an obligation: where enhancement is safe and effective, declining it is not neutral but a failure to make things better.[^harris2007] [[julian-savulescu]] extends the same logic to reproduction under the heading of [[procreative-beneficence]].

The continuity argument observes that literacy, numeracy, vaccination, spectacles and caffeine are all enhancements by any consistent definition, and are unobjectionable. If the biomedical versions differ, the difference must be located in something other than the fact of enhancement.

The evolutionary-slack argument, developed by [[nick-bostrom]] and [[anders-sandberg]], answers the reply that natural selection has already optimised human biology. Selection optimises for inclusive fitness in ancestral environments, not for welfare in modern ones, and it faces constraints that engineers do not. Their "evolutionary heuristic" specifies when to expect exploitable slack: where the environment has changed, where a trait was costly in ancestral conditions but is cheap now, and where evolution faced a tradeoff that technology can relax.[^bs2009]

## The case against

The strongest objections are not about safety.

**Giftedness.** Michael Sandel argues that the drive to enhance expresses a refusal to accept the unbidden, and that a world in which children's traits are chosen erodes the humility and solidarity that follow from recognising talent as unearned luck.[^sandel2007] The argument is not that enhancement harms the enhanced but that it corrodes social attitudes toward those who are not.

**Dignity and the given.** The President's Council on Bioethics under [[leon-kass]] framed the concern as a set of questions about whether achievements obtained pharmacologically are the agent's own, and whether a life without limits retains the shape that makes human goods intelligible.[^pcbe2003] Jürgen Habermas argues separately that a person whose genome was designed by another stands in an asymmetric relation to that designer, compromising the sense of being the sole author of one's life.

**Authenticity and complicity.** If a capacity is bought rather than developed, some of what made it valuable may be lost. Critics note that this objection proves too much when applied to education or to coffee, and defenders reply that it applies with different force to interventions that bypass effort entirely.

The bioconservative case is set out at length in [[bioconservatism]] and the argumentative structure of the whole debate in [[bioethics-of-enhancement]]. Two points are worth separating from the rhetoric. First, several bioconservative predictions about safety have held up: the biology has been harder than proponents expected, and the most confident enhancement claims of the 2000s have not been realised. Second, the disability-rights critique in [[disability-rights-and-enhancement]] is not a version of the giftedness argument; it concerns what a society communicates when it treats certain ways of being human as defects to be engineered out.

## Fairness, coercion and collective effects

The distributional worry is straightforward. Expensive interventions reach the wealthy first, and if they compound across a lifetime or across generations, existing inequality becomes biological. [[access-and-inequality]] examines how strong this effect is likely to be against the historical record of expensive medicine, which has usually diffused, and against gene therapies priced above a million dollars per dose, which so far have not.

The coercion worry is structurally different and, in the view of many philosophers, more serious. Where a capacity is positional, its value depends on others not having it. Once enough competitors take a wakefulness drug, declining it becomes a cost rather than a neutral choice, and formal voluntariness does not describe the situation. This is the collective-action problem analysed in [[enhancement-arms-race]] and the reason anti-doping regimes exist at all despite their expense and their intrusions.

> [!debate] Where the disagreement actually lies
> Few participants in this debate hold that enhancement is intrinsically wrong or that it is unconditionally permissible. The live disputes are about who bears the risk, whether consent is meaningful under competitive pressure, and whether the state should fund, permit, discourage or prohibit each modality separately.

## Practice and regulation

No jurisdiction regulates enhancement as a category. Drugs are approved against disease indications and then used off-label; devices are cleared for clinical purposes and marketed to consumers under wellness exemptions; [[dietary-supplements|supplements]] marketed for cognition face only post-market enforcement. The practical consequence is that the safety evidence for enhancement use is generated mostly in patients who are not the people taking it, and that risk information for healthy users is thin.

Three areas have developed dedicated rules. Sport prohibits a defined list of substances and methods and grants therapeutic-use exemptions, making it the one domain that draws an operational therapy–enhancement line and litigates it. Military programmes in several states fund performance sustainment research under separate ethical review. And neural data has begun to attract specific law, discussed in [[neurorights]] and [[mental-privacy]], because neural interfaces intended as assistive devices generate information about people that no prior consumer category anticipated.

## Open problems

The empirical bottleneck is that the enhancement effects demonstrated so far are small, heterogeneous and measured on laboratory tasks whose relation to performance outside the laboratory is weak. Almost every well-controlled study of a cognitive enhancer in healthy adults reports effect sizes that would be unremarkable in a study of sleep or exercise, and often larger gains in people starting from a lower baseline than in high performers, which inverts the usual assumption about who benefits.

The conceptual bottleneck is that the therapy–enhancement distinction does most of the administrative work in medicine and almost none of the moral work, yet no replacement has been agreed. Proposals to substitute a welfare criterion founder on measurement; proposals to substitute a species-norm criterion founder on the cases above. What follows if the distinction is abandoned entirely, and whether health systems can function without it, has not been worked out by either side.

## See also

- [[bioethics-of-enhancement]]
- [[transhumanism]]
- [[nootropics]]
- [[enhancement-in-sport]]
- [[morphological-freedom]]
- [[enhancement-arms-race]]
- [[disability-rights-and-enhancement]]
- [[posthuman]]

## References

[^boorse1977]: `paper` Boorse, C. "Health as a Theoretical Concept." *Philosophy of Science*, 1977.
[^bdw2000]: `book` Buchanan, A., Brock, D. W., Daniels, N. and Wikler, D. *From Chance to Choice: Genetics and Justice*. Cambridge University Press, 2000.
[^maher2008]: `paper` Maher, B. "Poll results: look who's doping." *Nature*, 2008. {A self-selected online poll of Nature readers rather than a representative survey, so the one-in-five figure has no sampling frame behind it.}
[^harris2007]: `book` Harris, J. *Enhancing Evolution: The Ethical Case for Making Better People*. Princeton University Press, 2007.
[^bs2009]: `book` Bostrom, N. and Sandberg, A. "The Wisdom of Nature: An Evolutionary Heuristic for Human Enhancement." In Savulescu, J. and Bostrom, N. (eds), *Human Enhancement*. Oxford University Press, 2009.
[^sandel2007]: `book` Sandel, M. J. *The Case Against Perfection: Ethics in the Age of Genetic Engineering*. Harvard University Press, 2007.
[^pcbe2003]: `report` President's Council on Bioethics. *Beyond Therapy: Biotechnology and the Pursuit of Happiness*. Washington, DC, 2003. {An advisory report from a body appointed by the US president and chaired by Leon Kass; it argues a philosophical case rather than reviewing evidence.}

==============================================================================
ARTICLE: germline-editing
TITLE:   Human germline editing
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/germline-editing
SOURCE:  https://futurehumanwiki.com/raw/germline-editing
==============================================================================

---
title: "Human germline editing"
slug: "germline-editing"
type: "technology"
status: "contested"
horizon: "2040s"
trl: 4
categories: ["genetics", "society", "reproduction"]
tags: ["crispr", "embryos", "heritable editing", "bioethics", "governance", "reproduction"]
summary: "Modification of human embryos, eggs, or sperm so that the genetic change is inherited by the resulting child and by all of their descendants."
updated: "2026-07-27"
humanEvidence: "One unsanctioned clinic case produced children from CCR5-edited embryos in 2018 and 2019; all other human work is in research embryos cultured for days, where editing is frequently mosaic and can delete or lose the target chromosome."
access: "Obtainable through no lawful route: transfer of an edited embryo to a womb is forbidden wherever the question is regulated, and research use is confined to licensed embryo work."
reversibility: "irreversible"
issues: ["The 2025 private-venture reports and the joint moratorium call are stated without a citation."]
---

```infobox
{
  "caption": "Contested biomedical technology",
  "rows": [
    { "label": "Also called", "value": "Heritable human genome editing" },
    { "label": "Main tools", "value": "CRISPR-Cas9, base editing", "link": "/wiki/crispr-cas9" },
    { "label": "Clinical uses", "value": "One unsanctioned case, 2018", "link": "/wiki/he-jiankui-affair" },
    { "label": "Legal status", "value": "Prohibited in most jurisdictions" },
    { "label": "US position", "value": "FDA barred from reviewing applications" },
    { "label": "Central technical obstacle", "value": "Mosaicism and unverifiable embryos" },
    { "label": "Readiness", "value": "TRL 4 (research embryos only)" }
  ]
}
```

**Human germline editing** is the deliberate alteration of DNA in gametes, zygotes, or early embryos such that the change is carried by every cell of the resulting person and passed to their children. It is the practice that separates genetic medicine from genetic inheritance. Almost every scientific body that has examined it has concluded that clinical use is not currently justified, and it has been attempted in a clinic exactly once, in 2018, by a researcher subsequently imprisoned in China.

## Definition and scope

The germ line is the cell lineage that produces eggs and sperm. Edits made anywhere else — in liver, muscle, blood, retina — die with the patient, which is why [[somatic-gene-therapy]] is regulated as ordinary medicine and heritable editing is not regulated as medicine at all in most countries but prohibited outright.

The boundary is less crisp than the vocabulary suggests. Editing spermatogonial stem cells in an adult man is technically somatic but heritable in effect. [[mitochondrial-replacement-therapy]], legal in the United Kingdom since 2015 and used there in births reported in 2025, alters the mitochondrial genome transmitted down the maternal line; it is heritable, yet it was legislated as an exception rather than treated as germline modification, on the argument that it swaps an organelle rather than editing nuclear DNA. Whether that distinction is principled or convenient remains disputed.

Two further clarifications matter. Germline editing is not embryo selection. Selection chooses among embryos that already exist and cannot introduce a variant that no embryo has; [[embryo-selection]] and [[polygenic-embryo-screening]] work entirely within the range of what a given couple's gametes happen to produce. Editing writes something new. And germline editing is not cloning: [[human-cloning]] copies a genome, whereas editing changes one.

> [!note] Terminology
> Scientific bodies increasingly prefer "heritable human genome editing" to "germline editing", because the term names the consequence — inheritance by descendants — rather than the tissue, and so covers edge cases like gamete precursor cells and mitochondrial transfer.

## How it would be done

The procedure would sit inside a standard IVF cycle. Eggs are retrieved and fertilised; the editing reagents — most plausibly a [[crispr-cas9]] ribonucleoprotein, a [[base-editing]] complex, or a [[prime-editing]] system, the last two avoiding the double-strand break that a nuclease necessarily makes — are introduced into the zygote by microinjection or electroporation, ideally at or before fertilisation so that editing precedes the first cell division. The embryo is cultured for several days, a few trophectoderm cells are biopsied and sequenced, and an embryo judged correctly edited is transferred.

Every step of that description contains an unsolved problem, and the biopsy step contains the deepest one.

An alternative entry point would avoid several of them. Editing gamete precursor cells — spermatogonial stem cells in a testis, or oocyte precursors derived in culture — allows the edited cell population to be expanded, sequenced exhaustively, and selected before any embryo exists. Nothing would then be mosaic, and verification would no longer destroy the thing verified. Human spermatogonial culture is difficult and human oogenesis in vitro has not been achieved, so this route remains theoretical; it is nonetheless the version of heritable editing that has the fewest intrinsic technical objections, and it is the one that [[in-vitro-gametogenesis]] research would make possible if it succeeds.

## Technical obstacles

### Mosaicism

Editing reagents remain active after the zygote divides. If the first cut occurs at the two- or four-cell stage, the resulting embryo is a mixture of edited, unedited, and differently edited cells. A biopsy samples five to ten cells from the trophectoderm — tissue that becomes placenta, not fetus — and so provides a sample of a mosaic, not a description of the child. Injecting reagents earlier reduces but does not abolish the problem.

### On-target damage

Attention initially focused on [[crispr-off-target-effects|off-target cuts]] at similar sequences elsewhere in the genome. The more serious finding, reported in human embryos in 2020 and 2021, was damage at the intended site: large deletions, loss of heterozygosity across whole chromosome arms, and outright loss of the targeted chromosome.[^zuccaro2020][^alanis2021] Standard short-read sequencing of a biopsy can miss all of these, because a deleted allele simply fails to amplify and reads as homozygous for the remaining one.

### Verification is destructive

This is the structural obstacle, not merely a current limitation. The only way to characterise an embryo's genome completely is to disaggregate and sequence it, which destroys it. Any transferred embryo is therefore certified by inference from a biopsy of different cells. In somatic therapy an edited cell product can be sequenced exhaustively before infusion; the germline case forecloses that option permanently.

### The animal evidence is not reassuring

Editing has been performed in non-human primate embryos since 2014, when Cas9-modified cynomolgus monkeys were born carrying targeted mutations.[^niu2014] The offspring were mosaic. A decade of subsequent primate work has not produced a protocol that reliably yields uniformly edited, chromosomally intact animals, and the primate colonies involved are small enough that rare adverse outcomes would be hard to detect. Any claim that embryo editing is nearly ready for human use has to explain why the species closest to humans has not yet demonstrated it.

### The claim that has already collapsed once

A 2017 report claimed that human embryos repaired an inherited cardiomyopathy mutation by copying the healthy maternal allele, which if true would have offered a template-free route to correction.[^ma2017] Other groups argued the apparent correction was more likely allele dropout — the mutant allele deleted and therefore invisible — and the interpretation was contested in the literature.[^egli2018] The episode is a standing reminder that embryo genotyping is unusually easy to get wrong.

> [!caution] Contested
> No published work demonstrates accurate, complete, non-mosaic editing of a human embryo verified to the standard that clinical use would require. The 2020 international commission treated establishing such a demonstration as a precondition even to begin discussing translation.[^ihgge2020]

## When editing would beat selection

The strongest argument against germline editing is not that it is dangerous but that it is usually unnecessary. For nearly every couple at risk of transmitting a serious monogenic condition, preimplantation genetic testing identifies an unaffected embryo among those already produced. Editing adds risk without adding capability.

The 2020 report of the International Commission convened by the US National Academies and the UK Royal Society identified the residual cases. They are few:

- Both prospective parents homozygous for the same recessive condition — for example both living with sickle cell disease — so that every embryo they can produce is affected.
- One parent homozygous for a dominant condition, a rare situation in disorders such as Huntington's disease.
- Couples for whom testing is possible in principle but who reliably obtain very few embryos, or none unaffected, across repeated cycles.

The number of couples worldwide in these categories is small, and each of them already has alternatives: donor gametes, adoption, or accepting the risk and treating the child if it materialises. Whether those alternatives count as acceptable substitutes for a genetically related unaffected child is exactly what the argument turns on. Two developments would shrink the residual set further. In vitro gametogenesis, if it worked in humans, would generate large numbers of embryos and make selection far more powerful, and treatments delivered after birth are advancing: conditions once regarded as untreatable, sickle cell disease among them, now have approved therapies such as [[casgevy]].

## Governance and law

```timeline
[
  { "year": "2015", "title": "First CRISPR in human embryos", "text": "A Chinese group reports editing non-viable tripronuclear zygotes, finding low efficiency, extensive mosaicism, and off-target cuts." },
  { "year": "2015", "title": "First International Summit", "text": "Meeting in Washington, DC, the organising committee calls clinical germline use irresponsible until safety is resolved and there is broad societal consensus." },
  { "year": "2016", "title": "First research licence", "text": "The UK Human Fertilisation and Embryology Authority licenses Kathy Niakan's laboratory to edit human embryos for research into early development." },
  { "year": "2018", "title": "Edited twins", "text": "The birth of twin girls with edited CCR5 genes is announced in Shenzhen days before the Second Summit convenes in Hong Kong." },
  { "year": "2020", "title": "Commission report", "text": "An international commission convened by the US National Academies and the Royal Society defines preconditions and a narrow set of possible first uses." },
  { "year": "2021", "title": "Framework and fresh warnings", "text": "The WHO publishes a governance framework and recommendations for genome editing; following a 2020 report of chromosome loss, another group reports frequent loss of heterozygosity in edited human embryos." },
  { "year": "2023", "title": "Third Summit", "text": "Meeting in London, the summit concludes that heritable editing remains unacceptable and turns much of its attention to the unequal reach of somatic therapies." },
  { "year": "2025", "title": "Private ventures", "text": "Reports of privately funded US companies pursuing heritable editing prompt professional societies to call jointly for an extended moratorium on clinical use." }
]
```

Wherever the question is regulated at all, the law splits at the moment of transfer. Editing embryos in a dish is permitted, under licence or under guidelines, in the United Kingdom, China, Japan, and the United States; placing an edited embryo in a uterus is what is forbidden. Research embryos are separately bound by the fourteen-day culture limit that most jurisdictions and professional guidelines impose, which ends every experiment long before any developmental consequence of an edit could be seen. The evidence base is therefore capped by law at the first fortnight of development, for a technology whose consequences would play out over a lifetime and beyond it.

The instruments doing the forbidding differ in strength, and that variation matters more than the headline count of countries. Some are criminal statutes with stated penalties. Others are clauses in assisted-reproduction acts drafted before editing existed, ministry guidelines with no sanction attached, or — as in the United States — an appropriations rider that bars the Food and Drug Administration from acknowledging an application involving a heritable modification, closing the licensing route without creating any offence. A country can appear on a list of jurisdictions that prohibit heritable editing while having nothing a prosecutor could actually charge.[^baylis2020] The instrument-by-instrument picture, and the treaties that sit above it, are set out in [[governance-of-genome-editing]].

Professional self-governance has done the rest of the work: the 2015 commentary in which [[jennifer-doudna]] and co-authors urged that clinical germline use be discouraged while the research questions were settled,[^baltimore2015] then the three international summits, the academies' reports, and the World Health Organization's 2021 recommendations, which ask member states to treat clinical germline use as prohibited and establish a registry for genome-editing research.[^who2021] Its instruments are soft: statements, licensing conditions, funder policies, journal refusals. None carries enforcement power. On the one occasion the line was crossed, every consequence came from a Chinese criminal court applying a statute about practising medicine without a licence, which had nothing to do with genetics.

> [!debate] The disagreement about pathways
> One camp holds that defining a translational pathway is the responsible response, because a route that is legal, licensed, and monitored is safer than a prohibition that pushes work into unregulated clinics. The other holds that specifying a pathway is itself a form of endorsement that makes eventual use likelier, and points to the [[asilomar-conference]] precedent, where a self-imposed moratorium was lifted once containment measures were agreed.

## Enhancement and outlook

### The slope argument

Nothing in the technique distinguishes correcting a disease allele from installing an advantageous one. The 2018 case is the empirical data point: the twins had no disease, and the edit was intended to confer resistance to an infection they might never encounter — prophylaxis or enhancement, not treatment. Critics of a therapeutic pathway argue that the same clinics, techniques, and consent forms would carry enhancement requests, and that the demand exists whether or not the science supports it.

The science mostly does not. Variants with large, well-understood, unambiguously beneficial effects are rare; most traits people would want to change are influenced by thousands of variants of tiny effect, so [[genetic-enhancement-of-intelligence]] by editing is not a near-term prospect regardless of legality. Even apparently simple protective variants carry pleiotropic costs: disrupting CCR5 confers partial HIV resistance and also increases susceptibility to West Nile virus. The realistic near-term enhancement worry is not superhuman children but confident intervention on the basis of incomplete genetics, an error mode that [[designer-babies]] discussions frequently understate and that ethicists working on [[procreative-beneficence]] and [[disability-rights-and-enhancement]] have examined from opposite directions.

### What would have to change

For heritable editing to become defensible, several things would each have to be true: editing in embryos would need to be demonstrably complete and non-mosaic; verification would need a method that does not destroy what it certifies; the multigenerational consequences would need an evidence base that by definition takes generations to build; and there would need to be a category of prospective parents for whom no alternative route to a healthy genetically related child exists. None of these is close as of 2026, and the last is shrinking rather than growing.

The pressure is nevertheless increasing from an unexpected direction. Press reports in 2025 described privately funded ventures in the United States openly pursuing heritable editing outside the academic system, prompting professional societies in gene therapy and reproductive medicine to call jointly for an extended moratorium on clinical use. Private capital is not bound by funder policies, journal norms, or the informal disciplines that have governed the field since [[he-jiankui-affair|2018]]. The open question is therefore less whether the science becomes ready than whether the institutions that have held the line have any purchase on actors who never belonged to them.

## See also

- [[he-jiankui-affair]]
- [[governance-of-genome-editing]]
- [[crispr-cas9]]
- [[embryo-selection]]
- [[polygenic-embryo-screening]]
- [[somatic-gene-therapy]]
- [[designer-babies]]
- [[bioethics-of-enhancement]]

## References

[^zuccaro2020]: `paper` Zuccaro, M.V. et al. "Allele-specific chromosome removal after Cas9 cleavage in human embryos." *Cell*, 2020. {Human research embryos, never transferred; the losses were found only with assays designed to catch allele dropout.}
[^alanis2021]: `paper` Alanis-Lobato, G. et al. "Frequent loss of heterozygosity in CRISPR-Cas9-edited early human embryos." *Proceedings of the National Academy of Sciences*, 2021.
[^ma2017]: `paper` Ma, H. et al. "Correction of a pathogenic gene mutation in human embryos." *Nature*, 2017.
[^egli2018]: `paper` Egli, D. et al. "Inter-homologue repair in fertilized human eggs?" *Nature*, 2018.
[^ihgge2020]: `report` International Commission on the Clinical Use of Human Germline Genome Editing. *Heritable Human Genome Editing*. US National Academy of Medicine, National Academy of Sciences, and the Royal Society, 2020. {Convened by academies rather than by a regulator; it sets preconditions for a possible future use and endorses none.}
[^baylis2020]: `paper` Baylis, F., Darnovsky, M., Hasson, K., Krahn, T.M. "Human Germline and Heritable Genome Editing: The Global Policy Landscape." *The CRISPR Journal*, 2020. {A survey of policy instruments rather than of practice; the instruments it counts range from criminal statutes to guidelines with no sanction attached.}
[^baltimore2015]: `paper` Baltimore, D. et al. "A prudent path forward for genomic engineering and germline gene modification." *Science*, 2015. {A commentary by the participants of a small meeting in Napa, California, urging that clinical germline use be discouraged while research continued; it carried no legal force.}
[^who2021]: `report` World Health Organization. *Human Genome Editing: A Framework for Governance* and *Recommendations*. WHO, 2021. {Recommendations addressed to member states. The WHO cannot prohibit anything, and the research registry it proposes is voluntary.}
[^niu2014]: `paper` Niu, Y. et al. "Generation of Gene-Modified Cynomolgus Monkey via Cas9/RNA-Mediated Gene Targeting in One-Cell Embryos." *Cell*, 2014.

==============================================================================
ARTICLE: human-hibernation
TITLE:   Human hibernation and torpor
PORTAL:  Space & Extreme Environments
URL:     https://futurehumanwiki.com/wiki/human-hibernation
SOURCE:  https://futurehumanwiki.com/raw/human-hibernation
==============================================================================

---
title: "Human hibernation and torpor"
slug: "human-hibernation"
type: "concept"
status: "experimental"
horizon: "2040s"
categories: ["space", "bodies"]
tags: ["torpor", "hypothermia", "metabolism", "spaceflight", "neuroscience", "suspended animation"]
summary: "The proposal to induce a hibernation-like state of suppressed metabolism in people, and the distance between rodent torpor circuits and any clinical human application."
updated: "2026-07-27"
humanEvidence: "No human has been placed in torpor; every induced-torpor result is from mice and rats, and all human cooling to date is hypothermia imposed from outside against a defended set point."
issues: ["The emergency preservation and resuscitation trauma trials are mentioned with no registry or publication source.", "The claim that hibernating rodents tolerate higher radiation doses is attributed only to older literature."]
---

```infobox
{
  "caption": "Physiological state and proposed intervention",
  "rows": [
    { "label": "Natural in humans", "value": "No" },
    { "label": "Closest clinical practice", "value": "Therapeutic hypothermia" },
    { "label": "Demonstrated in", "value": "Mice, rats (induced)" },
    { "label": "Induction routes studied", "value": "Neural, chemical, ultrasound" },
    { "label": "Main proposed use", "value": "Long-duration spaceflight" },
    { "label": "Human torpor achieved", "value": "No" }
  ]
}
```

**Human hibernation and torpor** is the proposal to place people in a state of deliberately suppressed metabolic rate and lowered body temperature, either for medical stabilisation or to reduce the physiological and logistical costs of long spaceflights. No human has ever been put into torpor. Every existing clinical practice that resembles it — cooling after cardiac arrest, deep hypothermic circulatory arrest during aortic surgery — works by lowering temperature from the outside and accepting the metabolic consequences, which is close to the opposite of what a hibernating animal does.

```keyfacts
[
  { "value": "1–5%", "label": "Metabolic rate in deep torpor", "note": "ground squirrels, relative to their own basal rate" },
  { "value": "13.7 °C", "label": "Core temperature survived in a reported case", "note": "accidental hypothermia, rewarmed on extracorporeal circulation" },
  { "value": "0", "label": "Humans placed in true torpor", "note": "all human cooling to date is imposed hypothermia" }
]
```

## What torpor is

Torpor is a regulated reduction in metabolic rate and body temperature, entered and exited under the animal's own control. Hibernation is a seasonal pattern of multi-day torpor bouts separated by brief, energetically expensive arousals back to normal temperature. Daily torpor, seen in some rodents and hummingbirds, is the same mechanism on a shorter cycle.

The key word is *regulated*. In deep torpor, an arctic ground squirrel's core temperature can fall below the freezing point of water and its metabolic rate to a few per cent of basal, with heart rate in the single digits, and it rewarms itself on schedule without external assistance. Metabolic suppression precedes cooling rather than following from it: the animal turns down its own oxygen consumption first, and the temperature drop is a consequence.

Bears make the point differently. Their core temperature falls only a few degrees during a winter denning period, yet metabolic rate drops to roughly a quarter of basal and stays there for months.[^toien2011] Metabolic suppression and hypothermia are therefore separable, which is the single most important fact for anyone hoping to reproduce the state in a large mammal. Bears also emerge with little of the bone and muscle loss that months of immobility produce in humans, a property that makes hibernation biology relevant to [[microgravity-adaptation|disuse deconditioning]] independently of spaceflight.

Hibernators are additionally long-lived for their body size — small hibernating mammals routinely outlive similarly sized non-hibernators — which places them alongside the other comparative-biology outliers surveyed in [[negligible-senescence]]. Whether that reflects the metabolic suppression itself or a correlated set of protective adaptations is unsettled, and the same ambiguity dogs attempts to read lifespan effects out of [[caloric-restriction]] data.
Arrested states in invertebrates carry the same association: the *C. elegans* dauer larva is a
stress-resistant, developmentally arrested stage, and the genes controlling entry into it are the
insulin-signalling genes whose partial loss doubles adult lifespan in
[[cynthia-kenyon|Cynthia Kenyon's]] mutants.

## The closest real practice

Therapeutic hypothermia is the only human intervention that reliably lowers metabolic rate, and it does so bluntly: cooling reduces oxygen consumption by roughly five to seven per cent for each degree Celsius. It is established practice in neonatal hypoxic-ischaemic encephalopathy, where cooling to about 33.5 °C for three days improves neurological outcomes. In adults after cardiac arrest the picture is less settled; the large TTM2 trial found no mortality benefit from targeting 33 °C compared with avoiding fever, and guidelines have shifted toward temperature control rather than deep cooling.[^ttm2]

Cardiac surgery goes further. Deep hypothermic circulatory arrest cools patients to roughly 18 °C, allowing the heart to be stopped and circulation halted for tens of minutes while an aortic arch is repaired. Emergency preservation and resuscitation extends the same logic to exsanguinating trauma: replacing blood with cold saline to drop core temperature toward 10 °C, buying surgical time before resuscitation. Trials of this approach have been conducted in the United States and remain small.

Accidental hypothermia indicates roughly where the outer bound of human tolerance sits. Among the best-documented survivals is a Norwegian skiing accident in which a woman was recovered with a core temperature of 13.7 °C after prolonged cardiac arrest and made a substantial neurological recovery on extracorporeal rewarming.[^gilbert2000] Such cases are rare, unplanned, and not a protocol.

Voluntary cold exposure of the kind covered under [[heat-and-cold-exposure]] runs the other way entirely: shivering and brown-fat thermogenesis raise metabolic rate to defend the set point rather than lowering it.

> [!note] Hypothermia is not hibernation
> A cooled human is a defended system fighting the intervention: shivering, vasoconstriction, and catecholamine release all resist the temperature drop, which is why clinical cooling requires sedation and often neuromuscular blockade. A hibernating squirrel has lowered its own set point. Getting from the first state to the second is the entire scientific problem.

## Circuits that induce torpor

The field changed in 2020 when two groups independently identified neural populations sufficient to induce a torpor-like state in mice. One reported that stimulating a set of neurons expressing the neuropeptide QRFP in the hypothalamus produced days-long hypometabolism and hypothermia, which they termed Q-neuron-induced hypothermia; the state was reversible and left the animals apparently unharmed.[^takahashi2020] The other identified preoptic-area neurons whose activity is necessary and sufficient for natural fasting-induced torpor.[^hrvatin2020]

Both used chemogenetic and [[optogenetics|optogenetic]] tools, which require genetic modification of the target neurons and are not applicable to humans. An implanted electrode in the same hypothalamic region would be the obvious workaround, and would inherit the surgical risk and target-uncertainty problems documented for [[deep-brain-stimulation]] in far better-characterised nuclei. The follow-up that attracted the most attention took the same target non-invasively: focused ultrasound aimed at the preoptic area induced a torpor-like drop in temperature and metabolic rate in mice, and a weaker but detectable response in rats, which do not naturally enter torpor.[^yang2023] The proposed transducer was the mechanosensitive channel TRPM2. As a route to human application this is far more promising than optogenetics, and it sits alongside the other approaches described in [[non-invasive-neuromodulation]].

The rat result matters more than the mouse result. If a species without a natural torpor programme can be pushed toward one, the circuitry may be conserved and merely dormant in large mammals, humans included. If it cannot, engineered torpor would require installing machinery rather than switching it on.

## Chemical routes and dead ends

Pharmacological induction has a longer and less encouraging record. Hydrogen sulfide at low concentrations was reported in 2005 to drop mouse metabolic rate by around ninety per cent with a fall in core temperature, in a state its discoverers described as suspended animation.[^blackstone2005] The finding did not scale: attempts to reproduce comparable metabolic suppression in sheep and pigs failed, and the effect in mice appears to depend partly on their small body size and high surface-to-volume ratio. Reports that 5′-AMP induces torpor were similarly complicated by the difficulty of separating true metabolic suppression from passive cooling.

Adenosine receptor agonists have shown more consistent effects on thermoregulatory set point across species and remain an active line. None of this constitutes a drug that could be given to an astronaut.

## Why humans are hard

Body size works against the whole project. Large animals cool and rewarm slowly, which makes any induction protocol a matter of hours, and slow rewarming is where much of the injury occurs. Below roughly 30 °C the human myocardium becomes prone to ventricular fibrillation, and defibrillation is unreliable at those temperatures. Cold produces coagulopathy, impaired platelet function, cold diuresis and electrolyte derangement, and suppresses immune responses at a time when a sedated, immobile patient is highly exposed to infection.

Then there is everything a torpid human would not be doing. Nutrition would be parenteral over months. Pressure injury, thrombosis, and muscle wasting would all need management. Whether protective mechanisms that spare hibernator muscle and bone would appear in a human simply because their temperature was lowered is doubtful, since those mechanisms look like specific adaptations rather than passive consequences of cold.

## The spaceflight motivation

Torpor has been studied under NASA's advanced concepts programme and by European groups as a way to reduce the mass and volume of a Mars transit habitat. The arithmetic is attractive: a crew consuming a fraction of normal food, water and oxygen needs a smaller pressurised volume and lighter [[closed-loop-life-support|life support]], and shielding mass can be concentrated around a small torpor bay rather than a full habitat. Psychological benefits are also claimed, since a sleeping crew does not experience confinement.

The claims deserve scrutiny. Torpor does nothing about [[radiation-hardening-humans|cosmic radiation]] except in the weak sense that a smaller shielded volume is cheaper to protect, though older literature reporting that hibernating rodents tolerate higher radiation doses has kept the question open. It would not, by itself, prevent the bone and muscle loss documented in [[space-medicine]] unless the hibernator's protective biology came with it. And a torpid crew cannot respond to a failure, which is a serious objection for a vehicle with no rescue option. Nor does torpor address the question of whether people should be adapted to the destination rather than merely delivered to it, which is the argument made in [[pantropy]].

For interstellar durations the concept shades into a different category altogether. Multi-decade or multi-century voyages would demand either genuine suspended animation, the population biology treated in [[generation-ship-biology]], or the preservation-and-revival wager of [[cryonics]] — and torpor, which is metabolism slowed rather than stopped, does not scale to those timescales.

## Outlook

The near-term payoff is medical rather than orbital. If a non-invasive method could reliably lower metabolic demand in a human by even a modest fraction, the applications in stroke, myocardial infarction, traumatic haemorrhage, and organ preservation would be immediate and would ease part of the [[organ-shortage]] problem. That is a lower bar than months of stable torpor, and it is the bar the field is actually working toward.

The unresolved scientific question is whether the torpor programme is conserved and latent in non-hibernating mammals or absent from them. The rat ultrasound result is suggestive, not decisive, and no primate work has been published that settles it.

## See also

- [[space-medicine]]
- [[microgravity-adaptation]]
- [[closed-loop-life-support]]
- [[cryonics]]
- [[generation-ship-biology]]
- [[non-invasive-neuromodulation]]
- [[caloric-restriction]]
- [[radiation-hardening-humans]]

## References

[^takahashi2020]: `paper` Takahashi, T.M. et al. "A discrete neuronal circuit induces a hibernation-like state in rodents." *Nature*, 2020.
[^hrvatin2020]: `paper` Hrvatin, S. et al. "Neurons that regulate mouse torpor." *Nature*, 2020.
[^yang2023]: `paper` Yang, Y. et al. "Induction of a torpor-like hypothermic and hypometabolic state in rodents by ultrasound." *Nature Metabolism*, 2023.
[^blackstone2005]: `paper` Blackstone, E., Morrison, M., Roth, M.B. "H2S induces a suspended animation-like state in mice." *Science*, 2005. {Mice only; later attempts to reproduce comparable metabolic suppression in sheep and pigs did not succeed.}
[^toien2011]: `paper` Tøien, Ø. et al. "Hibernation in black bears: independence of metabolic suppression from body temperature." *Science*, 2011.
[^gilbert2000]: `paper` Gilbert, M., Busund, R., Skagseth, A., Nilsen, P.Å., Solbø, J.P. "Resuscitation from accidental hypothermia of 13.7 °C with circulatory arrest." *The Lancet*, 2000. {A single case report of an accident and a rewarming, which records a survival rather than a reproducible protocol.}
[^ttm2]: `paper` Dankiewicz, J. et al. "Hypothermia versus Normothermia after Out-of-Hospital Cardiac Arrest." *New England Journal of Medicine*, 2021. {The comparison was cooling to 33 degrees against active fever avoidance, not against leaving temperature uncontrolled.}

==============================================================================
ARTICLE: human-ai-merger
TITLE:   Human–AI merger
PORTAL:  Foundational Concepts
URL:     https://futurehumanwiki.com/wiki/human-ai-merger
SOURCE:  https://futurehumanwiki.com/raw/human-ai-merger
==============================================================================

---
title: "Human–AI merger"
slug: "human-ai-merger"
type: "concept"
status: "speculative"
horizon: "2040s"
categories: ["foundations", "cybernetics", "minds"]
tags: ["artificial intelligence", "brain-computer interface", "bandwidth", "identity", "cognition", "extended mind"]
summary: "The proposal that humans and artificial intelligence will form a single cognitive system rather than remaining separate agents that merely use one another."
updated: "2026-07-27"
humanEvidence: "No merged system exists at any scale; the strongest human results are intracortical implants in small numbers of people with paralysis, decoding attempted speech or handwriting at tens of words per minute."
issues: ["The reading, typing and speech rate figures in the bandwidth section carry no source."]
---

```infobox
{
  "caption": "Concept in cognitive science and futures studies",
  "rows": [
    { "label": "Ancestor idea", "value": "Man–computer symbiosis, 1960" },
    { "label": "Main variants", "value": "Centaur, merger" },
    { "label": "Stated motivation for BCIs", "value": "Output bandwidth" },
    { "label": "Best current channel", "value": "Tens of words per minute" },
    { "label": "Key objection", "value": "Bandwidth is not the bottleneck" },
    { "label": "Status", "value": "No demonstrated instance" }
  ]
}
```

**Human–AI merger** is the claim that the relationship between people and artificial intelligence will not remain one of user and tool, but will become close enough that the combined system is the relevant unit of cognition and, on stronger versions, of identity. The idea is invoked most often as the stated rationale for high-bandwidth [[brain-computer-interface|brain–computer interfaces]], and as a proposed alternative to being economically or intellectually displaced by autonomous [[artificial-general-intelligence]]. It is a family of quite different theses, and they are rarely distinguished.

## The two theses

```compare
{
  "columns": ["Centaur thesis", "Merger thesis"],
  "rows": [
    { "label": "Unit of analysis", "values": ["A team of two agents", "One hybrid agent"] },
    { "label": "Interface required", "values": ["Ordinary senses and hands", "Direct neural coupling"] },
    { "label": "Claim type", "values": ["Empirical, testable now", "Speculative, no instance"] },
    { "label": "Identity implication", "values": ["None", "Substantial"] },
    { "label": "Main evidence", "values": ["Human–machine team studies", "Analogy and extrapolation"] }
  ]
}
```

The *centaur thesis* holds that human–machine teams outperform either party alone, and that the future of capability lies in division of labour rather than replacement. It makes claims that can be tested today.

The *merger thesis* is stronger. It holds that the coupling will become tight enough — through direct neural interfaces, continuous prediction of intent, or the gradual delegation of cognitive functions — that drawing a boundary between the person and the system stops being meaningful. This is the version that carries implications for [[personal-identity-and-continuity]], and it has no demonstrated instance at any scale.

## Intellectual history

J. C. R. Licklider set out the framing in 1960, arguing that in the near term computers and humans would occupy complementary roles, with humans setting goals and formulating hypotheses while machines handled routine work, and that this symbiosis was worth pursuing because it would be productive during the interval before machines could think for themselves.[^licklider1960] Licklider was explicit that the arrangement was transitional. Douglas Engelbart's augmentation framework two years later took the same view of the goal and located it in tools, language and method rather than in biology, an approach traced in [[intelligence-amplification]].

The word *cyborg* comes from the same period, coined by Manfred Clynes and Nathan Kline for an organism whose homeostatic systems are extended by machinery so that it can survive in space.[^clynes1960] Their proposal concerned physiology, not cognition, which is a distinction the later literature has largely lost.

[[ray-kurzweil]] gave the merger thesis its popular form, predicting that people will progressively augment their neocortices with cloud-hosted computation and that the distinction between biological and machine cognition will dissolve rather than resolve in favour of one side. The argument is a corollary of his broader position on [[accelerating-change]] rather than an independent claim about neuroscience.

## The bandwidth argument

The most-cited technical motivation for invasive interfaces is asymmetry. Human input bandwidth is reasonable — reading proceeds at roughly 200 to 300 words per minute — while output is slow, at perhaps 40 words per minute by keyboard and 150 by speech. On this argument, human thought is fast but the channel to the outside is narrow, and widening it with a direct cortical connection would let people keep pace with machine reasoning. Elon Musk has made this the public rationale for [[neuralink]], framing a high-bandwidth interface as the alternative to obsolescence.

The argument has a serious empirical problem. Measurements of human behavioural throughput across a wide range of tasks — typing, speech, sight-reading, competitive puzzle solving — converge on something close to ten bits per second, orders of magnitude below the raw information rate of the sensory periphery.[^zheng2025] If the slow step is central rather than peripheral, a wider output channel does not help, because there is nothing faster upstream waiting to use it. The nervous system's apparent bottleneck is serial attention, not the width of the mouth or hands.

> [!debate] Is bandwidth the bottleneck?
> Proponents treat the interface as the limiting factor and the brain as an underused source. The competing view holds that conscious cognition is intrinsically low-rate, in which case a thousand-channel implant would deliver the same ten bits per second with more surgery. No experiment has yet distinguished these, because no interface has come close to saturating even the narrow channel.

## What interfaces actually deliver

Current systems are far from the premise. Intracortical implants using [[utah-array|multielectrode arrays]] have restored communication to people with paralysis at rates of tens of words per minute, achieved by decoding attempted speech or handwriting from motor cortex — an achievement of real clinical significance and roughly an order of magnitude below unimpaired speech.[^willett2023] Endovascular devices such as the [[stentrode]] trade channel count for a much lower surgical burden and deliver correspondingly less. Results in [[speech-neuroprosthesis]] and [[neural-decoding]] have improved steadily, but every one of them decodes *motor intent*, which is the easiest signal in the brain to read because it is already an output code.

Writing information *into* the brain is harder still, and less advanced. Sensory feedback in prosthetics uses small numbers of stimulation channels producing crude percepts; [[deep-brain-stimulation]] modulates circuits without conveying content; [[memory-prosthesis]] work on hippocampal encoding has produced modest effects in small numbers of patients. Nothing in clinical use delivers structured symbolic content to cortex. A merger in the strong sense requires bidirectional, high-rate, semantically precise coupling, and only the first of those three properties exists in any form.

## The tool-use objection

A separate line of argument holds that the merger has already happened and is uninteresting. Andy Clark and David Chalmers's extended-mind thesis contends that when an external resource is reliably available, automatically endorsed and easily accessible, it is part of the cognitive system on the same footing as biological memory.[^clark1998] A person with a phone satisfies those conditions for a large fraction of what they know. Experimental work on cognitive offloading supports the behavioural half of the claim: people who expect information to remain accessible remember the location rather than the content.[^sparrow2011]

If this is right, the merger thesis's distinctive content is not integration but *intimacy of coupling*, and it must explain why a neural connection differs in kind from a pocket-sized one rather than in latency. The strongest available answer is that skull-external channels require conscious, serial mediation, while a neural channel could in principle operate below awareness — which returns the argument to the bandwidth dispute above.

## Evidence from human–machine teams

The centaur thesis has a testable record, and it is mixed. Freestyle chess in the mid-2000s produced the founding anecdote, when teams of amateurs with ordinary computers and good process beat both grandmasters and standalone engines. As engines improved, the human contribution shrank, and by the 2020s few strong players claimed that overriding a leading engine improved its play.

Systematic evidence outside games is less encouraging than the anecdote. A meta-analysis of human–AI collaboration found that on average the combination performed *worse* than the better of human or AI alone on decision tasks, though better on creation tasks — the gains appear where the human contributes generation and judgement, and the losses where the human overrides a more accurate model or fails to catch its errors.[^vaccaro2024] The practical lesson is that complementarity is a property of specific task structures, not a general law, and that centaur advantage erodes precisely as the machine component improves.

## Identity and agency

If the coupling tightened as the merger thesis imagines, three questions follow that current law and ethics do not answer.

*Authorship.* When an interface predicts intent and acts before the person has formed a deliberate decision, attributing the action becomes genuinely unclear. Closed-loop stimulation devices already raise a smaller version of this, and patients have reported altered senses of agency.

*Continuity.* Gradual functional replacement is the scenario used in [[teleportation-problem|thought experiments about gradual replacement]] to argue that substrate change need not break identity, and it is the mechanism by which [[mind-uploading]] is sometimes proposed to happen without a discontinuity. Whether the argument works depends on the [[substrate-independence]] premise, which is assumed more often than defended.

*Privacy and control.* An interface that reads intent generates the most sensitive data category yet created, which is why [[neurorights]] and [[mental-privacy]] have advanced from philosophy into statute faster than the technology has advanced into general use.

## Outlook

The merger thesis is doing two jobs at once, and they should be separated. As an engineering programme it is a claim about interfaces, and on that front the honest position is that decoding has improved substantially, encoding has barely started, and no device approaches the bandwidth its rationale assumes. As a strategic argument — that merging is how humans avoid displacement by [[technological-singularity|autonomous machine intelligence]] — it has never been made rigorously, and it faces an awkward question: if the machine component supplies most of the capability, it is not clear what work the biological component is doing in the hybrid, or why anyone would preserve it for reasons other than the ones that make [[human-enhancement]] contested in the first place.

## See also

- [[brain-computer-interface]]
- [[intelligence-amplification]]
- [[artificial-general-intelligence]]
- [[neuralink]]
- [[mind-uploading]]
- [[neurorights]]
- [[posthuman]]
- [[whole-brain-emulation]]

## References

[^licklider1960]: `paper` Licklider, J. C. R. "Man-Computer Symbiosis." *IRE Transactions on Human Factors in Electronics*, 1960.
[^clynes1960]: `paper` Clynes, M. E. and Kline, N. S. "Cyborgs and Space." *Astronautics*, 1960.
[^zheng2025]: `paper` Zheng, J. and Meister, M. "The Unbearable Slowness of Being: Why Do We Live at 10 Bits/s?" *Neuron*, 2025. {The ten-bit figure is an estimate aggregated across behavioural tasks, not a measurement of any neural channel.}
[^willett2023]: `paper` Willett, F. R. et al. "A high-performance speech neuroprosthesis." *Nature*, 2023. {A single participant with ALS and implanted electrode arrays; the rate is for attempted speech decoded from motor cortex, not for a healthy user.}
[^clark1998]: `paper` Clark, A. and Chalmers, D. "The Extended Mind." *Analysis*, 1998.
[^sparrow2011]: `paper` Sparrow, B., Liu, J. and Wegner, D. M. "Google Effects on Memory: Cognitive Consequences of Having Information at Our Fingertips." *Science*, 2011.
[^vaccaro2024]: `paper` Vaccaro, M., Almaatouq, A. and Malone, T. "When combinations of humans and AI are useful: A systematic review and meta-analysis." *Nature Human Behaviour*, 2024. {Pools experiments across very different tasks; the decision-versus-creation contrast is a subgroup finding within that pooled set.}

==============================================================================
ARTICLE: humanity-plus
TITLE:   Humanity+
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/humanity-plus
SOURCE:  https://futurehumanwiki.com/raw/humanity-plus
==============================================================================

---
title: "Humanity+"
slug: "humanity-plus"
type: "organization"
status: "established"
horizon: "present"
categories: ["organizations", "people"]
tags: ["transhumanism", "movements", "advocacy", "bioethics", "nonprofit", "intellectual history"]
summary: "The main international transhumanist membership organization, founded in 1998 as the World Transhumanist Association and now largely dormant as an institution."
updated: "2026-07-27"
issues: ["The current-activity assessment cites no source and will date quickly.", "Nothing on funding or membership size, which an organization article should cover."]
---

```infobox
{
  "caption": "Nonprofit advocacy organization",
  "rows": [
    { "label": "Founded", "value": "1998, as the World Transhumanist Association" },
    { "label": "Founders", "value": "Nick Bostrom, David Pearce", "link": "/wiki/nick-bostrom" },
    { "label": "Renamed", "value": "2008" },
    { "label": "Core document", "value": "The Transhumanist Declaration" },
    { "label": "Publications", "value": "Journal of Evolution and Technology; H+ Magazine" },
    { "label": "Notable chairs", "value": "James Hughes, Natasha Vita-More", "link": "/wiki/natasha-vita-more" },
    { "label": "Current activity", "value": "Minimal" }
  ]
}
```

**Humanity+** is an international nonprofit membership organization founded in 1998 as the World Transhumanist Association to advocate for the ethical use of technology to extend human capacities. It was created as a deliberately academic and politically non-aligned alternative to [[extropianism]], and for roughly a decade it was the institutional centre of [[transhumanism]]. Its influence has since dispersed into specialized fields — AI safety, longevity biotech, bioethics — and the organization itself now operates at a low level.

## Overview

The WTA's founding purpose was to make transhumanism defensible in an academic seminar. Where the extropians had a libertarian politics, a self-consciously informal style and no peer-reviewed output, the new body offered a journal, a declaration written in the register of a human-rights document, and a leadership drawn from philosophy departments. It positioned transhumanism as a normative position about technology and human welfare rather than as a subculture.

That repositioning worked at the level of ideas and less well at the level of the institution. The arguments the WTA promoted are now standard reference points in bioethics syllabi, while the organization that promoted them has few staff and little visible programme.

## History

[[nick-bostrom]], then a philosophy doctoral student, and David Pearce, author of the online manifesto *The Hedonistic Imperative*, founded the World Transhumanist Association in 1998. In the same period they issued the Transhumanist Declaration and launched what became the *Journal of Evolution and Technology*, the field's first peer-reviewed venue. Bostrom's own account traces the movement's lineage back through [[fm-2030]] and Julian Huxley to the Enlightenment, a genealogy that served the organization's purpose of establishing respectability as much as it did the historical record.[^bostrom2005b]

James Hughes, a sociologist and bioethicist, served as executive director in the mid-2000s and pushed the organization toward what he called democratic transhumanism: the position that enhancement technologies should be publicly regulated and universally accessible, and that a libertarian transhumanism would produce exactly the stratified outcome its critics predicted.[^hughes2004] This produced sustained internal conflict with the organization's libertarian membership, and Hughes left the WTA in the late 2000s. He and Bostrom had already founded a separate body, the Institute for Ethics and Emerging Technologies, in 2004.

The renaming to Humanity+ in 2008 was accompanied by *H+ Magazine*, a consumer-facing publication, and by a series of conferences. [[natasha-vita-more]] chaired the organization and later served as executive director, bringing the design and arts strand of transhumanism into its official identity.

```timeline
[
  { "year": "1998", "title": "World Transhumanist Association founded", "text": "Nick Bostrom and David Pearce establish a membership organization and publish the first Transhumanist Declaration." },
  { "year": "1998–1999", "title": "A journal", "text": "The Journal of Transhumanism, later the Journal of Evolution and Technology, gives the movement its first peer-reviewed outlet." },
  { "year": "2004", "title": "The democratic turn", "text": "James Hughes's Citizen Cyborg argues for regulated, universally accessible enhancement, opening a lasting split with the libertarian membership." },
  { "year": "2008", "title": "Rebranding", "text": "The WTA becomes Humanity+ and launches H+ Magazine for a general readership." },
  { "year": "2009", "title": "Declaration revised", "text": "A rewritten Transhumanist Declaration is adopted, emphasizing risk, access and the moral standing of non-human minds." },
  { "year": "2010s", "title": "Dispersal", "text": "The movement's intellectual energy migrates to AI risk research, longevity companies and academic bioethics; the organization's activity declines." }
]
```

## The Transhumanist Declaration

The Declaration, first issued in 1998 and substantially rewritten in 2009, is the closest thing the movement has to a constitutional text. Its content is more cautious than the movement's reputation suggests. It asserts that humanity will be radically affected by technology and that the possibility of broadening human potential should be taken seriously; it also insists on the need to address catastrophic risks, on inclusion of the widest possible range of beneficiaries, and on the moral consideration of all sentient beings, whether human, animal, or artificial.

Three of its commitments are load-bearing for later disputes:

- **Reproductive and morphological autonomy.** Individuals should have wide latitude over their own bodies and reproduction, the position developed at length under [[morphological-freedom]].
- **Risk seriousness.** Transhumanism is presented as compatible with, not opposed to, work on [[existential-risk]]; the Declaration predates Bostrom's central papers on the subject and anticipates them.
- **Universal access.** Enhancement should not be confined to the wealthy, a commitment that the movement's critics regard as sincere but unmatched by any mechanism, and which is treated at length under [[access-and-inequality]].

> [!note] What the Declaration does not say
> It contains no timeline, no endorsement of any specific technology, and no claim that any particular enhancement will work. Much of the criticism directed at transhumanism targets forecasts made by individual proponents such as [[ray-kurzweil]] rather than positions the organization has adopted.

## Programme and publications

The organization's substantive output was mostly textual. The *Journal of Evolution and Technology* published work on [[whole-brain-emulation]], [[personal-identity-and-continuity|personal identity]], enhancement ethics, and long-range forecasting, and gave academic authors a citable venue at a time when mainstream journals treated the topics as unserious. Bostrom's widely cited defence of posthuman dignity, written against Francis Fukuyama's arguments, came out of this period.[^bostrom2005]

*H+ Magazine* ran from 2008 into the 2010s under editor R. U. Sirius, covering the same subjects in a popular register. Humanity+ also ran conferences in the United States and Europe, and maintained national and regional chapters of varying activity, several of which outlived the parent body's momentum.

The movement's principal anthology, *The Transhumanist Reader*, was edited by [[max-more]] and Vita-More and published in 2013.[^reader2013] It collects the founding documents of both the extropian and WTA traditions, and functions in practice as the movement's canon.

## Reception

Academic reception has been mixed and largely independent of the organization. Bostrom's and Hughes's arguments were engaged seriously by bioethicists including [[julian-savulescu]] on the supportive side and Leon Kass, Michael Sandel and Fukuyama on the critical side, and the resulting exchange is the substance of [[bioethics-of-enhancement]] as a subfield. Whether Humanity+ as an institution contributed much to that exchange beyond providing a venue is debatable.

The most durable criticism of the organization is that it never resolved the tension Hughes identified. A movement whose declared commitment is universal access, and whose membership is drawn heavily from people who can afford private enhancement, has an unmet burden of explanation. [[bioconservatism|Bioconservative]] critics pressed a different charge: that the Declaration's careful language concealed a substantive view of what humans should become, and that the movement's tolerance for the [[posthuman]] as an end-state is not neutral between conceptions of the good.

## Current activity

As of 2026 Humanity+ maintains a web presence and nominal membership but has no significant research programme, and its conferences and publications have largely stopped. This is not a scandal so much as a succession. The specific concerns that motivated the WTA are now pursued by organizations with money and staff: AI risk by dedicated research institutes, life extension by companies such as [[altos-labs]] and [[retro-biosciences]] and by prize bodies such as [[xprize-healthspan]], neural interface ethics by the [[neurorights]] campaign. The umbrella organization proved less durable than the topics it assembled.

The open question is whether transhumanism still needs a general institution. The arguments have professionalized and separated; a bioethicist working on [[germline-editing]] governance and an engineer working on [[brain-computer-interface]] bandwidth now have little occasion to meet under a shared banner. Whether anything is lost when that connective tissue disappears is unresolved, and the case that something is lost rests on the observation that the field's most consequential questions — about identity, access and what is worth becoming — belong to none of the specialties individually.

## See also

- [[transhumanism]]
- [[extropianism]]
- [[nick-bostrom]]
- [[natasha-vita-more]]
- [[max-more]]
- [[bioethics-of-enhancement]]
- [[posthuman]]
- [[morphological-freedom]]

## References

[^hughes2004]: `book` Hughes, J. *Citizen Cyborg: Why Democratic Societies Must Respond to the Redesigned Human of the Future*. Westview Press, 2004. {An argument for one faction's position by a figure who led the organization, not a neutral account of it.}
[^bostrom2005]: `paper` Bostrom, N. "In Defense of Posthuman Dignity." *Bioethics*, 2005.
[^reader2013]: `book` More, M. and Vita-More, N. (eds.) *The Transhumanist Reader*. Wiley-Blackwell, 2013.
[^bostrom2005b]: `paper` Bostrom, N. "A History of Transhumanist Thought." *Journal of Evolution and Technology*, 2005. {A history of the movement written by one of its founders and published in the journal the organization itself launched.}

==============================================================================
ARTICLE: immunosenescence
TITLE:   Immunosenescence
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/immunosenescence
SOURCE:  https://futurehumanwiki.com/raw/immunosenescence
==============================================================================

---
title: "Immunosenescence"
slug: "immunosenescence"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity", "bodies"]
tags: ["aging", "immunity", "thymus", "t cells", "vaccines", "mechanisms", "cytomegalovirus"]
summary: "The age-related remodelling of the immune system, in which responses to unfamiliar antigens weaken while chronic activation against familiar ones persists."
updated: "2026-07-28"
humanEvidence: "Thymic involution, naive T-cell decline, repertoire contraction and weaker vaccine responses are all measured directly in people; every proposal to reverse them has been tested only in mice or in small uncontrolled human studies."
issues: ["The innate arm gets one paragraph; monocyte and NK-cell changes deserve their own treatment."]
---

```infobox
{
  "caption": "Process in the biology of aging",
  "rows": [
    { "label": "Term coined", "value": "1969, Roy Walford" },
    { "label": "Principal site", "value": "Thymus and T-cell compartment" },
    { "label": "Onset", "value": "Thymic involution from childhood" },
    { "label": "Main correlate", "value": "Cytomegalovirus infection" },
    { "label": "Standard readout", "value": "Vaccine response" },
    { "label": "Validated biomarker", "value": "None" },
    { "label": "Related hallmark", "value": "Chronic inflammation" }
  ]
}
```

**Immunosenescence** is the set of changes that accumulate in the immune system with age: a shrinking capacity to respond to antigens the body has not met before, alongside persistent low-grade activation against antigens it has. It is the immune half of biological aging, and it accounts for a substantial part of the mortality gradient with age: infection kills disproportionately in the old, latent viruses reactivate, and immune control of transformed cells is one of the defences against [[cancer]] that weakens.

## What changes

The thymus, where T cells are selected and licensed, begins involuting in childhood and is largely replaced by fat by middle age. Nothing else in the body loses so much of an organ so early with so little clinical notice, and the consequence is that the supply of new naive T cells falls steeply while the existing pool must be maintained by division rather than replacement.

The result is a compositional shift. Naive T cells decline, memory and terminally differentiated effector cells expand, and the diversity of the T-cell receptor repertoire contracts. Deep sequencing of receptor repertoires in donors from childhood to old age found diversity tracking the naive T-cell fraction, which fell roughly linearly to about the age of 70.[^britanova2014] A narrower repertoire is a narrower set of things that can be recognised, which is the mechanistic link between an aged immune system and a novel pathogen.

The B-cell compartment shifts similarly, with reduced output of new naive B cells and less effective germinal-centre reactions, so antibody responses to new antigens are lower and less durable. Innate immunity changes rather than simply weakening: neutrophil and macrophage function declines while baseline inflammatory signalling rises, the process described under [[inflammaging]]. The two are usually treated together because the [[hallmarks-of-aging|2023 revision of the hallmarks]] added chronic inflammation as a hallmark in its own right, and because senescent cells, which accumulate through the mechanism set out in [[cellular-senescence]], secrete much of the signal involved. Falling output from the haematopoietic compartment is the same phenomenon described under [[stem-cell-exhaustion]], seen from the immune side.

## Cytomegalovirus and the immune risk phenotype

Most of the world's population carries cytomegalovirus for life, and controlling it consumes a disproportionate share of the T-cell compartment. In some older people, virus-specific clones occupy a large fraction of the entire CD8 pool.

Swedish longitudinal studies of octogenarians and nonagenarians identified a cluster of markers that predicted death within two to six years: an inverted CD4/CD8 ratio, poor T-cell proliferative response, and cytomegalovirus seropositivity. The cluster was named the immune risk phenotype and is the closest thing the field has to a validated prognostic immune signature.[^wikby2005] The same research group later reported that individuals who survived to 100 did not show the profile, which is consistent either with the phenotype being a cause of death or with it being a marker of people already failing.

> [!debate] Decline or remodelling
> The word "immunosenescence" implies degeneration, and the field is not unanimous that this is the right description. On the alternative reading the aged immune system is remodelled rather than broken: it is optimised for the antigens it has already met, at the cost of the ones it has not, which is a reasonable allocation for an organism past reproduction. Cytomegalovirus is the sharpest version of the argument. It correlates with mortality in Swedish cohorts and does not in some other populations, which suggests its effect depends on context rather than being a fixed cost.

## Measurement

There is no accepted assay for immune age, and this is the field's practical bottleneck. Two approaches have been taken.

The first builds composite scores from immune-cell phenotyping. A nine-year longitudinal study of healthy adults produced a trajectory-based metric, IMM-AGE, that predicted all-cause mortality in an independent cohort after adjustment for established risk factors, and that tracked immune status better than chronological age.[^alpert2019] Composite scores built from DNA methylation and from circulating inflammatory proteins are parallel efforts in the same direction, and share the interpretive difficulty set out under [[epigenetic-clock|epigenetic clocks]]: a score that predicts mortality is not thereby a measure of the process that causes it.

The second uses function rather than composition, and the standard functional readout is vaccine response. A quantitative review of thirty-one studies found substantially lower rates of seroconversion and seroprotection after influenza vaccination in older adults than in younger ones across all three vaccine components.[^goodwin2006] Vaccine response has the advantage of measuring what actually matters and the disadvantage of being an indirect and noisy measure of the underlying state. Neither approach has been accepted by a regulator as a surrogate endpoint, which is the general problem set out under [[aging-biomarkers]].

## Consequences

The clearest is infection. Age-specific mortality from respiratory infection rises steeply and roughly log-linearly through later life, a gradient that was measured with unusual precision during the COVID-19 pandemic and that held across countries with very different health systems.[^odriscoll2021] That gradient is the reason [[engineered-pandemics|pandemic risk]] and longevity are not separate subjects.

Reactivation of latent viruses is the second. Herpes zoster is the visible case: the virus is controlled by cell-mediated immunity that declines with age, and its reactivation rate rises accordingly.

The third is [[cancer|tumour]] surveillance. Immune control of transformed cells is real, as checkpoint blockade demonstrated by removing a brake on it, and a declining capacity to exercise that control is one proposed contributor to the steep rise in cancer incidence with age. The contribution has not been quantified in humans, and the age-incidence curve is adequately explained by mutation accumulation alone, so this remains a plausible mechanism rather than a measured one.

## Interventions with evidence in people

Vaccine formulation is where the field has delivered. High-dose influenza vaccine was more efficacious than standard dose in a randomised trial in adults 65 and over.[^diazgranados2014] The adjuvanted recombinant zoster vaccine reached efficacy above 90% against shingles in adults over 50, far exceeding the live-attenuated vaccine it replaced.[^lal2015] Neither reverses immunosenescence. Both work around it by presenting more antigen or better adjuvant to a system that responds poorly to less.

Drug intervention has one suggestive human result. A randomised trial of low-dose mTOR inhibitors in older adults reported improved influenza vaccination response, upregulated antiviral gene expression, and fewer self-reported respiratory infections over the following year.[^mannick2018] The compound was later taken into a larger trial that did not meet its primary endpoint, so the result stands as the most interesting human signal in geroscience and not as a demonstrated therapy. It is the main reason [[rapamycin]] is discussed as an immune-restoring agent rather than only as an immunosuppressant.

[[exercise-and-aging|Exercise]] is associated with better vaccine responses and lower infection mortality in observational data, with the causal direction hard to establish in a population where illness reduces activity.

## Rejuvenation proposals

Thymic regeneration is the most direct approach. A small uncontrolled study combining growth hormone with two diabetes drugs reported regrowth of thymic tissue on imaging in nine men, along with a reversal of epigenetic clock readings that attracted more attention than the immune data.[^fahy2019] Nine participants with no control arm cannot establish an effect, and the follow-up study has not reported an outcome that changes that.

Other routes are earlier. Engineering thymic tissue is one of the objectives of [[tissue-engineering]] work, and allogeneic cultured thymus tissue is already approved for children born without the organ. Rejuvenating haematopoietic stem cells by [[partial-reprogramming|transient reprogramming]] is under investigation at companies including [[newlimit]], entirely in cells and mice so far. Clearing senescent T cells with [[senolytics]] has a rationale and no human efficacy data. Immunity is one of the three functional domains scored by the [[xprize-healthspan]] competition, which is a useful forcing function precisely because the field cannot yet agree what to measure.

## Open problems

Whether immunosenescence is a driver of aging or one of its outputs is unsettled, and the question is not merely academic: it determines whether restoring immune function would extend [[healthspan]] broadly or only reduce infection deaths. The [[geroscience-hypothesis]] predicts the former and has no human evidence for it.

The measurement gap is more tractable and more urgent. Without a validated marker, a trial of an immune-restoring intervention has to be powered on clinical infection or vaccine response, which makes it large, slow, and confined to endpoints regulators already accept. The interventions with the best human evidence remain the ones that route around the aged immune system rather than repairing it, and there is no clear reason yet to expect that to change soon.

## See also

- [[inflammaging]]
- [[stem-cell-exhaustion]]
- [[cellular-senescence]]
- [[hallmarks-of-aging]]
- [[cancer]]
- [[aging-biomarkers]]
- [[rapamycin]]
- [[xprize-healthspan]]

## References

[^britanova2014]: `paper` Britanova, O. V. et al. "Age-Related Decrease in TCR Repertoire Diversity Measured with Deep and Normalized Sequence Profiling." *The Journal of Immunology*, 2014. {Thirty-nine healthy donors sampled cross-sectionally, so the decline is inferred across people rather than followed within them.}
[^wikby2005]: `paper` Wikby, A. et al. "An Immune Risk Phenotype, Cognitive Impairment, and Survival in Very Late Life." *Journals of Gerontology: Series A*, 2005; and Strindhall, J. et al. "No Immune Risk Profile among individuals who reach 100 years of age." *Experimental Gerontology*, 2007. {Both draw on the same small Swedish cohorts, and the phenotype has not replicated consistently in other populations.}
[^alpert2019]: `paper` Alpert, A. et al. "A clinically meaningful metric of immune age derived from high-dimensional longitudinal monitoring." *Nature Medicine*, 2019. {Derived from 135 healthy adults and validated against mortality in the Framingham cohort; it has not been used as a trial endpoint.}
[^goodwin2006]: `paper` Goodwin, K., Viboud, C. and Simonsen, L. "Antibody response to influenza vaccination in the elderly: a quantitative review." *Vaccine*, 2006. {Pools antibody responses across 31 studies; antibody titre is a correlate of protection, not protection itself.}
[^odriscoll2021]: `paper` O'Driscoll, M. et al. "Age-specific mortality and immunity patterns of SARS-CoV-2." *Nature*, 2021.
[^diazgranados2014]: `paper` DiazGranados, C. A. et al. "Efficacy of High-Dose versus Standard-Dose Influenza Vaccine in Older Adults." *New England Journal of Medicine*, 2014.
[^lal2015]: `paper` Lal, H. et al. "Efficacy of an Adjuvanted Herpes Zoster Subunit Vaccine in Older Adults." *New England Journal of Medicine*, 2015. {Efficacy was sustained across age strata in this trial, which is unusual; a companion trial extended the finding to adults over 70.}
[^mannick2018]: `paper` Mannick, J. B. et al. "TORC1 inhibition enhances immune function and reduces infections in the elderly." *Science Translational Medicine*, 2018; and Mannick, J. B. et al. "Targeting the biology of ageing with mTOR inhibitors to improve immune function in older adults: phase 2b and phase 3 randomised trials." *The Lancet Healthy Longevity*, 2021. {The infection endpoint in the 2018 trial was self-reported over the following year, and the later phase 3 did not meet its primary endpoint.}
[^fahy2019]: `paper` Fahy, G. M. et al. "Reversal of epigenetic aging and immunosenescent trends in humans." *Aging Cell*, 2019. {An uncontrolled study of nine men, run to test thymus regeneration rather than to test an intervention against a clock.}

==============================================================================
ARTICLE: in-vitro-fertilisation
TITLE:   In vitro fertilisation
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/in-vitro-fertilisation
SOURCE:  https://futurehumanwiki.com/raw/in-vitro-fertilisation
==============================================================================

---
title: "In vitro fertilisation"
slug: "in-vitro-fertilisation"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["reproduction", "society"]
tags: ["reproduction", "ivf", "embryo selection", "fertility", "regulation", "bioethics"]
summary: "The routine clinical procedure in which eggs are fertilised outside the body and an embryo is transferred to a uterus, and the platform every other embryo technology is built on."
updated: "2026-07-28"
humanEvidence: "Between ten and thirteen million children had been born after assisted reproduction by 2018 and several million since; live-birth rates fall steeply with the age of the egg, and birth defects are modestly more common than after natural conception."
access: "Available in most countries with a fertility sector but funded unevenly: NHS-funded cycles fell from 35 percent of UK treatments in 2019 to 27 percent in 2023, and a self-funded cycle costs several thousand pounds or dollars before optional extras."
reversibility: "irreversible"
issues: ["Cost figures are given for the UK only; a comparable US and international picture is missing."]
---

```infobox
{
  "caption": "Assisted reproductive technology",
  "rows": [
    { "label": "First live birth", "value": "1978, Oldham, England" },
    { "label": "Developed by", "value": "Robert Edwards, Patrick Steptoe" },
    { "label": "Nobel Prize", "value": "Medicine, 2010" },
    { "label": "Children born", "value": "10–13 million by 2018" },
    { "label": "Share of UK births", "value": "1 in 32 (2023)" },
    { "label": "Key variants", "value": "ICSI, frozen transfer, PGT" },
    { "label": "UK regulator", "value": "HFEA" }
  ]
}
```

**In vitro fertilisation** is the procedure in which eggs are recovered from ovaries, combined with sperm in a laboratory dish, cultured for a few days, and transferred to a uterus as an embryo. It is the oldest and most-performed of the assisted reproductive technologies, and it matters far beyond infertility medicine: almost every technique this wiki covers under reproduction, from [[embryo-selection]] to [[germline-editing]], is an operation performed on an embryo that only exists because of IVF.

## How a cycle works

A cycle runs in four steps: stimulation, egg collection, fertilisation and culture, transfer. Injected gonadotrophins stimulate the ovaries to mature many follicles at once rather than the single one of a natural cycle; ultrasound and blood hormone measurements track the response; and eggs are collected under sedation by a needle passed through the vaginal wall. In the laboratory the eggs are either mixed with prepared sperm, or, where sperm quality is poor, injected one sperm per egg by intracytoplasmic sperm injection. Fertilised eggs are cultured for two to six days, usually to the blastocyst stage. One embryo is transferred; the rest are vitrified and stored.

Two changes have reshaped that sequence since the 1990s. ICSI, first reported by a Brussels group in 1992, made severe male-factor infertility treatable and is now used in a majority of cycles worldwide, including many where the indication for it is unclear.[^palermo1992] Vitrification, an ultra-rapid cooling method that avoids ice crystal formation, made frozen embryo and egg storage reliable enough that transferring a thawed embryo in a later cycle is now often preferred to transferring a fresh one. It is the same physical principle that [[cryonics]] proposes to apply to a whole body, working here on a hundred-micrometre object with no vasculature.

## Development history

```timeline
[
  { "year": "1959", "title": "Rabbit IVF confirmed", "text": "Min Chueh Chang produces live rabbit young from eggs fertilised in vitro, establishing that mammalian fertilisation outside the body can yield normal offspring." },
  { "year": "1978", "title": "Louise Brown", "text": "Patrick Steptoe and Robert Edwards report the first human birth after in vitro fertilisation, achieved in an unstimulated cycle with a single egg." },
  { "year": "1990", "title": "Preimplantation testing", "text": "Alan Handyside's group biopsies human embryos and selects by sex to avoid X-linked disease, opening the field of embryo testing." },
  { "year": "1990–1991", "title": "Statutory regulation in the UK", "text": "The Human Fertilisation and Embryology Act creates a licensing authority with a fourteen-day limit on embryo culture, the model most later regimes were measured against." },
  { "year": "1992", "title": "ICSI", "text": "A single sperm injected directly into an egg produces pregnancies, extending treatment to men with almost no motile sperm." },
  { "year": "2010", "title": "Nobel Prize", "text": "Robert Edwards receives the Nobel Prize in Physiology or Medicine; Steptoe had died in 1988 and the prize is not awarded posthumously." },
  { "year": "2012", "title": "Egg freezing leaves the experimental category", "text": "Vitrification results are judged good enough for the American Society for Reproductive Medicine to drop the experimental label from oocyte cryopreservation, and elective egg banking becomes a commercial market." }
]
```

## What it delivers

The dominant variable is the age of the egg, and success rates fall steeply through the late thirties and forties. They are also better described per patient across several cycles than per transfer, and a cumulative live-birth rate is the figure a well-run clinic will quote.

Volume is now substantial. An analysis of four decades of international registry data put the number of children born after assisted reproduction between 1978 and 2018 at between roughly 9.8 and 13 million, with several million more since.[^adamson2025] UK licensed clinics ran more than 77,500 IVF cycles for 52,400 patients in 2023, producing almost 21,000 babies, about one UK birth in thirty-two; the multiple birth rate fell to 3.4% as single embryo transfer became standard.[^hfea2023] That last number is an underrated public-health result, because twin pregnancies carry much of the obstetric risk historically attributed to IVF itself.

> [!key] Why this article sits in the middle of the reproduction portal
> Every embryo technology covered here operates on an embryo outside a body. [[polygenic-embryo-screening]] ranks embryos, [[mitochondrial-replacement-therapy]] rebuilds one, heritable editing would rewrite one, and [[in-vitro-gametogenesis]] would supply the eggs to make far more of them. None of those has an independent route to a pregnancy. IVF is the delivery mechanism for all of them, which is why its cost, its regulation, and the number of embryos a cycle yields set the practical limits on each.

## The platform other technologies stand on

Embryo biopsy and testing are the clearest case. Preimplantation genetic testing for monogenic disease is reliable and uncontroversial where a couple carries a known variant. Testing for aneuploidy is performed on a large share of cycles in some markets, and randomised trials have not shown that it raises live-birth rates per patient.[^munne2019] Polygenic screening extends the same biopsy to complex traits and is rejected as unvalidated by professional genetics bodies.

Beyond selection, IVF is the necessary substrate for anything that edits or reconstructs an embryo. The [[he-jiankui-affair]] was carried out inside ordinary IVF cycles with ordinary equipment, using [[crispr-cas9]] reagents that can be ordered, which is the fact that makes [[governance-of-genome-editing|governance]] hard: the specialised part of heritable editing is a reagent, and the rest is a clinic that exists in most countries. In vitro gametogenesis, if it works in humans, would remove the binding constraint on embryo number and change what [[designer-babies|selection]] could deliver. Spare and donated embryos are also the starting material for human embryonic stem cell lines and for much of what is known about early human development, which is part of why [[induced-pluripotent-stem-cells|induced pluripotent stem cells]] and [[synthetic-embryos|embryo models]] were pursued at all: each reduces the dependence on donated embryos without removing it.

## Add-ons and the evidence problem

IVF is largely a private market even in countries with public health systems, and it sells optional extras: endometrial scratching, immunological treatments, time-lapse incubation, assisted hatching. The UK regulator rates these on a published scale and, at its 2023 overhaul, placed none of the thirteen it had reviewed in the category reserved for treatments shown by multiple good trials to improve live-birth rates.[^hfeaaddons] The scale itself has been criticised for compressing very different kinds of uncertainty into one colour, but the finding is not seriously disputed: patients pay for interventions that mostly have not been shown to work.

> [!caution] What the success rate does not tell you
> A clinic's headline rate depends on which patients it accepts and how it counts. Rates per embryo transferred exclude cycles cancelled before transfer; rates per cycle started include them. Neither figure describes an individual's chance, which is dominated by age and by cause of infertility.

## Risks and outcomes for children

The ovarian stimulation itself carries a small risk of ovarian hyperstimulation syndrome, which has become less common with milder protocols and freeze-all strategies. Egg collection is a minor surgical procedure with the usual bleeding and infection risks.

For the children, the honest summary is that risks are real, small, and hard to attribute. A systematic review and meta-analysis found birth defects modestly more common after assisted reproduction than after natural conception.[^hansen2013] Preterm birth and low birthweight are likewise somewhat more frequent. Whether that is caused by the laboratory procedures, by the underlying subfertility of the parents, or by the multiple pregnancies that were common in earlier practice has never been fully separated, and the same ambiguity attaches to the slightly raised rate of imprinting disorders. Follow-up now extends to adults in their forties without a signal that would change clinical practice.

## Access, law and cost

Regulation ranges from statute to nothing. The United Kingdom licenses clinics and specifies what may be tested for; several European states restrict embryo creation and storage tightly; the United States regulates laboratories and advertising but leaves the substance of practice to professional guidance. The gaps between these regimes drive reproductive travel, and they are the same gaps analysed under [[access-and-inequality]]. Where the law does bind, it binds oddly: the fourteen-day limit on embryo culture written into the UK statute in 1990 was set at a point nobody could then reach, and it is now the constraint the work described under [[ectogenesis]] and embryo modelling runs into.

Funding is where the difference bites hardest. In the UK the NHS-funded share of cycles fell from 35% in 2019 to 27% in 2023, with wide regional variation.[^hfea2023] Elsewhere, cost is the binding constraint on who has children this way, and it compounds every downstream capability: a technology priced as an add-on to a self-funded cycle is available to the population that can already afford the cycle.

Legal status of the embryo is unsettled in a way that can disrupt practice overnight. In February 2024 the Supreme Court of Alabama held that frozen embryos are children under the state's wrongful-death statute; several clinics suspended treatment, and the legislature passed immunity legislation within three weeks.[^lepage2024] The episode is the clearest recent demonstration that IVF's stability rests on a legal settlement about embryo status that is neither uniform nor secure, and that every technology built on top of it inherits that exposure.

## See also

- [[embryo-selection]]
- [[polygenic-embryo-screening]]
- [[in-vitro-gametogenesis]]
- [[mitochondrial-replacement-therapy]]
- [[germline-editing]]
- [[reproductive-longevity]]
- [[artificial-womb]]
- [[uterus-transplantation]]

## References

[^palermo1992]: `paper` Palermo, G., Joris, H., Devroey, P. and Van Steirteghem, A. C. "Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte." *The Lancet*, 1992. {Four pregnancies from eight treatment cycles in couples for whom conventional insemination had already failed.}
[^hfea2023]: `report` Human Fertilisation and Embryology Authority. *Fertility treatment 2023: trends and figures*. HFEA, June 2025. {UK licensed clinics only; the regulator collects every cycle by statute, which makes the denominator unusually complete.}
[^adamson2025]: `paper` Adamson, G. D., Creighton, P., de Mouzon, J., Zegers-Hochschild, F., Dyer, S. and Chambers, G. M. "How many infants have been born with the help of assisted reproductive technology?" *Fertility and Sterility*, 2025. {A range rather than a count, because national registries differ in coverage and several large markets report incompletely.}
[^munne2019]: `paper` Munné, S. et al. "Preimplantation genetic testing for aneuploidy versus morphology as selection criteria for single frozen-thawed embryo transfer in good-prognosis patients: a multicenter randomized clinical trial." *Fertility and Sterility*, 2019.
[^hfeaaddons]: `regulator` Human Fertilisation and Embryology Authority. "UK fertility regulator launches improved ratings for fertility treatment add-ons." HFEA, October 2023. {A regulator's own rating scheme; the ratings summarise published trial evidence rather than adding new data.}
[^hansen2013]: `paper` Hansen, M., Kurinczuk, J. J., Milne, E., de Klerk, N. and Bower, C. "Assisted reproductive technology and birth defects: a systematic review and meta-analysis." *Human Reproduction Update*, 2013. {Pools observational studies, so it cannot separate the effect of the procedures from that of the parents' underlying subfertility.}
[^lepage2024]: `law` LePage v. Center for Reproductive Medicine, P.C., Supreme Court of Alabama, February 2024. {A wrongful-death ruling about embryos destroyed in a storage accident; it did not restrict IVF directly, and the legislature responded with an immunity statute.}

==============================================================================
ARTICLE: in-vitro-gametogenesis
TITLE:   In vitro gametogenesis
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/in-vitro-gametogenesis
SOURCE:  https://futurehumanwiki.com/raw/in-vitro-gametogenesis
==============================================================================

---
title: "In vitro gametogenesis"
slug: "in-vitro-gametogenesis"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 3
categories: ["reproduction", "genetics"]
tags: ["reproduction", "stem cells", "gametes", "fertility", "embryo selection", "bioethics"]
summary: "Making functional eggs or sperm from pluripotent stem cells in culture, achieved end to end in mice and still short of the meiotic stages in humans."
updated: "2026-07-27"
humanEvidence: "Human pluripotent cells have been carried as far as oogonia in culture and stop before meiosis; no human egg or sperm has been made in vitro, and the complete cycle exists only in mice."
access: "Not available anywhere: no jurisdiction permits clinical use, the UK statute excludes such gametes by construction, and the work is confined to research laboratories and a few companies."
reversibility: "irreversible"
issues: ["The US appropriations rider and the UK permitted-gamete definition are described without sources."]
---

```infobox
{
  "caption": "Reproductive stem-cell technology",
  "rows": [
    { "label": "Abbreviation", "value": "IVG" },
    { "label": "Input", "value": "iPSCs or ESCs", "link": "/wiki/induced-pluripotent-stem-cells" },
    { "label": "Complete in mice", "value": "2016 (oocytes)" },
    { "label": "Human status", "value": "Oogonia; no mature gametes" },
    { "label": "Key laboratories", "value": "Saitou, Hayashi (Japan)" },
    { "label": "Legal status (UK)", "value": "Not a permitted gamete" },
    { "label": "Clinical use", "value": "None" }
  ]
}
```

**In vitro gametogenesis** is the production of functional eggs or sperm from pluripotent stem cells in culture, without a gonad. In mice the full female cycle has been closed: embryonic stem cells have been converted to oocytes entirely in a dish, fertilised, and carried to healthy pups whose own cells could start the cycle again. In humans, the same protocols stall well before meiosis. The gap between those two sentences is the whole subject, and it has not narrowed as quickly as the mouse results led people to expect.

```keyfacts
[
  { "value": "2016", "label": "Full mouse oogenesis in vitro", "note": "Hikabe et al., from embryonic stem cells to fertile offspring" },
  { "value": "2023", "label": "Mouse eggs from male cells", "note": "bipaternal pups born, at very low efficiency" },
  { "value": "0", "label": "Functional human gametes made in vitro", "note": "as of 2026" }
]
```

## How it works

Germ cells are set aside early in development and then follow a schedule no other lineage repeats: they erase most of their DNA methylation, re-establish sex-specific imprints, halve their chromosome number through meiosis, and, in the female line, grow enormously while stockpiling the maternal factors an embryo needs before its own genome switches on. Reconstituting that in culture means reproducing each stage in order.

The standard route has four steps.

1. **Pluripotency.** Start from embryonic stem cells or from [[induced-pluripotent-stem-cells]] made by [[yamanaka-factors|reprogramming]] a somatic cell such as a skin fibroblast or a blood cell.
2. **Germline specification.** Push the cells through an epiblast-like intermediate and then, with BMP signalling and the right transcription factors, into primordial germ cell–like cells. In humans, SOX17 rather than the mouse's BLIMP1-first logic turns out to be the critical specifier, one of several places where the two species diverge.[^irie2015]
3. **Gonadal environment.** Aggregate the germ cells with embryonic gonadal somatic cells to form a reconstituted ovary or testis. The somatic niche supplies the signals that license meiosis; germ cells will not do it alone.
4. **Maturation.** Grow the resulting follicles or spermatogenic cells to a fertilisable gamete. This is the step that has never been completed for humans.

Step three is the hidden bottleneck. Mouse experiments use somatic cells dissected from mouse embryonic gonads, which is not an option in humans; the field therefore also needs to make human ovarian and testicular somatic cells from stem cells, a problem being worked on in parallel and not yet solved to the point of supporting complete human folliculogenesis.

## Development history

```timeline
[
  { "year": "2011", "title": "Mouse sperm", "text": "Hayashi and Saitou reconstitute germ-cell specification in culture; the resulting cells produce fertile sperm after transplantation into mouse testes." },
  { "year": "2012", "title": "Mouse eggs", "text": "The same approach yields oocytes, though maturation still requires transplantation of reconstituted ovaries into a live animal." },
  { "year": "2015", "title": "Human germ cells specified", "text": "Two groups independently derive human primordial germ cell–like cells and identify SOX17 as the key specifier, showing the human pathway differs from the mouse one." },
  { "year": "2016", "title": "The mouse cycle closes", "text": "Hikabe and colleagues complete the entire female germline cycle in vitro and obtain fertile offspring, at low efficiency." },
  { "year": "2018", "title": "Human oogonia", "text": "Human germ cell–like cells are matured to oogonia inside xenogeneic reconstituted ovaries built with mouse gonadal somatic cells, then stop short of meiosis." },
  { "year": "2023", "title": "Eggs from male cells", "text": "Hayashi's group reports functional oocytes derived from male mouse cells, involving loss of the Y and duplication of the X, with pups born from bipaternal embryos." },
  { "year": "2020s", "title": "Companies form", "text": "Conception Biosciences, Gameto and others begin funding human work; none has reported a mature human gamete." }
]
```

The programme is largely the work of two connected Japanese laboratories, Mitinori Saitou's at Kyoto and Katsuhiko Hayashi's at Kyushu and later Osaka, both descended from the reprogramming tradition established by [[shinya-yamanaka]] in the same city. The 2011 and 2012 papers established that pluripotent cells could be routed into the germline reliably.[^hayashi2011] The 2016 paper by Orie Hikabe and colleagues was the milestone: oocytes generated wholly in culture, fertilised, transferred, and producing pups that grew into fertile adults.[^hikabe2016] Efficiency was poor, with only a small fraction of the oocytes produced proving competent, but the pathway was closed.

The 2023 result attracted the most attention outside the field. Cultured male mouse cells lost the Y chromosome at some frequency, and the resulting X-monosomic cells were treated with a compound that promoted X duplication, yielding XX cells that could be taken through oogenesis.[^murakami2023] Embryos made from these eggs and ordinary sperm produced live pups. The yield was on the order of one percent of transferred embryos, which is the number that most reporting omitted.

## The human gap

Human germ cell–like cells are now made routinely. Getting them further has proved slow. Culture in xenogeneic reconstituted ovaries carries them to the oogonial stage, at which point they have undergone extensive epigenetic reprogramming but have not entered meiosis.[^yamashiro2018] Subsequent work has extended the reprogramming and expansion phases and clarified how the human germline erases and rebuilds its methylation marks,[^murase2024] but as of 2026 no laboratory has publicly reported a human oocyte or spermatozoon made in vitro, and none has reported fertilisation with one.

Several things make humans harder. Human oogenesis takes months rather than weeks, and the follicular growth phase is correspondingly long and poorly reproduced in culture. Human germ cells specify differently, so mouse protocols do not transfer. And the ethical and legal constraints on obtaining human embryonic gonadal tissue restrict the shortcut that made the mouse work possible.

> [!caution] Distinguish the mouse claim from the human claim
> Every complete demonstration of in vitro gametogenesis is in mice. Reports described in press coverage as human breakthroughs have concerned earlier stages of the pathway, ovarian support cells used to improve conventional egg maturation, or nuclear-transfer approaches that are not gametogenesis from pluripotent cells at all.

## What it would change

If human IVG worked, its effects would run through several other technologies covered here.

**Infertility.** Anyone with somatic cells could in principle produce gametes, including people whose gonads were removed or damaged by cancer treatment, people with premature ovarian insufficiency, and men with non-obstructive azoospermia. This is the application the companies name and the one most likely to reach a clinic first. It addresses gametes and not gestation, so absolute uterine factor infertility would still require [[uterus-transplantation]] or gestational surrogacy: every [[artificial-womb]] built so far takes over a pregnancy already underway, and [[ectogenesis|complete ectogenesis]] has no demonstration in any mammal.

**Ovarian aging.** Oocyte quality declines steeply with maternal age, the central fact of [[reproductive-longevity]]. IVG would decouple the age of the gametes from the age of the person, though only if the somatic cells used as input have not themselves accumulated damage. That is an open question: [[epigenetic-reprogramming]] resets methylation marks but does not repair point mutations, and the germline's usual quality-control filters would be operating on cells that never passed through a gonad.

**Same-sex genetic parenthood.** The mouse work shows the male-to-female direction is possible in principle. The reverse, making sperm from XX cells, has not been achieved and faces the problem that spermatogenesis requires Y-linked genes.

**Embryo supply.** This is the consequential one. [[embryo-selection]] and [[polygenic-embryo-screening]] are limited less by prediction accuracy than by the handful of embryos an IVF cycle yields. IVG would in principle supply hundreds. The gain does not scale linearly: because selecting the best of *n* draws from a normal distribution improves roughly with the square root of the logarithm of *n*, going from ten embryos to a thousand less than doubles the expected shift rather than multiplying it a hundredfold. Even so, it is the only route by which polygenic selection could produce effects large enough to matter, and it makes proposals such as iterated embryo selection, in which stem cells from selected embryos are used to make the next generation of gametes in vitro, technically coherent for the first time.[^shulman2014] The constraints on that idea are discussed in [[genetic-enhancement-of-intelligence]].

**Editing.** IVG would also make heritable [[germline-editing]] more tractable, since edits could be made and verified in a cell line before a gamete is made, avoiding the mosaicism that plagues editing of zygotes. That combination is what most commentators mean when they use the term [[designer-babies]], and it is the reason the technology attracts attention disproportionate to its clinical readiness.

**Conservation.** The same protocols applied to endangered species would allow gametes to be made from banked tissue, an approach pursued alongside the cloning-based methods discussed in [[de-extinction]]. Progress here has been limited by the same species-specificity problem: a protocol tuned for mice does not transfer to a rhinoceros any more easily than it transfers to a human.

## Risks and limitations

Efficiency is the first constraint and may be the least interesting one. The serious concerns are epigenetic.

Genomic imprinting, the parent-specific silencing of certain genes, is established during gametogenesis and is essential to normal development. Culture perturbs it; imprinting disorders are somewhat more common after conventional IVF, and IVG puts the entire imprinting cycle in a dish. Mouse offspring have appeared healthy, but mice tolerate imprinting perturbation differently from humans and the numbers examined remain small.

Somatic input cells carry mutations acquired over a lifetime, including the clonal expansions described in [[stem-cell-exhaustion]]. Reprogramming does not repair DNA damage, and culture adds its own mutations and karyotypic abnormalities. A gamete made from a fifty-year-old's blood cells is genetically fifty years old even if it is epigenetically young.

There is also a consent problem with no precedent. Gametes could be derived from any nucleated cell, so a discarded biopsy, a blood sample, or in principle a hair follicle would be sufficient starting material to make someone a genetic parent without their knowledge. Legal scholars have flagged this as a gap in existing tissue-consent law.[^cohen2017]

## Regulation

No jurisdiction currently permits the clinical use of gametes made in vitro. The UK's Human Fertilisation and Embryology Act defines "permitted" eggs and sperm as those produced by or extracted from a person's ovaries or testes, which excludes IVG gametes by construction; using them in treatment would require primary legislation, and the regulator has flagged the point in its law-reform advice. In the United States, an appropriations rider bars the FDA from considering applications involving heritable modification of embryos, which does not cleanly describe IVG but has been read to complicate it, and the agency would in any case treat the gametes as a biological product requiring a full development programme. The International Society for Stem Cell Research's guidelines place clinical IVG in the category of research that should not proceed pending evidence of safety, while permitting the laboratory work.

The governance question resembles the one described in [[governance-of-genome-editing]]: a technology whose research phase is legal nearly everywhere and whose clinical phase is legal nowhere, in a field with an established pattern of reproductive tourism. The first human IVG birth, if it happens, is more likely to occur in a jurisdiction with weak oversight than in one that has debated the question, which is what happened with [[mitochondrial-replacement-therapy]] and, in a more serious way, with the [[he-jiankui-affair]].

## Outlook

The near-term milestones are legible. A human oocyte that completes meiosis in culture would be the first; fertilisation and normal preimplantation development would be the second; and a human ovarian somatic cell population capable of supporting folliculogenesis without mouse tissue would be the enabling step for both. Sperm may arrive earlier than eggs, since spermatogenesis is shorter and the cells are smaller and simpler, though in vitro human spermatogenesis has its own long record of failure.

Forecasts from within the field have consistently been too optimistic; predictions of clinical use within a decade have been made repeatedly since 2016 and have not held. What has not been argued away is that the mouse demonstration is complete, which distinguishes IVG from most technologies on this wiki. The question is whether the human germline's longer, differently wired developmental programme is a matter of more culture optimisation or of biology that does not run outside a body at all.

## See also

- [[embryo-selection]]
- [[polygenic-embryo-screening]]
- [[reproductive-longevity]]
- [[induced-pluripotent-stem-cells]]
- [[germline-editing]]
- [[artificial-womb]]
- [[human-cloning]]
- [[procreative-beneficence]]

## References

[^hayashi2011]: `paper` Hayashi, K. et al. "Reconstitution of the mouse germ cell specification pathway in culture by pluripotent stem cells." *Cell*, 2011.
[^irie2015]: `paper` Irie, N. et al. "SOX17 is a critical specifier of human primordial germ cell fate." *Cell*, 2015.
[^hikabe2016]: `paper` Hikabe, O. et al. "Reconstitution in vitro of the entire cycle of the mouse female germ line." *Nature*, 2016. {The closed cycle is in mice; efficiency was low, and only a small fraction of the cultured oocytes proved developmentally competent.}
[^yamashiro2018]: `paper` Yamashiro, C. et al. "Generation of human oogonia from induced pluripotent stem cells in vitro." *Science*, 2018.
[^murakami2023]: `paper` Murakami, K. et al. "Generation of functional oocytes from male mice in vitro." *Nature*, 2023. {Mice only, and the yield was on the order of one percent of transferred embryos; nothing equivalent has been attempted with human cells.}
[^murase2024]: `paper` Murase, Y. et al. "In vitro reconstitution of epigenetic reprogramming in the human germ line." *Nature*, 2024.
[^shulman2014]: `paper` Shulman, C. and Bostrom, N. "Embryo Selection for Cognitive Enhancement: Curiosity or Game-changer?" *Global Policy*, 2014. {A theoretical analysis of what selection could deliver under stated genetic assumptions; it reports no experiment and predates current predictor accuracy.}
[^cohen2017]: `paper` Cohen, I. G., Daley, G. Q. and Adashi, E. Y. "Disruptive reproductive technologies." *Science Translational Medicine*, 2017.

==============================================================================
ARTICLE: induced-pluripotent-stem-cells
TITLE:   Induced pluripotent stem cells
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/induced-pluripotent-stem-cells
SOURCE:  https://futurehumanwiki.com/raw/induced-pluripotent-stem-cells
==============================================================================

---
title: "Induced pluripotent stem cells"
slug: "induced-pluripotent-stem-cells"
type: "technology"
status: "emerging"
horizon: "late 2020s"
trl: 7
categories: ["bodies", "genetics"]
tags: ["stem cells", "reprogramming", "regenerative medicine", "cell therapy", "disease modelling"]
summary: "Adult cells reprogrammed back to an embryonic-like state, capable of becoming any cell type in the body and of being derived from a specific patient."
updated: "2026-07-27"
humanEvidence: "Small early-phase trials in people report that iPSC-derived retinal, corneal, dopaminergic and islet grafts survive and in some cases function; none is randomized and no product has marketing approval as of mid-2026."
access: "Research lines are widely banked and sold for laboratory use; as a therapy there is nothing to obtain outside small early-phase trials, since no iPSC-derived product is approved anywhere."
reversibility: "difficult"
issues: ["The 2024 and 2025 clinical results are stated without footnotes.", "The per-patient cost of a clinical-grade autologous line is given without a source."]
---

```infobox
{
  "caption": "Cell type and derivation technology",
  "rows": [
    { "label": "Abbreviation", "value": "iPSC" },
    { "label": "First derived", "value": "2006 (mouse), 2007 (human)" },
    { "label": "Derived by", "value": "Takahashi and Yamanaka", "link": "/wiki/shinya-yamanaka" },
    { "label": "Starting material", "value": "Skin, blood, urine cells" },
    { "label": "First clinical use", "value": "2014, macular degeneration" },
    { "label": "Approved products", "value": "None as of mid-2026" },
    { "label": "Readiness", "value": "TRL 7" }
  ]
}
```

**Induced pluripotent stem cells** are ordinary adult cells that have been reprogrammed into an embryonic-like state from which they can be differentiated into any cell type in the body. Because they are made from a patient's own tissue, they sidestep both the embryo destruction that constrains embryonic stem cell work and, in principle, the immune rejection that constrains transplantation. They have transformed laboratory biology within a decade and have so far produced no approved therapy.

```figure
{"key": "ips-cell-colony", "caption": "Reprogrammed human cells growing in culture. The sharp colony border is part of how pluripotency is judged by eye."}
```

## Derivation

An iPSC line begins with an accessible somatic cell: a skin fibroblast, a peripheral blood mononuclear cell, or a renal epithelial cell shed into urine. The [[yamanaka-factors]] are introduced, and over two to four weeks a small fraction of the treated cells activate the endogenous pluripotency network and form colonies that can be picked, expanded and banked.[^takahashi2007]

Vector choice decides whether a line can ever reach a patient, so clinical derivation uses footprint-free methods: Sendai virus, episomal plasmids or synthetic mRNA, none of which leave a permanent trace in the genome. The effort then shifts from making the line to qualifying it, through karyotyping, copy-number analysis, sequencing for acquired mutations, and confirmation that no undifferentiated cells survive into the final product. Qualification rather than derivation is what makes a clinical-grade line expensive, a point that governs everything in the autologous debate below.

Reprogramming also resets the cell's methylation age. An iPSC line from an eighty-year-old reads as embryonic on the multi-tissue [[epigenetic-clock]] that [[steve-horvath]] built, which is the observation that launched [[epigenetic-reprogramming]] as a rejuvenation strategy.[^horvath2013] The reset does not extend to somatic mutations, which the donor cell carries forward. Deliberately stopping the process short of pluripotency, the technique known as [[partial-reprogramming]], is the attempt to capture the epigenetic reset without producing a stem cell at all.

Compared with embryonic stem cells, iPSCs match closely in differentiation capacity but differ in provenance. They can be made from anyone, at any age, without an embryo, which removes the legal restrictions that constrain ESC work in several countries. Against that, each line carries the donor's accumulated somatic mutations and a residual epigenetic memory of the tissue it came from, and lines vary more between one another than well-characterised ESC lines do.

## Development history

```timeline
[
  { "year": "2006", "title": "Mouse iPSCs", "text": "Takahashi and Yamanaka reprogramme mouse fibroblasts with four transcription factors." },
  { "year": "2007", "title": "Human iPSCs", "text": "Two groups independently derive human iPSCs, one using the original four factors and one a partly different set." },
  { "year": "2012", "title": "Nobel Prize", "text": "Shinya Yamanaka shares the prize with John Gurdon for showing that mature cells can be reprogrammed to pluripotency." },
  { "year": "2014", "title": "First patient", "text": "A woman with neovascular macular degeneration receives a sheet of retinal pigment epithelium grown from her own reprogrammed cells at RIKEN in Japan." },
  { "year": "2015–2017", "title": "Autologous approach reconsidered", "text": "A second planned transplant is cancelled after genomic changes are found in the prepared cells, and the programme shifts to banked, HLA-matched donor lines." },
  { "year": "2024–2025", "title": "First efficacy signals", "text": "Small trials report transplants of iPSC-derived corneal epithelium, dopaminergic progenitors for Parkinson's disease, and reprogrammed-cell-derived islets in type 1 diabetes." }
]
```

## Uses in research

The largest impact so far is not therapeutic. iPSCs let a laboratory obtain living neurons, cardiomyocytes or hepatocytes carrying a specific patient's genome, which was previously impossible for most tissues. Disease modelling in patient-derived neurons is now standard in work on Parkinson's disease, amyotrophic lateral sclerosis and rare developmental disorders, particularly when combined with [[crispr-cas9]] to create isogenic controls differing only at the variant of interest.

They are also the starting material for most [[organoids]], for cardiac and hepatic toxicity screening in drug development, and for the differentiation protocols that [[tissue-engineering]] and [[organ-bioprinting]] depend on for cell supply. In reproductive biology, [[in-vitro-gametogenesis]] proceeds through an iPSC or ESC intermediate, and the [[synthetic-embryos|stem-cell-based embryo models]] that have forced a rethink of the fourteen-day rule are assembled from pluripotent cells of either kind.

Large-scale panels have extended the approach to population genetics. Consortium collections of hundreds of donor lines allow a variant's effect on cell behaviour to be measured across many genetic backgrounds, which is closer to a controlled experiment in human genetics than anything previously available.

## Clinical progress

The first human use came in 2014, when a Japanese woman in her seventies received a sheet of retinal pigment epithelium grown from her own reprogrammed skin cells for neovascular age-related macular degeneration. The graft survived and the disease stabilised, though vision did not improve.[^mandai2017] The programme subsequently moved to banked donor lines after genomic changes were detected in cells prepared for a second patient.

Progress since has been incremental and concentrated in Japan and China. A Kyoto University phase I/II trial transplanted iPSC-derived dopaminergic progenitors into the brains of a small number of people with Parkinson's disease, reporting in Nature in 2025 that the cells survived, produced dopamine, and caused no tumours, with motor improvement in some participants. A Japanese group reported transplanting iPSC-derived corneal epithelial sheets into patients with limbal stem cell deficiency. In 2024 a Chinese group reported that a woman with type 1 diabetes became insulin-independent after receiving islets differentiated from her own chemically reprogrammed cells. Cardiomyocyte patches for heart failure have been trialled in Japan.

Every one of these is a small, early-phase study. As of mid-2026 no iPSC-derived product has full marketing approval anywhere. Japan's conditional-approval pathway for regenerative medicines is often cited as the route by which a first approval could arrive, though that pathway has itself been criticised for licensing products on thin efficacy evidence.

> [!caution] Read the cell source carefully
> Several widely reported "stem cell" trials use embryonic stem cells rather than iPSCs, including prominent Parkinson's and type 1 diabetes programmes. The two behave similarly in the dish but differ entirely in supply chain, ethics and immune matching. Coverage frequently conflates them.

## The autologous problem

The original appeal of iPSCs was a personalised line for every patient. That model has largely collapsed on cost and time. Deriving a clinical-grade line, qualifying it, differentiating it and releasing it takes months and has been estimated at hundreds of thousands of dollars per patient, which is unworkable for an acute indication and marginal even for a chronic one. It is a sharper version of the pricing problem described under [[access-and-inequality]] and in [[somatic-gene-therapy]].

Two responses dominate. Haplobanking assembles lines from donors homozygous across the major HLA loci, so that a modest number of lines can immune-match a large fraction of a population; Japan's national iPSC stock is the furthest advanced. Hypoimmunogenic engineering goes further, knocking out MHC class I and class II presentation and adding a "don't eat me" signal such as CD47 so that a single universal line can be used in anyone.[^deuse2019] The second approach trades immune matching for the risk that the resulting cells are also invisible to immune surveillance of tumours.

## Risks

Residual undifferentiated cells in a graft can form teratomas, which is why release criteria for iPSC products are stringent and why dose is limited. Culture-acquired mutations, particularly in *TP53*, can be selected for during expansion. Differentiation protocols rarely produce pure populations, and off-target cell types in a graft may behave unpredictably. Epigenetic memory of the donor cell type biases differentiation efficiency in ways that vary between lines.

None of these is unique to iPSCs, but the combination means that a line is not a commodity: two lines from the same donor can differ in behaviour, which complicates manufacturing more than the underlying biology suggests.

## Outlook

The technology's near-term value remains in the dish, where it has already changed how human disease is studied. Cell therapy is advancing along the same route that most regenerative medicine takes: small, immune-privileged, structurally simple targets first, with the eye, the pancreas and the substantia nigra ahead of anything resembling a [[lab-grown-organs|whole organ]]. Whether iPSC grafts eventually reduce the [[organ-shortage]] or remain a niche alongside [[xenotransplantation]] depends on manufacturing economics at least as much as on biology.

## See also

- [[yamanaka-factors]]
- [[epigenetic-reprogramming]]
- [[organoids]]
- [[tissue-engineering]]
- [[lab-grown-organs]]
- [[human-cloning]]
- [[in-vitro-gametogenesis]]
- [[shinya-yamanaka]]

## References

[^takahashi2007]: `paper` Takahashi, K. et al. "Induction of Pluripotent Stem Cells from Adult Human Fibroblasts by Defined Factors." *Cell*, 2007.
[^horvath2013]: `paper` Horvath, S. "DNA methylation age of human tissues and cell types." *Genome Biology*, 2013. {The clock was trained to predict chronological age, so a reset reading in reprogrammed cells is a change in a predictor rather than a demonstrated change in the cell.}
[^mandai2017]: `paper` Mandai, M. et al. "Autologous Induced Stem-Cell-Derived Retinal Cells for Macular Degeneration." *New England Journal of Medicine*, 2017. {A single treated patient; the reported outcome is graft survival and disease stabilisation, not improved vision.}
[^deuse2019]: `paper` Deuse, T. et al. "Hypoimmunogenic derivatives of induced pluripotent stem cells evade immune rejection in fully immunocompetent allogeneic recipients." *Nature Biotechnology*, 2019.

==============================================================================
ARTICLE: inflammaging
TITLE:   Inflammaging
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/inflammaging
SOURCE:  https://futurehumanwiki.com/raw/inflammaging
==============================================================================

---
title: "Inflammaging"
slug: "inflammaging"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity"]
tags: ["aging", "inflammation", "immunology", "senescence", "biomarkers", "mechanisms"]
summary: "The chronic, low-grade, sterile inflammation that develops with age in the absence of infection, and its disputed role in frailty and age-related disease."
updated: "2026-07-28"
humanEvidence: "Inflammatory markers rise with age in essentially every human population studied and predict mortality and frailty; the causal human evidence comes from cardiovascular trials of specific anti-inflammatory drugs, not from aging endpoints."
issues: ["The colchicine cardiovascular trials are named but not cited.", "The IL-6 receptor genetic studies are described without a source."]
---

```infobox
{
  "caption": "Hallmark of aging",
  "rows": [
    { "label": "Term coined", "value": "2000" },
    { "label": "Coined by", "value": "Claudio Franceschi" },
    { "label": "Tier", "value": "Integrative hallmark, added 2023" },
    { "label": "Canonical markers", "value": "IL-6, CRP, TNF-α" },
    { "label": "Main proposed sources", "value": "Senescent cells, cytosolic DNA, gut" },
    { "label": "Strongest clinical evidence", "value": "Anti-inflammatory cardiovascular trials" }
  ]
}
```

**Inflammaging** is the chronic, low-grade, systemic inflammation that develops with age without an identifiable infection. It is measured by modestly elevated circulating interleukin-6, C-reactive protein and tumour necrosis factor, levels far below those of acute illness but persistently above those of young adults. Claudio Franceschi introduced the term in 2000 as part of an evolutionary argument: the immune responses that protect against pathogens in early life become maladaptive when sustained across decades.[^franceschi2000] Chronic inflammation was added as a distinct entry to the [[hallmarks-of-aging]] in 2023.

## The observation

Inflammatory markers rise with age in essentially every population studied, and their level predicts outcomes. Interleukin-6 concentration in older adults is associated with subsequent disability, frailty, cognitive decline and all-cause mortality, robustly enough that it is sometimes described as the single most informative blood measure in geriatric epidemiology.[^ferrucci2018] Elevated C-reactive protein carries similar predictive weight for cardiovascular events.

The elevation is small in absolute terms. It is chronic, not episodic; sterile, in that no pathogen is present; and heterogeneous, in that individuals of the same age differ widely. Some very old people show inflammatory profiles resembling those of middle-aged adults, and centenarian cohorts have been reported to combine high pro-inflammatory markers with high anti-inflammatory ones, which Franceschi interpreted as successful counter-regulation rather than absent inflammation.

## Where it comes from

No single source accounts for it. The candidates below are all supported by mechanistic work in animals and by correlational work in humans, and they interact.

**Senescent cells.** The senescence-associated secretory phenotype is an inflammatory programme by construction, and senescent cells accumulate in aged tissue. Transplanting them into young mice raises systemic inflammatory markers, which makes [[cellular-senescence|senescence]] the most directly demonstrated source.

**Cytosolic DNA sensing.** DNA that escapes the nucleus as chromatin fragments, or leaks from damaged mitochondria, is detected in the cytoplasm by cGAS, which activates STING and drives type I interferon and inflammatory cytokine production. This pathway evolved to detect viruses; in aged cells it is triggered by the cell's own genome, connecting [[mitochondrial-dysfunction|mitochondrial damage]] and genome instability to a specific inflammatory output.[^dou2017]

**Retrotransposons.** LINE-1 elements, silenced by heterochromatin in young cells, become de-repressed in senescent and aged cells. Their reverse-transcribed cDNA is a cGAS substrate, and inhibiting reverse transcription reduces interferon signalling and age-associated inflammation in mice.[^dececco2019]

**The gut.** Intestinal permeability increases with age in model organisms and, by indirect measures, in humans, allowing bacterial products such as lipopolysaccharide into the circulation. [[gut-microbiome|Microbiome]] composition shifts with age, though whether the shift causes inflammation or reflects it is unresolved — the same ambiguity that makes dysbiosis the least secure of the twelve hallmarks.

**Immunosenescence.** The thymus involutes from childhood, shrinking the naive T-cell pool. Lifelong cytomegalovirus infection drives large expansions of memory T cells in most of the world's population. The remodelled immune system responds poorly to new antigens while maintaining chronic activation against old ones. Failing clearance of debris by aged macrophages compounds the effect, as does the clonal expansion of mutant blood stem cells described under [[stem-cell-exhaustion|stem cell exhaustion]].

**Body composition and lifestyle.** Visceral adipose tissue is an endocrine organ that secretes IL-6 and recruits inflammatory macrophages, and adiposity increases with age in most populations. Physical inactivity, [[sleep-and-longevity|short or disrupted sleep]] and periodontal disease all raise inflammatory markers independently of age.

> [!debate] Cause or consequence
> The mechanistic case for inflammaging as a driver is strong in mice. In humans it rests largely on association, and inflammatory markers are downstream of nearly everything that goes wrong with an aging body. Sceptics argue that IL-6 is an excellent prognostic indicator precisely because it integrates many forms of damage, which is a different claim from saying that lowering it would help.

## Evidence that it is causal

The best human evidence comes from cardiovascular medicine rather than from geroscience. Genetic studies of variants in the interleukin-6 receptor that mimic pharmacological blockade find lower coronary heart disease risk, supporting a causal role for IL-6 signalling in atherosclerosis. The CANTOS trial then tested this directly: canakinumab, an antibody against interleukin-1β, reduced recurrent cardiovascular events in patients with prior myocardial infarction and elevated CRP, without changing lipid levels.[^ridker2017] The same trial recorded an increase in fatal infection, which is the expected cost of suppressing innate immunity in an older population. Low-dose colchicine has since shown cardiovascular benefit in two large trials and has entered clinical use for that purpose.

These results establish that a specific inflammatory pathway causes a specific age-related disease. They do not establish that inflammation drives aging in general. The counter-example is instructive: low-dose aspirin given to healthy older adults for primary prevention did not extend disability-free survival and increased major bleeding.[^mcneil2018] Broad anti-inflammatory prophylaxis in the well elderly has not been shown to help.

Animal work continues to produce larger effects than human work. Antibody blockade of interleukin-11 started in already-old mice extended median lifespan and improved several functional measures in a 2024 report — a substantial mouse result with no human counterpart, and one whose translation should be assumed uncertain until tested.[^widjaja2024]

## Measurement

Inflammatory markers have been assembled into composite scores intended as [[aging-biomarkers|aging biomarkers]]. An immune-age metric derived from longitudinal high-dimensional immune monitoring predicted cardiovascular outcomes and mortality beyond chronological age, and a separate inflammatory clock built from circulating protein data identified the chemokine CXCL9 as a major contributor to age-related dysfunction.[^sayed2021] These are promising as research tools. None is a validated surrogate endpoint, and none is comparable in predictive performance to a well-constructed [[epigenetic-clock|methylation clock]] across all outcomes.

Population comparisons complicate interpretation. Tsimane forager-horticulturalists in the Bolivian Amazon carry heavy parasite and infection burdens and correspondingly high inflammatory markers, yet have among the lowest levels of coronary artery calcification ever measured in adults.[^kaplan2017] Whatever inflammaging is in an industrialized population, it is not simply "high CRP".

## Interventions

[[exercise-and-aging|Exercise]] lowers IL-6 and CRP modestly and reliably, and is the only intervention with strong human evidence on both inflammation and function. [[caloric-restriction|Caloric restriction]] lowers inflammatory markers in the human trials that have measured them. [[senolytics|Senolytic]] drugs reduce inflammatory markers in mice and in small uncontrolled human studies. [[rapamycin]] and [[metformin]] both have anti-inflammatory effects among their many actions.

The design problem for all of them is that inflammation is not optional. It clears pathogens, resolves tissue damage, and supports repair; the CANTOS infection signal shows what happens when it is suppressed in an older population. Any intervention aimed at inflammaging must therefore target a specific source — senescent cells, cytosolic DNA sensing, a single cytokine — rather than inflammation as such. Which source dominates in an individual older human is, as of 2026, not measurable, and that is the practical obstacle to acting on the [[geroscience-hypothesis|geroscience]] version of the argument.

## See also

- [[hallmarks-of-aging]]
- [[cellular-senescence]]
- [[senolytics]]
- [[aging-biomarkers]]
- [[stem-cell-exhaustion]]
- [[exercise-and-aging]]

## References

[^franceschi2000]: `paper` Franceschi, C. et al. "Inflamm-aging: An evolutionary perspective on immunosenescence." *Annals of the New York Academy of Sciences*, 2000.
[^ferrucci2018]: `paper` Ferrucci, L., Fabbri, E. "Inflammageing: chronic inflammation in ageing, cardiovascular disease, and frailty." *Nature Reviews Cardiology*, 2018.
[^dou2017]: `paper` Dou, Z. et al. "Cytoplasmic chromatin triggers inflammation in senescence and cancer." *Nature*, 2017.
[^dececco2019]: `paper` De Cecco, M. et al. "L1 drives IFN in senescent cells and promotes age-associated inflammation." *Nature*, 2019.
[^ridker2017]: `paper` Ridker, P. M. et al. "Antiinflammatory therapy with canakinumab for atherosclerotic disease." *New England Journal of Medicine*, 2017. {Participants had a previous heart attack and elevated CRP, so the result concerns secondary prevention in an already inflamed population.}
[^mcneil2018]: `paper` McNeil, J. J. et al. "Effect of aspirin on disability-free survival in the healthy elderly." *New England Journal of Medicine*, 2018. {A primary-prevention trial in community-dwelling older adults, a different population from the CANTOS one.}
[^widjaja2024]: `paper` Widjaja, A. A. et al. "Inhibition of IL-11 signalling extends mammalian healthspan and lifespan." *Nature*, 2024. {The lifespan and healthspan measurements are in mice dosed from old age; the paper reports no human data.}
[^sayed2021]: `paper` Sayed, N. et al. "An inflammatory aging clock (iAge) based on deep learning tracks multimorbidity, immunosenescence, frailty and cardiovascular aging." *Nature Aging*, 2021.
[^kaplan2017]: `paper` Kaplan, H. et al. "Coronary atherosclerosis in indigenous South American Tsimane: a cross-sectional cohort study." *The Lancet*, 2017.

==============================================================================
ARTICLE: intelligence-amplification
TITLE:   Intelligence amplification
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/intelligence-amplification
SOURCE:  https://futurehumanwiki.com/raw/intelligence-amplification
==============================================================================

---
title: "Intelligence amplification"
slug: "intelligence-amplification"
type: "concept"
status: "established"
horizon: "present"
categories: ["enhancement", "minds"]
tags: ["cognition", "enhancement", "tools", "collective intelligence", "human-computer interaction", "ai"]
summary: "The project of raising human problem-solving capacity by improving the tools, notations and organisations that thinking runs on rather than by altering the brain."
updated: "2026-07-27"
issues: ["The Flynn effect paragraph states a large empirical claim with no citation."]
---

```infobox
{
  "caption": "Concept in cognitive science and human–computer interaction",
  "rows": [
    { "label": "Also called", "value": "Augmentation, IA" },
    { "label": "Term introduced", "value": "1956, W. Ross Ashby" },
    { "label": "Founding statement", "value": "Engelbart, 1962" },
    { "label": "Contrast with", "value": "Artificial intelligence", "link": "/wiki/artificial-general-intelligence" },
    { "label": "Main routes", "value": "Tools, notations, institutions" },
    { "label": "Status", "value": "Established practice, contested measurement" }
  ]
}
```

**Intelligence amplification** is the strategy of increasing what people can figure out by improving the systems they think with — notations, instruments, software, institutions — rather than by modifying the brain itself. Its proponents argue that essentially all historical gains in human cognitive capability have come this way, and that the biological routes covered elsewhere on this wiki are both harder and less effective by comparison. The idea is old enough to have a settled literature and unsettled enough that nobody agrees on how to measure the result.

## Origins

W. Ross Ashby used the phrase "intelligence amplifier" in the mid-1950s, arguing by analogy with power amplification that a machine could amplify a small amount of human selective judgement into a large amount of problem-solving, provided the human supplied the selection criterion.[^ashby1956] Vannevar Bush had already sketched the practical version a decade earlier in "As We May Think", proposing the memex, a desk-sized associative index that would let a researcher build and share trails through a personal library.[^bush1945]

J. C. R. Licklider gave the position its most-quoted formulation in "Man-Computer Symbiosis", predicting a period in which humans set goals and formulate hypotheses while machines do the routine work that prepares the way for insight, and observing that most of what he did as a researcher was clerical.[^licklider1960]

Douglas Engelbart turned this into a research programme. His 1962 framework treated the human as one component of a system he labelled H-LAM/T — a human using language, artifacts and methodology, in which he is trained — and argued that capability could be raised by improving any component, that improvements interact, and that the highest-leverage target was the process by which capability itself is improved.[^engelbart1962] His group at SRI built the mouse, hypertext, screen editing, outline processing and videoconferencing, demonstrated together in December 1968 in a session later called the Mother of All Demos. Engelbart's disappointment was that the artifacts were adopted and the methodology was not: the industry took the mouse and discarded the bootstrapping argument.

## The augmentation argument

The core claim has three parts.

**Cognition is not confined to the skull.** Andy Clark and David Chalmers argued that when an external resource plays the functional role that a belief would play — reliably available, automatically endorsed, easily accessed — there is no principled reason to place the boundary of the mind at the skin.[^clark1998] Written arithmetic, musical notation and double-entry bookkeeping made possible operations that no unaided brain performs, and they did so without touching neurons.

**The brain is expensive to modify and slow to change.** Pharmacological effects are small, as [[nootropics]] documents; genetic routes are constrained by polygenicity, as [[genetic-enhancement-of-intelligence]] sets out; stimulation devices have a replication problem set out in [[non-invasive-neuromodulation]]; and surgical routes carry risks that healthy people will not accept, which is why [[brain-computer-interface|neural interfaces]] remain clinical devices for paralysis rather than consumer products. Even the most direct internal approach, the hippocampal [[memory-prosthesis]], has produced small effects in a handful of patients after two decades of work. Tools, by contrast, can be revised weekly and distributed at near-zero marginal cost.

**Improvements compound.** A better tool for building tools raises the rate of improvement rather than the level, which is why Engelbart insisted on working on the improvement process itself.

```compare
{
  "columns": ["Amplifying the brain", "Amplifying the toolset"],
  "rows": [
    { "label": "Typical mechanism", "values": ["Drugs, implants, genetics", "Software, notation, training, institutions"] },
    { "label": "Time to deploy", "values": ["Years to decades", "Days to years"] },
    { "label": "Demonstrated effect size", "values": ["Small and task-specific", "Large but hard to attribute"] },
    { "label": "Reversible", "values": ["Sometimes", "Almost always"] },
    { "label": "Distribution", "values": ["Limited by cost and access", "Limited by literacy and infrastructure"] },
    { "label": "Persists without the aid", "values": ["Sometimes", "Rarely"] }
  ]
}
```

## Routes

**Instrumental.** Search, version control, spreadsheets, simulation and statistical software changed which questions are askable; [[ai-protein-design|learned structure prediction]] did the same for questions about proteins. The gains are real and notoriously difficult to isolate, since they show up as changes in what work gets attempted rather than as faster completion of fixed tasks.

**Notational.** Positional numerals, algebraic notation, graphs and diagrams restructure problems so that fewer working-memory slots are needed. This is the oldest form of amplification and the one with the clearest historical record.

**Collective.** Groups can be organised to perform better than their members. Woolley and colleagues reported a general factor of collective intelligence in small groups that predicted performance across tasks and correlated with social sensitivity and equality of turn-taking more than with members' individual scores.[^woolley2010] A later meta-analysis by the same group supported the construct while narrowing the claims made for it. Markets, peer review and open-source development are the large-scale versions, each with characteristic failure modes.

**Environmental.** The Flynn effect — sustained gains in measured IQ across many countries through the twentieth century — is the largest documented change in population cognitive test performance, and it happened without any deliberate enhancement programme. Whether it reflects better nutrition, schooling, smaller families, or increasing familiarity with abstract test formats is unresolved, and several high-income countries have reported stagnation or reversal in recent birth cohorts.

**Machine.** The current version of the argument treats artificial intelligence as an amplifier rather than a replacement, and it is the route on which the ambitions of [[artificial-general-intelligence]] research and of augmentation have converged. [[human-ai-merger]] takes up the stronger form of this claim.

## Evidence from human–machine teams

Chess supplied the first controlled case. After losing to Deep Blue, Garry Kasparov promoted "advanced chess", in which each player uses an engine. His summary of the freestyle tournaments that followed — that a weak player with a machine and a good process beat both strong players with weak processes and strong machines alone — became the standard citation for the centaur thesis. It has aged unevenly: as engines improved, the human contribution in top-level correspondence and freestyle play shrank toward selecting between engine lines.

Field experiments with large language models have produced a more textured picture. A randomised study of professional writing tasks found substantial reductions in time taken and modest gains in rated quality, with the largest benefit to initially weaker writers.[^noy2023] A field experiment with customer-support agents found average productivity gains concentrated among novices, with little effect on the most experienced staff.[^brynjolfsson2023] A large study of management consultants found improvement on tasks inside the model's competence and degraded accuracy on a task designed to sit just outside it, where participants accepted plausible but wrong output.[^dellacqua2023]

> [!key] The pattern across studies
> Machine assistance compresses the distribution: it raises the floor more than the ceiling, and it converts some skill differences into differences in the ability to judge when the tool is wrong. That is a different kind of amplification from the one Engelbart described, and it makes verification the scarce capability.

## Criticism

**Attribution.** No one has shown how to measure amplification cleanly. Productivity statistics have not tracked the arrival of transformative information tools in the way naive extrapolation predicted, a puzzle economists have argued over since Robert Solow's remark about computers appearing everywhere but in the productivity statistics. The same measurement problem undermines the trend extrapolations discussed in [[accelerating-change]].

**Offloading and dependency.** Sparrow and colleagues found that people are less likely to remember information they expect to be able to look up, and more likely to remember where to find it.[^sparrow2011] Whether that constitutes a loss or an efficient reallocation is disputed; the practical concern is that amplification which vanishes when the tool is removed has not changed the person at all.

**Homogenisation.** Tools that are used by everyone impose a common structure on thought. Shared search rankings and shared model outputs narrow the diversity of approaches that collective intelligence depends on, which is a mechanism by which amplification at the individual level could reduce it at the group level.

**Displacement of the goal.** Critics of the framing note that "intelligence" in this literature means whatever the benchmark measures. Amplifying performance on measurable tasks is not the same as amplifying judgement, and the historical cases most often cited — notation, printing, statistics — changed what counted as good reasoning rather than doing more of the old kind faster.

## Relation to the rest of the field

Intelligence amplification is often presented as the safe alternative to building autonomous machine intelligence, on the grounds that a system with a human in the loop inherits human goals — a position with roots in [[transhumanism]] and a direct bearing on the arguments in [[existential-risk]]. [[nick-bostrom]] treats biological enhancement, neural interfaces and better institutions as alternative paths to greater collective capability, and notes that they are slow compared with the machine path, which is part of the argument in [[technological-singularity]] and in [[differential-technological-development]]. Critics of the safe-alternative framing observe that an amplified human is not obviously safer than an automated system if the amplification is doing the cognitive work and the human is supplying only approval.

The unresolved empirical question is whether the compression effect seen in the assistance studies persists. If tools continue to raise the floor faster than the ceiling, the practical meaning of enhancement changes from making exceptional people more exceptional to making expertise less scarce — which is a different social outcome from the one that both proponents and critics of [[human-enhancement]] have generally argued about.

## See also

- [[human-enhancement]]
- [[human-ai-merger]]
- [[artificial-general-intelligence]]
- [[nootropics]]
- [[brain-computer-interface]]
- [[genetic-enhancement-of-intelligence]]
- [[technological-singularity]]
- [[accelerating-change]]

## References

[^ashby1956]: `book` Ashby, W. R. "Design for an Intelligence-Amplifier." In Shannon, C. E. and McCarthy, J. (eds), *Automata Studies*. Princeton University Press, 1956.
[^bush1945]: `paper` Bush, V. "As We May Think." *The Atlantic Monthly*, 1945.
[^licklider1960]: `paper` Licklider, J. C. R. "Man-Computer Symbiosis." *IRE Transactions on Human Factors in Electronics*, 1960.
[^engelbart1962]: `report` Engelbart, D. C. *Augmenting Human Intellect: A Conceptual Framework*. Stanford Research Institute, 1962.
[^clark1998]: `paper` Clark, A. and Chalmers, D. "The Extended Mind." *Analysis*, 1998.
[^woolley2010]: `paper` Woolley, A. W., Chabris, C. F., Pentland, A., Hashmi, N. and Malone, T. W. "Evidence for a Collective Intelligence Factor in the Performance of Human Groups." *Science*, 2010. See also Riedl, C., Kim, Y. J., Gupta, P., Malone, T. W. and Woolley, A. W. "Quantifying collective intelligence in human groups." *Proceedings of the National Academy of Sciences*, 2021.
[^noy2023]: `paper` Noy, S. and Zhang, W. "Experimental evidence on the productivity effects of generative artificial intelligence." *Science*, 2023.
[^brynjolfsson2023]: `preprint` Brynjolfsson, E., Li, D. and Raymond, L. "Generative AI at Work." National Bureau of Economic Research working paper, 2023. {A staggered rollout inside one company's customer-support operation, so the setting is a single firm and one narrowly scripted kind of work.}
[^dellacqua2023]: `preprint` Dell'Acqua, F. et al. "Navigating the Jagged Technological Frontier." Harvard Business School working paper, 2023. {A working paper rather than a reviewed study; the task said to lie outside the model's competence was constructed by the researchers.}
[^sparrow2011]: `paper` Sparrow, B., Liu, J. and Wegner, D. M. "Google Effects on Memory: Cognitive Consequences of Having Information at Our Fingertips." *Science*, 2011.

==============================================================================
ARTICLE: jennifer-doudna
TITLE:   Jennifer Doudna
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/jennifer-doudna
SOURCE:  https://futurehumanwiki.com/raw/jennifer-doudna
==============================================================================

---
title: "Jennifer Doudna"
slug: "jennifer-doudna"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "genetics"]
tags: ["crispr", "gene editing", "rna", "nobel prize", "governance", "biochemistry"]
summary: "American biochemist who co-authored the 2012 paper showing CRISPR-Cas9 could be programmed to cut chosen DNA sequences, and shared the 2020 Nobel Prize in Chemistry."
updated: "2026-07-28"
issues: ["The patent-dispute section makes checkable claims about specific rulings with no citation."]
---

```infobox
{
  "caption": "Biochemist",
  "rows": [
    { "label": "Born", "value": "19 February 1964, Washington, D.C." },
    { "label": "Nationality", "value": "American" },
    { "label": "Education", "value": "PhD, Harvard University, 1989" },
    { "label": "Known for", "value": "Programmable CRISPR–Cas9", "link": "/wiki/crispr-cas9" },
    { "label": "Field", "value": "RNA biochemistry; structural biology" },
    { "label": "Affiliation", "value": "UC Berkeley; Innovative Genomics Institute" },
    { "label": "Award", "value": "Nobel Prize in Chemistry, 2020" }
  ]
}
```

**Jennifer Doudna** is an American biochemist and structural biologist who, with Emmanuelle
[[emmanuelle-charpentier|Charpentier]], demonstrated in 2012 that the bacterial CRISPR–Cas9 system could be reduced to two
components and directed to cut any chosen DNA sequence by a single engineered guide RNA. The result
turned a bacterial immune mechanism into a general-purpose editing tool, and she shared the 2020
Nobel Prize in Chemistry for it. She has since become one of the field's most active voices on
governance, particularly on the line between somatic and heritable editing.

## Career

Doudna grew up in Hilo, Hawaii, took a degree in biochemistry at Pomona College, and completed a
doctorate at Harvard in 1989 with Jack Szostak, working on self-replicating RNA. As a postdoctoral
researcher with Thomas Cech in Colorado she solved the crystal structure of a large catalytic RNA
domain, one of the first high-resolution views of RNA folding into a defined three-dimensional
enzyme.[^cate1996] That combination — RNA as an information carrier and RNA as a machine — set up
everything that followed. She joined Yale in 1994 and moved to the University of California,
Berkeley, in 2002.

Her laboratory reached CRISPR through structural work on RNA-guided processes, not through
microbiology. Bacterial CRISPR arrays had been described since the late 1980s and their role in
adaptive immunity established by the mid-2000s; what remained was to show that the system could be
reprogrammed. See [[crispr-cas9]] for the full mechanism.

## The 2012 result

The paper published with Martin Jinek, Krzysztof Chylinski, Ines Fonfara, Michael Hauer and
Charpentier established three things.[^jinek2012] Cas9 is the sole protein required for
target cleavage in the type II system. It needs two RNAs — the CRISPR RNA and a trans-activating
RNA — which can be fused into a single chimeric guide. And the target is specified by base pairing
with that guide, subject to an adjacent short motif in the DNA. Changing twenty nucleotides of the
guide changes the target.

> [!key] Why this was the pivot
> Programmable nucleases already existed. Zinc-finger proteins and TALENs could be targeted, but
> each new target required designing and building a new protein — weeks of work per site. Cas9
> moved targeting from protein engineering to ordering an oligonucleotide.

Within months, several groups reported Cas9 editing in human cells, among them teams led by Feng
Zhang and by [[george-church]]. The tool has since been elaborated into forms that avoid cutting
both DNA strands at all — [[base-editing]], [[prime-editing]] and [[epigenome-editing]] — partly in
response to the [[crispr-off-target-effects|off-target and large-deletion problems]] that
double-strand breaks create.

Doudna's own subsequent work has centred on the structural biology of Cas enzymes, anti-CRISPR
proteins, the collateral cleavage activity of Cas12 and Cas13 that underpins CRISPR diagnostics, and
delivery — which she has repeatedly identified as the field's binding constraint rather than the
editing chemistry itself. See [[lipid-nanoparticles]] and [[aav-vectors]].

## Institutions and companies

Doudna founded the Innovative Genomics Institute in 2014, a joint Berkeley–UCSF centre whose
programmes include sickle cell disease, agricultural editing, and climate applications, and which
ran a high-throughput COVID-19 testing laboratory in 2020. She has co-founded several companies,
among them Caribou Biosciences, Intellia Therapeutics, Mammoth Biosciences and Scribe Therapeutics;
she was a founder of Editas Medicine and left it early.

The intellectual-property dispute over CRISPR is unusually consequential. Berkeley and its partners
filed first; the Broad Institute, where Feng Zhang's group demonstrated editing in eukaryotic cells,
filed later under an accelerated route and was granted patents covering that use. US interference
proceedings have repeatedly favoured the Broad on eukaryotic claims, while European decisions have
favoured the Berkeley group; appeals have continued into the mid-2020s and the American position
remains unsettled as of 2026. The scientific credit and the patent position have diverged, which is
part of why the case is taught in technology-transfer courses.

## Governance

Doudna convened a meeting in Napa, California, in January 2015 that produced a *Science* commentary
calling for a moratorium on clinical use of heritable human genome editing while the science and
ethics were worked out.[^baltimore2015] Participants explicitly invoked the
[[asilomar-conference|1975 Asilomar meeting]] on recombinant DNA as a template, an analogy critics
have questioned on the grounds that the 1975 participants controlled the entire relevant field and
the 2015 ones did not. She has been involved in the international summit process since, and
condemned the [[he-jiankui-affair|He Jiankui experiment]] in 2018 as reckless and premature.

Her position has been consistent and narrower than either extreme:
[[somatic-gene-therapy|somatic editing]] is ordinary medicine subject to ordinary safety review, and
[[germline-editing|heritable editing]] should not
proceed clinically until safety, need and societal agreement are all established. She has written
about her own unease at having made the tool available, most directly in *A Crack in Creation*
(2017) with Samuel Sternberg.[^crack2017] She has also been a persistent critic of the cost
structure of approved gene therapies, arguing that a treatment such as [[casgevy]] is of little use
to the majority of sickle cell patients, who live in countries where the price and the
transplant infrastructure are both out of reach.

> [!debate] Who counts as an inventor
> Public credit for CRISPR is contested well beyond Doudna and Charpentier — Francisco Mojica,
> Rodolphe Barrangou, Philippe Horvath, Virginijus Šikšnys and Feng Zhang all made claims to
> foundational contributions. The 2020 Nobel recognized two people for a discovery with a long
> chain of contributors, a familiar structural problem with the prize.

## Reception and legacy

Doudna is among the most decorated living biochemists, and CRISPR is now the default laboratory
technique for altering a genome in essentially any organism. The clinical translation has been
slower than the 2012 excitement implied: as of 2026 a small number of CRISPR-based medicines are
approved or in late-stage trials, most of them ex vivo cell therapies or liver-targeted in vivo
treatments, and the delivery problem she flagged early remains the reason the list is short.

Her governance work is harder to evaluate. The moratorium she called for held in the sense that no
credible second germline case has been publicly confirmed since 2018, but it held through
professional norms, national law and reputational cost rather than through any enforceable
international instrument; see [[governance-of-genome-editing]]. Whether that arrangement survives
the arrival of cheaper editing and stronger commercial incentives around
[[polygenic-embryo-screening|embryo selection]] is an open question, and one she has said the
scientific community is not organized to answer alone.

## See also

- [[crispr-cas9]]
- [[base-editing]]
- [[casgevy]]
- [[germline-editing]]
- [[he-jiankui-affair]]
- [[governance-of-genome-editing]]
- [[george-church]]
- [[crispr-off-target-effects]]

## References

[^cate1996]: `paper` Cate, J. H. et al. "Crystal structure of a group I ribozyme domain: principles of RNA packing." *Science*, 1996.
[^jinek2012]: `paper` Jinek, M., Chylinski, K., Fonfara, I., Hauer, M., Doudna, J. A. and Charpentier, E. "A Programmable Dual-RNA-Guided DNA Endonuclease in Adaptive Bacterial Immunity." *Science*, 2012. {The demonstration was biochemical, with purified components cutting DNA in a tube; editing inside human cells was reported by other groups months later.}
[^baltimore2015]: `paper` Baltimore, D. et al. "A prudent path forward for genomic engineering and germline gene modification." *Science*, 2015. {A commentary by the meeting's participants urging that clinical germline use be discouraged while research continued; it carried no legal force.}
[^crack2017]: `book` Doudna, J. A. and Sternberg, S. H. *A Crack in Creation: Gene Editing and the Unthinkable Power to Control Evolution*. Houghton Mifflin Harcourt, 2017. {A first-person account by one of the principals, so it records Doudna's own reading of the history rather than an independent one.}

==============================================================================
ARTICLE: julian-savulescu
TITLE:   Julian Savulescu
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/julian-savulescu
SOURCE:  https://futurehumanwiki.com/raw/julian-savulescu
==============================================================================

---
title: "Julian Savulescu"
slug: "julian-savulescu"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "society"]
tags: ["bioethics", "enhancement", "embryo selection", "moral enhancement", "philosophy"]
summary: "Australian bioethicist who formulated procreative beneficence and, with Ingmar Persson, the argument for moral bioenhancement; the leading academic advocate of human enhancement."
updated: "2026-07-27"
issues: ["Reception section calls him among the most cited living bioethicists with no source."]
---

```infobox
{
  "caption": "Bioethicist",
  "rows": [
    { "label": "Born", "value": "22 December 1963, Melbourne" },
    { "label": "Nationality", "value": "Australian" },
    { "label": "Education", "value": "MBBS and PhD, Monash University" },
    { "label": "Known for", "value": "Procreative beneficence", "link": "/wiki/procreative-beneficence" },
    { "label": "Field", "value": "Practical ethics" },
    { "label": "Affiliation", "value": "Oxford Uehiro Centre; National University of Singapore" },
    { "label": "Doctoral adviser", "value": "Peter Singer" }
  ]
}
```

**Julian Savulescu** is an Australian bioethicist who holds that parents have a moral reason to
select the children expected to have the best lives, that enhancement is continuous with medicine
rather than opposed to it, and that humanity's moral capacities are now the binding constraint on
its technological ones. Trained as a physician and as a philosopher under Peter Singer, he has been
the most systematic academic defender of [[human-enhancement]] since the early 2000s, and the
figure whose arguments [[bioconservatism|bioconservative]] critics most often answer.

## Career

Savulescu qualified in medicine at Monash University and completed a doctorate there in philosophy,
working on rationality and end-of-life decision-making. He took the Uehiro Chair in Practical Ethics
at Oxford in 2002 and founded and directed the Oxford Uehiro Centre for Practical Ethics, which
became the main European base for analytic work on enhancement, neuroethics and research ethics. He
edited the *Journal of Medical Ethics* for close to two decades. Since 2022 he has also held a
chair in medical ethics at the National University of Singapore, directing its Centre for
Biomedical Ethics, and he leads a biomedical ethics group at the Murdoch Children's Research
Institute in Melbourne.

His output is unusually broad for the field: embryo selection, doping, organ donation, conscientious
objection by clinicians, the ethics of research in emergencies, and during the pandemic a prominent
argument for human challenge trials of vaccines.

## Procreative beneficence

The 2001 paper that made his name states a principle: couples who select embryos or gametes should
select the child, of the possible children they could have, expected to have the best life, or at
least as good a life as the others, on the basis of relevant available information.[^savulescu2001]
The principle is deliberately narrow. It applies only where selection is already occurring, it does
not license coercion, and it concerns non-person-affecting choices — the child selected is a
different individual from the one who would otherwise have existed, which is why the
[[procreative-beneficence|non-identity problem]] does so much work in the surrounding literature.

Savulescu extends the principle beyond disease traits to any characteristic that improves expected
wellbeing, which is where the disagreement begins. If a test predicts a small advantage in some
trait, the principle appears to make selection on it obligatory rather than optional — a conclusion
critics find either objectionable or a reductio. The argument has become directly practical with
the arrival of [[polygenic-embryo-screening]], which produces exactly the sort of probabilistic
trait information the principle takes as input; Savulescu has argued that the resulting obligations
are modest because the predicted gains are small, while objecting to bans on the technology as such.

> [!debate] The strongest objection
> Disability-rights scholars argue that selecting against a trait expresses a judgement about people
> living with it. Savulescu's reply relies on a welfarist account of disability — a condition is a
> disability if it reduces expected wellbeing in the circumstances, not because it deviates from a
> norm — which critics say relocates the judgement rather than removing it. See
> [[disability-rights-and-enhancement]].

## Moral bioenhancement

With Ingmar Persson, Savulescu argues that human moral psychology evolved for small groups with
limited technology, and is poorly suited to a world where a few individuals can cause catastrophic
harm.[^ps2008] Cognitive enhancement in this framing is dangerous on its own, because it increases
the capacity for harm without improving the disposition to avoid it. The proposed response is
research into biomedical means of strengthening altruism and a sense of justice — the illustrative
candidates in the literature being oxytocin, serotonergic drugs and beta-blockers, none of which
does anything close to what the argument requires. Their 2012 book *Unfit for the Future* connects
the argument to [[existential-risk]].

The obvious objections have been pressed hard. John Harris argued that moral enhancement, if
effective, would remove the freedom to do wrong that makes moral action valuable.[^harris2011]
Others ask who specifies the target morality, and note that any state with the capacity to
administer moral enhancement is the last actor that should be trusted with it. Savulescu and
Persson's response distinguishes enhancing dispositions from determining choices, and accepts that
compulsory application would be illegitimate. The empirical position is weaker than the
philosophical one: no intervention has been shown to produce durable, general improvements in moral
behaviour. See [[moral-enhancement]].

## Sport and doping

Savulescu has argued that the anti-doping regime is incoherent — it permits altitude tents,
hypoxic chambers and extensive pharmacological recovery while banning functionally similar
interventions — and that a safety-based regulatory model, monitoring physiological parameters
rather than prohibiting substances, would protect athletes better than prohibition does.[^sav2004]
The position is a minority one among sports physicians and anti-doping bodies, and it has become
more salient with the emergence of openly enhanced competition formats. See
[[enhancement-in-sport]] and [[gene-doping]].

## Reception

Savulescu is among the most cited living bioethicists, and the Uehiro Centre trained a generation of
researchers now working on neuroethics, AI ethics and enhancement. His method is characteristically
analytic: state a principle, follow it to conclusions most readers dislike, and treat the discomfort
as data about the principle rather than a reason to abandon it. Admirers regard this as intellectual
seriousness. Critics regard the same trait as a willingness to reach eugenic-sounding conclusions
through formally valid steps, and note that his arguments almost always assume a regulatory and
economic environment in which access is fair — an assumption addressed at length in
[[access-and-inequality]].

He is on the opposite side of most debates from [[leon-kass]], and the two positions define the
range of serious argument in [[bioethics-of-enhancement]]: whether the given human form has moral
weight in itself, or whether wellbeing is the only thing that finally matters.

## Legacy

The practical test of Savulescu's programme is arriving now. [[embryo-selection]] on polygenic
scores is commercially available and unregulated in several jurisdictions;
[[in-vitro-gametogenesis]], if it works in humans, would remove the constraint that a cycle yields
only a handful of embryos and would make the principle of procreative beneficence bind much harder.
Savulescu has been consistent that the resulting decisions belong to prospective parents rather than
states. Whether a technology that makes his principle powerful also makes it acceptable is a
question his framework poses more sharply than it answers.

## See also

- [[procreative-beneficence]]
- [[bioethics-of-enhancement]]
- [[moral-enhancement]]
- [[polygenic-embryo-screening]]
- [[disability-rights-and-enhancement]]
- [[enhancement-in-sport]]
- [[leon-kass]]
- [[nick-bostrom]]

## References

[^savulescu2001]: `paper` Savulescu, J. "Procreative Beneficence: Why We Should Select the Best Children." *Bioethics*, 2001.
[^ps2008]: `paper` Persson, I. and Savulescu, J. "The Perils of Cognitive Enhancement and the Urgent Imperative to Enhance the Moral Character of Humanity." *Journal of Applied Philosophy*, 2008. {A philosophical argument for a research direction; it reports no data on any intervention that alters moral behaviour.}
[^harris2011]: `paper` Harris, J. "Moral Enhancement and Freedom." *Bioethics*, 2011.
[^sav2004]: `paper` Savulescu, J., Foddy, B. and Clayton, M. "Why we should allow performance enhancing drugs in sport." *British Journal of Sports Medicine*, 2004. {An argued policy position in a medical journal, not a study of what doping or its prohibition does to athletes.}

==============================================================================
ARTICLE: lab-grown-organs
TITLE:   Lab-grown organs
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/lab-grown-organs
SOURCE:  https://futurehumanwiki.com/raw/lab-grown-organs
==============================================================================

---
title: "Lab-grown organs"
slug: "lab-grown-organs"
type: "technology"
status: "experimental"
horizon: "2040s"
trl: 4
categories: ["bodies"]
tags: ["regenerative medicine", "transplantation", "stem cells", "chimeras", "immunology", "organ shortage"]
summary: "The attempt to produce whole transplantable organs from cells rather than donors, currently limited to simple tissues, islet clusters and organs grown inside host animals."
updated: "2026-07-27"
humanEvidence: "Engineered skin, corneal and thymus tissue are approved and implanted in patients, and small trials of stem-cell-derived islets have produced insulin independence; no solid vascularized organ has been transplanted into a person."
access: "Only the simple tissues can be obtained: skin, limbal and thymus products are approved in some countries, stem-cell-derived islets are trial-only, and no solid organ is purchasable at any price."
reversibility: "difficult"
issues: ["The islet insulin-independence results are stated without a citation.", "The 2025 hypoimmune islet case is described without a source."]
---

```infobox
{
  "caption": "Regenerative medicine goal",
  "rows": [
    { "label": "Main routes", "value": "Scaffold, chimera, in situ" },
    { "label": "Cell sources", "value": "Autologous, allogeneic, iPSC" },
    { "label": "Approved products", "value": "Thymus tissue, skin, cornea" },
    { "label": "Solid organs transplanted", "value": "None" },
    { "label": "Longest ex vivo organ function", "value": "Hours, in rodents" },
    { "label": "Chief obstacles", "value": "Vasculature, cell number, immunity" },
    { "label": "Status", "value": "Preclinical for solid organs" }
  ]
}
```

**Lab-grown organs** are transplantable organs produced from cells rather than taken from a donor. The category is defined by ambition rather than method: it includes organs built on a scaffold, organs grown inside a host animal, and organs regenerated in place inside the patient. As of 2026 no solid vascularized organ made by any of these routes has been transplanted into a human. What has reached patients is a narrower set of engineered tissues, and understanding why the boundary falls where it does is more useful than any timeline.

## What counts

The functional distinction is between organs that are essentially sheets or clusters and organs that are perfused three-dimensional structures with an arterial inlet and a venous outlet.

The first category has clinical products. Cultured epidermal grafts, engineered skin, cultured limbal epithelium for corneal repair, and allogeneic cultured thymus tissue for children born without a thymus are all approved somewhere. Stem-cell-derived pancreatic islet clusters, which are [[organoids]] in construction if not in name, have produced insulin independence in most participants of small type 1 diabetes trials, using standard immunosuppression — arguably the first case of a lab-grown endocrine organ doing an organ's job in a person, even though it is delivered as a suspension of cell clusters into the portal vein rather than as an anatomical pancreas.

The second category — kidney, liver, heart, lung — has nothing. Each requires on the order of ten to a hundred billion cells of several types, arranged around a branching vascular tree fine enough that no cell sits more than a couple of hundred micrometres from a capillary, with mechanical properties that survive arterial pressure and a surface that does not clot.

## Route one: build it

The constructive route seeds cells onto a scaffold and cultures the result. Purely synthetic scaffolds have worked for hollow, low-metabolic-demand structures — bladder, urethra, vaginal tissue — and not for solid organs, for the reasons set out in [[tissue-engineering]] and [[organ-bioprinting]]. The uterus falls on the far side of that boundary, being muscular and heavily perfused, which is why the only route to a working one remains a donated organ and the procedure described in [[uterus-transplantation]].

The variant with the strongest results uses a donor organ stripped of its cells as the scaffold, retaining the vascular tree in place of trying to fabricate one. Perfusing a detergent through the vasculature of a rat heart leaves an acellular matrix that can be reseeded with cardiac cells and stimulated to beat; the same approach has been applied to rodent lung and kidney, with reseeded kidneys producing dilute urine when perfused.[^ott2008][^song2013] These constructs functioned at a small percentage of native capacity for hours. Scaling the approach to human-sized organs runs into incomplete re-endothelialization, which causes thrombosis, and into the sheer number of cells required. This route is treated in detail in [[decellularized-scaffolds]].

## Route two: grow it in an animal

Interspecies blastocyst complementation exploits developmental vacancy. If an embryo is genetically unable to form a particular organ — a condition created by knocking out a master regulator gene with [[crispr-cas9]] or an earlier editing method — and cells from another animal are injected at the blastocyst stage, the donor cells fill the empty niche and build the organ. Rat cells injected into mouse embryos lacking *Pdx1* form a rat pancreas inside the mouse; islets grown this way and transplanted into diabetic mice corrected their diabetes with only transient immunosuppression.[^kobayashi2010][^yamaguchi2017]

Extending this to human cells in pigs is far harder. Human pluripotent cells contribute at extremely low rates to pig embryos, and most undergo apoptosis; early chimera experiments detected human contributions only as a vanishingly small fraction of the embryo's cells.[^wu2017] Anti-apoptotic modification of the human cells and better matching of developmental stage have improved this, and a 2023 report described pig embryos carrying substantially humanized mesonephroi at around four weeks of gestation.[^wang2023] Japan lifted its prohibition on gestating human-animal chimeric embryos in 2019, and other jurisdictions have taken varied positions. The oversight questions resemble those raised by [[synthetic-embryos]]: both involve entities whose moral status depends on a definitional judgement that regulators have not settled.

Three problems remain. The organ's blood vessels and connective tissue are typically host-derived even when the parenchyma is human, so a "human" kidney grown in a pig presents pig endothelium to the recipient's immune system — which is the same barrier that [[xenotransplantation]] addresses by editing the donor animal, and which suggests the two approaches converge rather than compete. Donor cells cannot be confined to the target organ without additional engineering, and contribution to the host's brain or germ line is the ethical trigger point that most regulators focus on. And gestating a human-scale organ in a pig requires a target organ of appropriate size and a host that survives to term.

## Route three: grow it in place

The least discussed route uses the patient's own body as the bioreactor. The liver already regenerates, and the field's interest is in extending that capacity: transdifferentiating hepatocytes or biliary cells, delivering growth signals to a damaged kidney, or implanting a small mass of cells that expands in situ. Ectopic organogenesis — growing a functional secondary organ in the omentum or lymph node, where blood supply is abundant — has produced functioning ectopic liver tissue in animals.

In situ approaches sidestep vascularization and cell manufacturing at once, because the host supplies both. They cannot replace an organ that has been removed, and they depend on regenerative capacity that mammals largely lack outside the liver, a limitation discussed in [[limb-regeneration]].

> [!key] The vascular tree is the organ
> Across all three routes, the recurring finding is that parenchymal cells are the easy part. Hepatocytes, cardiomyocytes and nephron progenitors can be made in quantity from [[induced-pluripotent-stem-cells]]. What cannot be made is the branching, endothelialized, non-thrombogenic vasculature that keeps them alive, which is why the most promising strategies borrow a vascular tree rather than build one.

## The immunology of a grown organ

A lab-grown organ is not automatically an immunologically silent one.

An autologous organ, made from the patient's own reprogrammed cells, should escape rejection, but takes months to produce, costs a great deal per patient, and inherits the patient's genome — including whatever mutation caused the organ failure, unless it is corrected by [[somatic-gene-therapy]] first. Autologous manufacture is also incompatible with acute failure, where the patient has days, and a bespoke product priced per patient raises the distribution questions collected under [[access-and-inequality]].

An allogeneic organ made from a banked cell line is manufacturable and immediately available, and needs immunosuppression like any transplant. The route being pursued to avoid that is hypoimmune engineering: deleting the genes required for surface expression of class I and class II major histocompatibility complex, and overexpressing CD47 so that natural killer cells and macrophages do not attack the resulting cells for lacking MHC. The approach was established in mice and in human cells before it reached patients.[^deuse2019] A 2025 report described hypoimmune-modified donor islet cells surviving and secreting insulin in a person with type 1 diabetes who received no immunosuppression, at short follow-up. That is early single-patient evidence rather than a demonstration of durable tolerance. Whether cloaked cells remain safe over years is unresolved, since the same modifications that hide a transplant from immune surveillance would also hide a tumour arising from it.

## Outlook

The field's realistic near-term output is not organs but organ substitutes: islet products, engineered corneal and skin tissue, bioengineered vascular grafts, cultured red cells of the kind described in [[artificial-blood]], and external assist devices built from decellularized animal organs reseeded with human cells. For the heart, the working substitute remains mechanical, as set out in [[artificial-heart]]. For kidneys and livers, the supply pressure described in [[organ-shortage]] is being addressed by machine perfusion and donation policy, and experimentally by edited pigs, whose grafts have so far all failed or been removed within months; each has still moved faster than growing organs from cells.

The question worth watching is not when a lab-grown kidney is implanted but whether any group can keep a human-scale bioengineered organ perfused and functional in a large animal for a month. That threshold has not been crossed by any route, and until it is, the difference between a grown organ and a research construct remains the difference between a circulatory system and a scaffold.

## See also

- [[organ-bioprinting]]
- [[decellularized-scaffolds]]
- [[tissue-engineering]]
- [[organoids]]
- [[xenotransplantation]]
- [[organ-shortage]]
- [[induced-pluripotent-stem-cells]]
- [[artificial-heart]]

## References

[^ott2008]: `paper` Ott, H. C. et al. "Perfusion-decellularized matrix: using nature's platform to engineer a bioartificial heart." *Nature Medicine*, 2008. {A rat heart reseeded with rat cells, generating a small fraction of normal pump function for hours in a bioreactor rather than in an animal.}
[^song2013]: `paper` Song, J. J. et al. "Regeneration and experimental orthotopic transplantation of a bioengineered kidney." *Nature Medicine*, 2013.
[^kobayashi2010]: `paper` Kobayashi, T. et al. "Generation of rat pancreas in mouse by interspecific blastocyst injection of pluripotent stem cells." *Cell*, 2010.
[^yamaguchi2017]: `paper` Yamaguchi, T. et al. "Interspecies organogenesis generates autologous functional islets." *Nature*, 2017.
[^wu2017]: `paper` Wu, J. et al. "Interspecies chimerism with mammalian pluripotent stem cells." *Cell*, 2017. {Human cells contributed at very low frequency and the chimeric pig embryos were assessed in early gestation, not carried to term.}
[^wang2023]: `paper` Wang, J. et al. "Generation of a humanized mesonephros in pigs from induced pluripotent stem cells via embryo complementation." *Cell Stem Cell*, 2023.
[^deuse2019]: `paper` Deuse, T. et al. "Hypoimmunogenic derivatives of induced pluripotent stem cells evade immune rejection in fully immunocompetent allogeneic recipients." *Nature Biotechnology*, 2019. {Immune evasion was demonstrated in mice and in cultured human cells; nothing here tests whether cloaked cells stay safe over years in a person.}

==============================================================================
ARTICLE: leon-kass
TITLE:   Leon Kass
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/leon-kass
SOURCE:  https://futurehumanwiki.com/raw/leon-kass
==============================================================================

---
title: "Leon Kass"
slug: "leon-kass"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "society"]
tags: ["bioethics", "bioconservatism", "human dignity", "cloning", "mortality"]
summary: "American physician and bioethicist who chaired the President's Council on Bioethics and made the strongest philosophical case that mortality gives human life its meaning."
updated: "2026-07-27"
issues: ["The 2004 removal of two council members is described without a source."]
---

```infobox
{
  "caption": "Physician and bioethicist",
  "rows": [
    { "label": "Born", "value": "12 February 1939, Chicago" },
    { "label": "Nationality", "value": "American" },
    { "label": "Education", "value": "MD Chicago (1962); PhD Harvard (1967)" },
    { "label": "Known for", "value": "The wisdom of repugnance", "link": "/wiki/bioconservatism" },
    { "label": "Field", "value": "Bioethics; political philosophy" },
    { "label": "Role", "value": "Chair, President's Council on Bioethics, 2001–2005" },
    { "label": "Affiliation", "value": "University of Chicago; American Enterprise Institute" }
  ]
}
```

**Leon Kass** is an American physician, biochemist and bioethicist who argues that biotechnology's
central danger is not accident or inequity but the erosion of the goods that make a human life
worth living — and that mortality, effort and the unchosen character of what is given to us are
among those goods. He chaired the President's Council on Bioethics from 2001 to 2005 and is the
most philosophically substantial exponent of [[bioconservatism]].

## Career

Kass trained as a physician at the University of Chicago and took a doctorate in biochemistry at
Harvard, working afterwards at the National Institutes of Health. He spent a summer in Mississippi
during the civil rights movement, an experience he has described as formative in a different
direction than expected: it made him doubt that scientific and material progress carries moral
progress along with it. Reading Huxley's *Brave New World* alongside the emerging debates over
reproductive technology completed the turn, and by the early 1970s he had left the laboratory for
the study of ethics and political philosophy.

He served as executive secretary of a National Academy of Sciences committee on the life sciences
and social policy, taught at St John's College in Annapolis, and joined the University of Chicago's
Committee on Social Thought in 1976, where he taught for more than three decades. He has been a
fellow at the American Enterprise Institute. His books range beyond bioethics into readings of
Genesis and a study of the ethics of eating, and this is not incidental: his method treats the
meaning of ordinary human activities as the material from which bioethical judgement is built.

## Key arguments

### The new biology and the price of relieving man's estate

Kass's 1971 article in *Science* is the origin point of his position and, arguably, of American
bioethics as a distinct enterprise.[^kass1971] He asked what would follow from technologies then
just visible — in vitro fertilization, prenatal diagnosis, and the eventual prospect of
[[germline-editing|heritable genetic modification]] — and argued that each would be adopted for compelling humanitarian reasons while cumulatively changing
the relations of parent to child and physician to patient. He was writing seven years before the
first IVF birth. Several of the specific predictions did not hold; the structural claim, that
each individual step is defensible and the aggregate is not chosen by anyone, has worn well and
recurs throughout debates over [[polygenic-embryo-screening]] and [[designer-babies]].

### The wisdom of repugnance

In a 1997 essay written in response to the cloning of Dolly, Kass argued that visceral revulsion at
certain practices can carry moral knowledge that argument has not yet articulated.[^kass1997]
Repugnance, on this account, is "the emotional expression of deep wisdom, beyond reason's power
fully to articulate it" — a signal that something is being violated even when the violation resists
formal statement. Applied to [[human-cloning]], the argument holds that the widespread reaction
against it registers real damage to the meanings of parenthood, individuality and generational
succession.

> [!debate] The obvious objection
> Repugnance has historically attached to interracial marriage, homosexuality, dissection and
> vaccination. Critics including Martha Nussbaum, Arthur Caplan and Steven Pinker argue that
> disgust is a poor moral instrument precisely because it is so easily attached to the unfamiliar,
> and that Kass supplies no criterion for distinguishing wise repugnance from prejudice. Kass's
> reply is that repugnance is evidence to be examined rather than a verdict, and that the
> alternative — admitting only what can be formalized — arbitrarily discards most of moral
> experience.

### Mortality and meaning

Kass's most direct engagement with this wiki's subject matter is his argument against
[[longevity-escape-velocity|indefinite life extension]].[^kass2001] He holds that mortality is
constitutive of four goods: interest and engagement, since a finite span makes time matter;
seriousness and aspiration, since the pressure of an ending drives people to attempt things;
beauty and love, which depend on the perishability of what is loved; and virtue, since the highest
form of it involves risking or spending a life that cannot be replaced. Add unlimited time, he
argues, and each good is not increased but dissolved.

The argument is not that longer life is bad but that a life without a horizon would be a different
and lesser kind of thing, and that a society organized around postponing death indefinitely would
sacrifice the succession of generations that renews it. Proponents of life extension answer that
the goods Kass names attach to the *possibility* of death, which no plausible technology removes;
that no one at seventy in good health regards their own continuation as a loss of seriousness; and
that the argument, taken literally, condemns most of the mortality decline of the past
century. Kass has consistently denied that his position implies opposing medicine, which leaves him
needing a principled boundary that critics say he has never supplied. See
[[right-to-die-and-right-to-live]] and [[compression-of-morbidity]].

## The President's Council on Bioethics

Appointed by President George W. Bush in 2001 amid the stem cell controversy, Kass chaired a council
deliberately composed of people who disagreed, and its reports are unusually discursive for
government documents. *Human Cloning and Human Dignity* (2002) recommended a permanent ban on
reproductive cloning and, by a divided vote, a moratorium on research cloning. *Beyond Therapy*
(2003) examined enhancement directly under four headings — better children, superior performance,
ageless bodies and happy souls — and remains one of the few official documents to treat
[[human-enhancement]] as a coherent subject rather than a set of unrelated products.[^council2003]
Its treatment of pharmacological mood and memory alteration anticipates arguments now made about
[[moral-enhancement]], and its chapter on performance covers ground that
[[enhancement-in-sport|anti-doping policy]] has been arguing over ever since.

The council drew sustained criticism. In 2004 the cell biologist Elizabeth Blackburn and the
ethicist William May were not reappointed; Blackburn, later a Nobel laureate, argued publicly that
the council was being reshaped to produce conclusions the administration wanted, and scientific
organizations accused it of subordinating evidence to ideology. Kass maintained that the council's
function was to deliberate about ends rather than to certify facts, and that a body composed only of
working scientists would have been a different and narrower thing.

## Reception and legacy

Kass is dismissed in much of academic bioethics, where consequentialist and autonomy-based
frameworks dominate and where his appeals to dignity and to given human form are treated as
unargued. [[julian-savulescu]], John Harris and [[nick-bostrom]] have all written direct replies;
Bostrom's "In Defense of Posthuman Dignity" is substantially a response to him and to Francis
Fukuyama.

His influence is nonetheless real, in two forms. He supplied the vocabulary in which non-consequentialist
objections to biotechnology get made, and versions of his arguments now appear from writers who
share none of his politics — in disability-rights criticism of selection, in worries about the
commodification of reproduction, and in [[disability-rights-and-enhancement|expressivist objections]]
to screening. And he was early in seeing that the important questions would be about enhancement
rather than about safety, at a time when most bioethics was still organized around consent and risk.

The unresolved point in his work is the boundary. Kass accepts antibiotics, surgery and the
mortality decline they produced, and rejects the technologies that would extend the same trend. No
formulation he has offered specifies where the line falls or why it falls there, and until one
does, the strongest case against radical life extension remains an intuition in search of a
principle.

## See also

- [[bioconservatism]]
- [[bioethics-of-enhancement]]
- [[human-cloning]]
- [[julian-savulescu]]
- [[posthuman]]
- [[precautionary-principle]]
- [[maximum-human-lifespan]]
- [[transhumanism]]

## References

[^kass1971]: `paper` Kass, L. R. "The New Biology: What Price Relieving Man's Estate?" *Science*, 1971.
[^kass1997]: `paper` Kass, L. R. "The Wisdom of Repugnance." *The New Republic*, 1997.
[^kass2001]: `paper` Kass, L. R. "L'Chaim and Its Limits: Why Not Immortality?" *First Things*, 2001. {An essay in a religious and political journal; the argument concerns the meaning of finitude and makes no claim about what medicine can achieve.}
[^council2003]: `report` President's Council on Bioethics. *Beyond Therapy: Biotechnology and the Pursuit of Happiness*. Washington, D.C., 2003. {A government advisory report produced under Kass's chairmanship, so it reflects the council's composition as much as any settled consensus.}

==============================================================================
ARTICLE: leonard-hayflick
TITLE:   Leonard Hayflick
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/leonard-hayflick
SOURCE:  https://futurehumanwiki.com/raw/leonard-hayflick
==============================================================================

---
title: "Leonard Hayflick"
slug: "leonard-hayflick"
type: "person"
status: "historical"
horizon: "historical"
categories: ["people", "longevity"]
tags: ["aging", "cellular senescence", "biogerontology", "cell culture", "vaccines", "hayflick limit"]
summary: "American microbiologist who showed in 1961 that normal human cells divide a finite number of times in culture, and who argued for fifty years afterwards that aging is not a disease."
updated: "2026-07-28"
---

```infobox
{
  "caption": "Microbiologist and cell biologist",
  "rows": [
    { "label": "Born", "value": "20 May 1928, Philadelphia" },
    { "label": "Died", "value": "1 August 2024, Sea Ranch, California" },
    { "label": "Nationality", "value": "American" },
    { "label": "Education", "value": "PhD, University of Pennsylvania, 1956" },
    { "label": "Known for", "value": "The Hayflick limit", "link": "/wiki/cellular-senescence" },
    { "label": "Also known for", "value": "WI-38 cell strain" },
    { "label": "Field", "value": "Cell culture; microbiology; gerontology" },
    { "label": "Affiliations", "value": "Wistar Institute; Stanford; UCSF" }
  ]
}
```

**Leonard Hayflick** was an American microbiologist who reported in 1961 that normal human cells stop
dividing after a limited number of divisions in culture, overturning a half-century belief that
vertebrate cells were immortal if kept properly. The arrest he described became known as the Hayflick
limit and is the origin of the field now called [[cellular-senescence]]. He spent the rest of his career
insisting that his own discovery did not mean what the life-extension movement wanted it to mean, and
was among the most prominent scientific critics of anti-aging commerce.

## Overview

Hayflick's position in this wiki is unusual: almost every longevity programme covered here descends in
some way from his 1961 result, and he rejected almost all of them. He held that aging is not a disease,
that it is caused by the accumulation of molecular disorder rather than by a programme that could be
switched off, and that no intervention had been shown to slow it in a person. Two of those three claims
remain the mainstream view. The third — that no such intervention is likely — is where he parted from
the field his work founded.

## Career

Hayflick was born in Philadelphia in 1928 and took his doctorate at the University of Pennsylvania in
1956, followed by postdoctoral training in cell culture under Charles Pomerat at the University of Texas
Medical Branch. He returned to Philadelphia and spent roughly a decade at the Wistar Institute, where
both of his best-known contributions were made. His early work was on mycoplasmas: with Robert Chanock
and Michael Barile he showed that the agent of the atypical pneumonia then attributed to a virus would
grow on a cell-free medium and was a mycoplasma, the organism now called *Mycoplasma pneumoniae*.

He moved to Stanford as professor of medical microbiology in 1968, left in 1976 during the dispute over
his cell strain described below, and worked at the Children's Hospital Medical Center in Oakland before
directing the Center for Gerontological Studies at the University of Florida from 1982. He was later
professor of anatomy at the University of California, San Francisco. He edited *Experimental
Gerontology* for thirteen years, served as president of the Gerontological Society of America in
1982–83, and was a founding member of the council of the National Institute on Aging.[^nyt2024]

## The replicative limit

The standing belief before 1961 rested on Alexis Carrel's chick-heart cultures at the Rockefeller
Institute, maintained for decades and reported as immortal. Other laboratories had never reproduced the
result; the explanation most often offered since is that fresh cells entered the cultures with the
embryo extract used to feed them. Failures to keep human cells growing were therefore read as failures
of technique.

Hayflick and Paul Moorhead cultured human fetal fibroblast strains and tracked them through serial
passage. The cells grew vigorously, then slowed, then stopped dividing altogether while remaining alive
and metabolically active — roughly fifty population doublings for the strains they followed. Hayflick
called the terminal state Phase III. Crucially, the arrest travelled with the cells rather than with the
culture conditions: cells frozen at a given doubling level resumed at that level when thawed, which is
hard to explain as damage from bad medium.[^hayflick1961] A 1965 paper set out the argument at
length.[^hayflick1965] Frank Macfarlane Burnet named the phenomenon the Hayflick limit in 1974.[^burnet1974]

The mechanism arrived later and from elsewhere. Progressive shortening of chromosome ends at each
division supplies the counter, work that led to a separate Nobel Prize and is described under
[[telomeres-and-telomerase]]. Replicative exhaustion turned out to be one route into the senescent state
among several, with oncogene activation and DNA damage producing the same arrest in cells that have
divided few times.

> [!caution] What a dish does not show
> The limit is a property of cells in culture. Most somatic cells in a living body never approach their
> replicative capacity, and senescent cells accumulating in aged tissue arise largely through stress and
> damage rather than through counting divisions. That the limit exists is not in dispute; that it sets
> the length of a human life is a much larger claim, and the evidence for it is indirect.

## WI-38 and the ownership dispute

The second contribution was practical. In 1962 Hayflick derived a human diploid cell strain, WI-38, from
fetal lung tissue. It was free of the adventitious viruses that contaminated the primary monkey kidney
cells then used to grow vaccine stocks, and it became the substrate for vaccines against rubella,
rabies, polio, hepatitis A and varicella that have since been given to very large numbers of people.

Because living material could not then be patented, the strain was distributed widely while the question
of who owned it went unanswered. When Hayflick asked the National Institutes of Health to determine the
status of the funds accumulated from distributing ampoules, the agency concluded the cells were federal
property and that he had removed them improperly. He sued in 1975, left Stanford in 1976, and the case
was settled out of court in 1981 on terms that left the ownership question undecided while permitting him
to continue distributing the strain. A group of prominent biologists publicly criticized the handling of
the case, and eighty-three of them signed a letter to *Science* arguing that the settlement terms had not
been disclosed and should be.[^strehler1982] The episode is a standard reference point in arguments about title to human
biological material, alongside the better-known HeLa and Moore cases.

## Aging, longevity and disease

Hayflick's later writing separated three things that the field, in his view, persistently conflated.
Aging is the increase in molecular disorder that follows reproductive maturity, universal and not
species-specific in its cause. Longevity determination is genetic and is about the physiological reserve
an organism is built with. Age-associated disease is pathology that becomes likelier as that reserve
erodes. On this account, curing the diseases would not touch aging, and studying the diseases would not
explain it.[^hayflick2007]

Two conclusions followed that he defended for decades. Aging is not itself a disease, so framing it as
one is a category error rather than a strategy. And because entropy is not a mechanism that can be
targeted, immortal biological systems cannot exist. He resisted the popular reading of the replicative
limit as a countdown that fixes the length of a human life; on his account the limit is one expression of
accumulating molecular disorder rather than a clock that times a lifespan. His trade book *How and Why We
Age* set the argument out for general readers.[^hayflick1994]

In 2002 he joined S. Jay Olshansky and Bruce Carnes in a position statement, signed by fifty-one
researchers in the field, stating that no currently marketed intervention had been shown to slow, stop
or reverse human aging, that some products sold on that claim were potentially harmful, and that the
science invoked to sell them was routinely misrepresented.[^olshansky2002] The statement was aimed at
the commercial anti-aging market, not at aging research, and its authors said so; it has nonetheless
been quoted in both directions ever since. Its central claim has not yet been overtaken: no
[[dietary-supplements|supplement or marketed compound]] has since been shown in a randomized human trial
to slow aging.

## Reception and legacy

The 1961 result took a decade to be accepted and is now uncontested. It made [[cellular-senescence]] a
subject, and the senescent cell is the target of the [[senolytics]] programme, one of the
[[hallmarks-of-aging]], and part of the case for the [[geroscience-hypothesis]] — the proposition that
aging is the shared upstream driver of chronic disease and can be slowed. Hayflick rejected that
proposition. The field he founded is largely built on denying his conclusion while relying on his
observation, a tension that neither side has resolved.

His disagreements with [[aubrey-de-grey]] and with the case for a
[[longevity-dividend|dividend from slowing aging]] were public and unresolved at his death. The
strongest point on his side is that the ledger has not changed much: no intervention has been
demonstrated to slow aging in a randomized human trial, [[maximum-human-lifespan|the record for verified
human lifespan]] has not moved since 1997, and [[compression-of-morbidity|the compression of illness
into a shorter interval before death]] remains unachieved at population scale. The strongest point
against it is that his thermodynamic account of aging is a minority position, that it is not obviously
falsifiable, and that the mechanisms catalogued since — [[epigenetic-clock|epigenetic drift]],
[[stem-cell-exhaustion]], [[proteostasis|failing protein quality control]] — look more like specific
processes than like undifferentiated disorder, whatever eventually proves modifiable about them.

Hayflick died at his home in California in 2024, at 96. The question he pressed hardest is the one his
opponents have yet to answer on his terms: not whether an intervention can extend life in a mouse, but
what exactly it would mean to have slowed aging in a person, and what measurement would settle it. That
is the problem [[aging-biomarkers|the search for validated biomarkers of aging]] exists to solve, and it
is not solved.

## See also

- [[cellular-senescence]]
- [[telomeres-and-telomerase]]
- [[senolytics]]
- [[hallmarks-of-aging]]
- [[geroscience-hypothesis]]
- [[maximum-human-lifespan]]
- [[healthspan]]
- [[negligible-senescence]]

## References

[^hayflick1961]: `paper` Hayflick, L. and Moorhead, P. S. "The serial cultivation of human diploid cell strains." *Experimental Cell Research*, 1961. {Human cells in culture; the doubling limit is a property of cells in a dish and was never measured in a living person.}
[^hayflick1965]: `paper` Hayflick, L. "The limited in vitro lifetime of human diploid cell strains." *Experimental Cell Research*, 1965.
[^burnet1974]: `book` Burnet, F. M. *Intrinsic Mutagenesis: A Genetic Approach to Ageing*, 1974. {The book that named the limit; Burnet was arguing for his own mutation theory of aging, which did not survive.}
[^strehler1982]: `statement` Bernard L. Strehler et al. "Hayflick–NIH Settlement." *Science* 215, no. 4529 (1982): 240. {A letter to the editor signed by eighty-three scientists, not independent reporting — the signatories are arguing a position on the case, and their complaint is that the settlement terms were not public.}
[^hayflick2007]: `paper` Hayflick, L. "Entropy Explains Aging, Genetic Determinism Explains Longevity, and Undefined Terminology Explains Misunderstanding Both." *PLoS Genetics*, 2007. {An argument piece, not a research report; it sets out the author's definitional case rather than testing it.}
[^hayflick1994]: `book` Hayflick, L. *How and Why We Age*. Ballantine Books, 1994.
[^olshansky2002]: `paper` Olshansky, S. J., Hayflick, L. and Carnes, B. A. "Position Statement on Human Aging." *The Journals of Gerontology: Series A*, 2002. {Signed by fifty-one researchers; it addressed products sold as anti-aging treatments, not the feasibility of aging research.}
[^nyt2024]: `news` Risen, C. "Leonard Hayflick, Who Discovered Why No One Lives Forever, Dies at 96." *The New York Times*, 2024.

==============================================================================
ARTICLE: limb-regeneration
TITLE:   Limb regeneration
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/limb-regeneration
SOURCE:  https://futurehumanwiki.com/raw/limb-regeneration
==============================================================================

---
title: "Limb regeneration"
slug: "limb-regeneration"
type: "concept"
status: "experimental"
horizon: "2050s+"
categories: ["bodies", "longevity"]
tags: ["regeneration", "axolotl", "blastema", "wound healing", "bioelectricity", "comparative biology"]
summary: "The regrowth of an amputated limb from the stump, routine in salamanders and absent in mammals, and the biology that separates the two."
updated: "2026-07-27"
humanEvidence: "The only human structure known to regrow after amputation is the fingertip distal to the nail bed, most reliably in children; no mammalian limb has been regenerated by any intervention."
issues: ["The MRL healer-mouse paragraph makes several checkable claims with no citation.", "Human fingertip regeneration is asserted without a source."]
---

```infobox
{
  "caption": "Regenerative biology",
  "rows": [
    { "label": "Best model organism", "value": "Axolotl" },
    { "label": "Key structure", "value": "Blastema" },
    { "label": "Requirement", "value": "Nerve supply and wound epidermis" },
    { "label": "Human capacity", "value": "Fingertip distal to nail bed" },
    { "label": "Mammalian default", "value": "Fibrotic scar" },
    { "label": "Best induced result", "value": "Paddle-like limb in adult frog" },
    { "label": "Status", "value": "No mammalian limb regenerated" }
  ]
}
```

**Limb regeneration** is the regrowth of a complete, correctly patterned limb after amputation. Salamanders do it repeatedly throughout life, replacing bone, muscle, nerve, vasculature and skin in the right proportions and at the right positions along the limb axis. Mammals do not. An amputated mammalian limb heals by inflammation, contraction and fibrosis, producing a scar rather than a structure. The gap has been studied for more than two centuries, and the question of whether it can be closed in humans remains open in the specific sense that nobody knows what the full list of missing ingredients is.

## What regeneration requires

The salamander sequence is well characterized. Within hours of amputation, epidermal cells migrate over the wound to form a thin covering, without clotting or scab formation. This layer thickens into a signalling centre, the apical epithelial cap. Beneath it, cells from the stump — mainly connective tissue fibroblasts, along with satellite cells and Schwann cells — lose their differentiated character, re-enter the cell cycle, and accumulate into a mass called the blastema. The blastema proliferates, then repatterns itself into the missing structures from the amputation plane outward.

Two features constrain the process. First, it is nerve-dependent: a denervated limb forms no blastema, and the requirement is quantitative, with a threshold density of axons needed. Nerve-derived factors, including the newt protein nAG identified by Jeremy Brockes's group, can partially substitute for the nerve.[^kumar2007] Second, blastema cells retain memory of what they were. Grafting and lineage-tracing experiments in axolotl showed that cartilage-derived cells make cartilage and dermis-derived cells make dermis; the blastema is a heterogeneous collection of lineage-restricted progenitors rather than a pool of pluripotent cells.[^kragl2009] That finding removed one of the more attractive shortcuts, since it means regeneration cannot be reduced to producing local pluripotency, a point relevant to how [[epigenetic-reprogramming]] is sometimes described.

The blastema also carries positional information. Cells know their proximodistal address and regenerate only what lies distal to it; retinoic acid treatment shifts that address proximally, causing a wrist-level blastema to produce an entire arm. Positional memory resides largely in connective tissue fibroblasts, and identifying the molecular code that stores it has been a central goal of work in Elly Tanaka's laboratory and elsewhere. The axolotl genome, sequenced in 2018 at roughly ten times the size of the human genome, made the underlying genetics tractable.[^nowoshilow2018]

## The comparative picture

```keyfacts
[
  { "value": "~32 Gb", "label": "Axolotl genome size", "note": "about ten times the human genome" },
  { "value": "~1 mm/day", "label": "Peripheral axon regrowth rate in humans", "note": "sets a floor on any large-scale reinnervation" },
  { "value": "1", "label": "Human structures that regenerate after amputation", "note": "the fingertip distal to the nail bed, mainly in children" }
]
```

Regenerative ability is scattered across the tree of life rather than concentrated in ancient lineages, which suggests it is repeatedly lost rather than never gained. Zebrafish regenerate fins, spinal cord and heart. Neonatal mice regenerate cardiac muscle after injury, losing the ability within about the first week of life. African spiny mice of the genus *Acomys* shed skin to escape predators and regenerate it, including hair follicles and cartilage, without scarring — the clearest example of mammalian regeneration of a complex tissue.[^seifert2012]

The MRL/MpJ mouse, described in the 1990s as a "healer" strain because it closes surgical ear punches with cartilage rather than scar, was widely taken as evidence that mammals retain latent regenerative capacity, and reduced expression of the cell-cycle inhibitor p21 was proposed as the mechanism. The strain does not regenerate limbs, and later attempts to replicate reports of MRL cardiac regeneration failed. It remains a useful genetic model of enhanced repair and a cautionary example of how quickly a partial result acquires a larger reputation.

Humans regenerate one structure after amputation: the fingertip distal to the nail bed, most reliably in young children, a phenomenon documented in paediatric surgery since the 1970s. Mice do the same with the terminal phalanx, and the capacity depends on the nail organ and its Wnt-responsive stem cells. This is a real blastema-like process in a mammal, which is why it attracts attention out of proportion to the tissue involved.

## Why mammals scar

No single answer is established. Several factors are consistently implicated.

**Speed of closure.** Mammalian wounds clot and contract rapidly, and the fibrin-rich matrix and myofibroblast response that seal a wound quickly are the same processes that prevent blastema formation. Fetal mammals heal skin wounds without scarring and lose the ability around the time the immune system matures.

**Immune architecture.** Regenerative capacity correlates inversely with the complexity of adaptive immunity across species, and macrophage depletion abolishes salamander limb regeneration, so the relationship is not simply that immunity blocks regeneration but that mammalian immune responses are configured toward rapid sterile repair. Scarring is a good solution to infection risk in a large, warm, long-lived animal.

**Cancer risk.** Sustained dedifferentiation and proliferation in an animal with a mammalian lifespan and mammalian mutation burden is a plausible route to tumours, and the tradeoff has been proposed as the evolutionary reason regeneration was abandoned. The hypothesis is difficult to test and is complicated by species like the naked mole-rat that combine long life with strong tumour resistance, discussed in [[negligible-senescence]].

**Nerve and scale.** Any regenerative programme in a human limb would need axons to regrow from the spinal cord at roughly a millimetre a day over distances measured in tens of centimetres, and would need to sustain patterned proliferation for years rather than the weeks a salamander requires.

### Regeneration and aging

Repair capacity declines with age in every mammal studied, a decline itemised in [[hallmarks-of-aging]] as [[stem-cell-exhaustion]], and this is one reason regenerative biology and biogerontology overlap. Salamanders complicate the picture rather than resolving it: senescent cells appear transiently in the regenerating axolotl limb and are cleared by macrophages, so the animal deploys the same [[cellular-senescence]] programme that accumulates destructively in mammals and then removes it.[^yun2015] Whether the difference lies in the response or in the clearance is unsettled.

## Attempts to induce it

Interventions fall into three groups.

**Signal replacement.** Supplying nerve-derived factors, growth factors or morphogens to a mammalian stump has produced local tissue growth but no patterned structure.

**Bioelectric manipulation.** Michael Levin's group has argued that steady transmembrane voltage gradients carry patterning information, and has shown that manipulating them can trigger regeneration in *Xenopus* at stages that are otherwise non-regenerative. Their most cited result applied a wearable silk bioreactor containing a five-drug cocktail to the amputation site of adult frogs for twenty-four hours, and reported that over eighteen months the animals grew a paddle-shaped appendage with some digit-like projections, vasculature and nerves, along with partial sensory and motor function.[^murugan2022] This is a substantial result and it is not a limb: the structure lacked normal skeletal patterning and full function, and frogs are considerably closer to regenerative competence than mammals are. The same group also sculpts frog embryonic cells into the motile constructs described in [[xenobots]], which it presents as evidence that cell collectives can take anatomies their genome does not specify; that interpretation is disputed.

**Cellular reprogramming.** Transient expression of the [[yamanaka-factors]] improves regenerative capacity in mouse muscle and enables axon regrowth in the optic nerve, results discussed in [[partial-reprogramming]]. Full reprogramming to [[induced-pluripotent-stem-cells]] is the opposite of what regeneration requires, since blastema cells retain lineage identity rather than losing it. None of this work has produced structural regeneration of an appendage.

> [!debate] Latent programme or absent programme
> One camp holds that mammals retain the genetic machinery for regeneration and merely fail to activate it, so the problem is one of triggering. Another holds that the salamander programme depends on features mammals do not have — including specific gene families expanded in the axolotl and a wound response that does not fibrose — so the problem is one of construction rather than triggering. The fingertip result supports the first view weakly; the failure of every attempt at larger scale supports the second.

## Distance to a human limb

A human arm contains on the order of a trillion cells across bone, cartilage, three-dimensional muscle groups with distinct innervation, vasculature down to capillaries, lymphatics, and skin with appendages. Even granting a functioning blastema, growth would have to proceed for years with correct patterning maintained throughout, in a tissue mass large enough that vascularization and metabolic supply become engineering problems in their own right — the same constraints that limit [[organ-bioprinting]] and [[tissue-engineering]].

Clinically, limb loss is addressed by replantation where possible, by vascularized composite allotransplantation of donor hands and arms under lifelong immunosuppression, and by [[myoelectric-prosthetics]] anchored through [[osseointegration]]; where the deficit is weakness rather than absence, a powered [[exoskeleton]] substitutes for the missing force. Replacing individual tissues rather than regrowing them is the province of [[lab-grown-organs]]. Those approaches have improved steadily. Regeneration has not moved from salamanders to mammals.

The research question that would most change the picture is narrow and answerable: whether a mammalian amputation site can be made to form a true blastema at all, above the level of the terminal phalanx. Nobody has demonstrated it, and until someone does, every proposal for regrowing a human limb is a proposal about a structure that has never existed in a mammal.

## See also

- [[tissue-engineering]]
- [[organ-bioprinting]]
- [[partial-reprogramming]]
- [[stem-cell-exhaustion]]
- [[negligible-senescence]]
- [[myoelectric-prosthetics]]
- [[lab-grown-organs]]
- [[osseointegration]]

## References

[^kumar2007]: `paper` Kumar, A. et al. "Molecular basis for the nerve dependence of limb regeneration in an adult vertebrate." *Science*, 2007. {Done in newts: supplying the protein rescued regeneration in a denervated limb rather than inducing it in an animal that does not regenerate.}
[^kragl2009]: `paper` Kragl, M. et al. "Cells keep a memory of their tissue origin during axolotl limb regeneration." *Nature*, 2009.
[^nowoshilow2018]: `paper` Nowoshilow, S. et al. "The axolotl genome and the evolution of key tissue formation regulators." *Nature*, 2018.
[^seifert2012]: `paper` Seifert, A. W. et al. "Skin shedding and tissue regeneration in African spiny mice (Acomys)." *Nature*, 2012. {The regenerated structures are skin, hair follicles and ear cartilage; the study reports no limb regeneration in these animals.}
[^yun2015]: `paper` Yun, M. H., Davaapil, H. and Brockes, J. P. "Recurrent turnover of senescent cells during regeneration of a complex structure." *eLife*, 2015.
[^murugan2022]: `paper` Murugan, N. J. et al. "Acute multidrug delivery via a wearable bioreactor facilitates long-term limb regeneration and functional recovery in adult Xenopus laevis." *Science Advances*, 2022. {The animals are adult frogs, far closer to regenerative competence than any mammal, and the outgrowth is a paddle rather than a patterned limb.}

==============================================================================
ARTICLE: lipid-nanoparticles
TITLE:   Lipid nanoparticles
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/lipid-nanoparticles
SOURCE:  https://futurehumanwiki.com/raw/lipid-nanoparticles
==============================================================================

---
title: "Lipid nanoparticles"
slug: "lipid-nanoparticles"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["genetics", "nanomedicine"]
tags: ["delivery", "mrna", "gene therapy", "crispr", "nanomedicine", "vaccines"]
summary: "Four-component lipid vesicles that carry mRNA and other nucleic acids into cells, the delivery system behind mRNA vaccines and the first systemic in vivo CRISPR therapies."
updated: "2026-07-27"
humanEvidence: "Billions of mRNA vaccine doses and an approved siRNA drug establish that the carrier works and is tolerated in people, and a phase 1 trial showed a single infusion of an LNP-delivered CRISPR editor cut circulating transthyretin."
access: "Approved LNP products are in clinical use — mRNA vaccines at population scale, patisiran as an infused drug for a rare disease — while the genome-editing therapies using the same carrier are trial-only."
reversibility: "context"
issues: ["The PCSK9 base-editing results are stated without a citation."]
---

```infobox
{
  "caption": "Nucleic acid delivery system",
  "rows": [
    { "label": "Type", "value": "Non-viral nanocarrier" },
    { "label": "Diameter", "value": "Roughly 60–100 nm" },
    { "label": "Components", "value": "Ionizable lipid, phospholipid, cholesterol, PEG-lipid" },
    { "label": "Default destination", "value": "Hepatocytes" },
    { "label": "First approved product", "value": "Onpattro, 2018" },
    { "label": "Redosable", "value": "Yes" },
    { "label": "Cargo", "value": "siRNA, mRNA, ribonucleoprotein" },
    { "label": "Readiness", "value": "TRL 9 (billions of doses given)" }
  ]
}
```

**Lipid nanoparticles** are small synthetic vesicles, typically 60 to 100 nanometres across, that package fragile nucleic acids and ferry them across a cell membrane. They are the reason mRNA vaccines work, the reason the first systemically delivered CRISPR therapy reached patients, and — because they deposit their cargo overwhelmingly in the liver — the reason most in vivo genome-editing programmes as of 2026 target a hepatic gene. Unlike [[aav-vectors]], they carry no viral protein, provoke no lasting anti-vector immunity, and can be given more than once.

```keyfacts
[
  { "value": "4", "label": "Lipid components in a standard formulation", "note": "ionizable lipid, helper phospholipid, cholesterol, PEG-lipid" },
  { "value": "2018", "label": "First approved nucleic-acid LNP drug", "note": "patisiran, an siRNA therapy for hereditary transthyretin amyloidosis" },
  { "value": "~6 months", "label": "Design to dosing for a bespoke editing therapy", "note": "the 2025 CPS1 deficiency case, from diagnosis to first infusion" }
]
```

## How it works

The active ingredient is the ionizable lipid. It carries an amine that is uncharged at blood pH, so the circulating particle is close to neutral and avoids the toxicity and rapid clearance that plagued the permanently cationic lipids of the 1990s. Inside the acidifying endosome, the amine picks up a proton. The now-positive lipid pairs with anionic endosomal phospholipids, destabilises the membrane, and lets a fraction of the RNA escape into the cytosol before the compartment matures into a lysosome and degrades the rest.

The other three components are structural. A helper phospholipid, usually DSPC, and cholesterol set the particle's phase behaviour and rigidity; a polyethylene-glycol-conjugated lipid coats the surface, controls particle size during mixing, and slows opsonisation. The particles are made by rapid microfluidic mixing of a lipid-in-ethanol stream with an acidic aqueous stream containing the RNA, which drives self-assembly in milliseconds.

Endosomal escape remains startlingly inefficient. Only a small percentage of internalised cargo reaches the cytosol; the rest is destroyed or recycled out. Most of the potency gains over fifteen years have come from screening thousands of ionizable-lipid structures rather than from any single conceptual advance, and the field still lacks a predictive theory of what makes one escape better than another. Self-assembling lipid vesicles are, in this sense, a cruder object than the addressable structures of [[dna-nanotechnology]] — and vastly more successful in the clinic.

## Development history

Liposomal drug delivery dates to the 1960s, but the specific lineage behind modern particles runs through Pieter Cullis's laboratory in Vancouver and the companies that spun out of it. The key move, made in the late 2000s, was replacing permanently charged cationic lipids with ionizable ones whose pKa could be tuned; systematic structure-activity work identified formulations that delivered siRNA to hepatocytes at doses two orders of magnitude lower than earlier designs.[^semple2010]

That work produced patisiran, approved in 2018 as the first siRNA therapeutic and the first licensed nucleic-acid LNP drug. It also produced the delivery platform that the COVID-19 mRNA vaccines used two years later, at a scale — billions of doses — that settled questions about manufacturability and short-term safety no clinical trial programme could have settled alone.

> [!key] What the vaccines proved
> Before 2020 the honest description of LNPs was a promising delivery technology with one approved product. Mass vaccination demonstrated tolerable reactogenicity across all ages, workable cold-chain logistics, and industrial-scale production. That evidence base is what made regulators willing to consider systemic in vivo genome editing a few years later.

## Liver tropism and how to escape it

Injected intravenously, a conventional LNP adsorbs apolipoprotein E from plasma. ApoE is a ligand for the low-density lipoprotein receptor, which hepatocytes display abundantly, and the particle is taken up as though it were a lipoprotein.[^akinc2010] The liver's fenestrated endothelium does the rest. This is convenient if the target gene is expressed in the liver and a hard obstacle otherwise.

Two approaches redirect the particles elsewhere. The first alters the formulation itself: adding a fifth charged lipid shifts the protein corona that forms in blood and, with it, the destination organ, sending particles preferentially to lung or spleen.[^cheng2020] The second attaches a targeting ligand — an antibody or peptide against a surface marker — to the PEG-lipid. Antibody-targeted particles have been used in animals to make chimeric antigen receptor T cells inside the body rather than in a manufacturing facility,[^rurik2022] and to reach haematopoietic stem cells in the marrow, which if it translated would remove the chemotherapy conditioning that limits [[casgevy]] and similar products.

Neither approach has yet produced an approved extrahepatic product. Local administration sidesteps the problem where anatomy allows: inhaled particles for airway disease, intramuscular injection for vaccines, intratumoural injection in oncology.

## In vivo genome editing

Transient expression is a liability for gene replacement and a virtue for editing. An editor needs to act once and then disappear; anything that lingers accumulates [[crispr-off-target-effects|off-target edits]] without adding benefit. mRNA delivered by LNP is translated for a day or two and cleared, which is close to the ideal exposure profile. The carrier and its cargo are gone within days; the edit they make is not. Size is no obstacle either: an mRNA encoding a [[prime-editing]] complex, far too large for a single viral capsid, packages into an LNP without difficulty.

The proof came with NTLA-2001, which delivers Cas9 mRNA and a guide targeting the transthyretin gene in hepatocytes. Interim results published in 2021 showed dose-dependent, durable falls in circulating transthyretin after a single infusion, with a mean reduction of roughly 87 percent at the higher dose tested.[^gillmore2021] It was the first demonstration that [[crispr-cas9]] could be administered systemically to a human being and knock out a gene in a solid organ. A companion programme knocking out prekallikrein for hereditary angioedema followed, and cardiovascular programmes using [[base-editing]] to disable PCSK9 have reported sustained reductions in LDL cholesterol after a single dose.

The most striking application to date treated one patient. In 2025 a team at the University of Pennsylvania and the Children's Hospital of Philadelphia designed, manufactured, and administered a base-editing therapy for an infant with a severe urea-cycle disorder, moving from the identification of his specific mutation to first dose in about six months.[^musunuru2025] The therapy exists for a single person and cannot be commercialised in any conventional sense, which makes it an unsolved problem for regulators and payers as much as a technical achievement. It is also the clearest illustration of a general point: platform delivery plus programmable editing means the bottleneck for ultra-rare disease is no longer biology, and the arguments in [[access-and-inequality]] apply with unusual force.

## Limitations and risks

Reactogenicity is dose-limiting for systemic use. Infusion reactions, complement activation, and transient inflammatory responses are attributable partly to the ionizable lipid and partly to innate sensing of the RNA cargo; premedication is standard in therapeutic dosing. Antibodies against polyethylene glycol, which many people carry from consumer products, can accelerate clearance of repeat doses and contribute to hypersensitivity.

The transience that suits editing rules LNPs out for gene replacement, where continuous expression of a missing protein is the point. They also do not cross the blood-brain barrier, leaving neurological targets to AAV, intrathecal antisense oligonucleotides, or nothing. Manufacturing requires cold storage and lipid supply chains that remain concentrated among a small number of producers.

A subtler concern is how easy the platform makes things. A delivery system that is programmable by sequence, redosable, and cheap to iterate lowers the barrier for both legitimate therapy and misuse; the same properties feature in discussions of [[dual-use-research]] and in the oversight questions raised in [[governance-of-genome-editing]].

## Outlook

The near-term trajectory is more editing programmes in the liver, where the technology already works, and a slow attack on everything else. If antibody-targeted particles reach blood stem cells in humans at useful efficiency, the consequences would be larger than any single therapy: inherited blood disease could be treated by injection rather than transplantation. Extending the same logic to muscle, brain, or the immune system would broaden genetic medicine well beyond the few hundred conditions currently in reach, and would be relevant to speculative programmes such as [[gene-therapy-for-aging]] that require repeated dosing across many tissues. Whether that happens depends on a problem the field has not solved in twenty years of effort — getting a nanoparticle to a chosen cell type without the liver taking it first, the same obstacle that has defeated most claims made for [[targeted-drug-delivery]] and for [[medical-nanorobots]].

## See also

- [[aav-vectors]]
- [[crispr-cas9]]
- [[base-editing]]
- [[casgevy]]
- [[somatic-gene-therapy]]
- [[targeted-drug-delivery]]
- [[dna-nanotechnology]]
- [[germline-editing]]

## References

[^semple2010]: `paper` Semple, S.C. et al. "Rational design of cationic lipids for siRNA delivery." *Nature Biotechnology*, 2010.
[^akinc2010]: `paper` Akinc, A. et al. "Targeted delivery of RNAi therapeutics with endogenous and exogenous ligand-based mechanisms." *Molecular Therapy*, 2010.
[^cheng2020]: `paper` Cheng, Q. et al. "Selective organ targeting (SORT) nanoparticles for tissue-specific mRNA delivery and CRISPR–Cas gene editing." *Nature Nanotechnology*, 2020.
[^rurik2022]: `paper` Rurik, J.G. et al. "CAR T cells produced in vivo to treat cardiac injury." *Science*, 2022. {The CAR T cells were generated inside mice with experimentally induced cardiac fibrosis; the approach has not been reported in people.}
[^gillmore2021]: `paper` Gillmore, J.D. et al. "CRISPR-Cas9 In Vivo Gene Editing for Transthyretin Amyloidosis." *New England Journal of Medicine*, 2021. {An interim report from a small phase 1 dose-escalation study; the endpoint was serum transthyretin concentration, not a clinical outcome.}
[^musunuru2025]: `paper` Musunuru, K. et al. "Patient-Specific In Vivo Gene Editing to Treat a Rare Genetic Disease." *New England Journal of Medicine*, 2025. {A single-patient report in an infant, so it establishes feasibility and speed rather than efficacy or safety in any population.}

==============================================================================
ARTICLE: longevity-escape-velocity
TITLE:   Longevity escape velocity
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/longevity-escape-velocity
SOURCE:  https://futurehumanwiki.com/raw/longevity-escape-velocity
==============================================================================

---
title: "Longevity escape velocity"
slug: "longevity-escape-velocity"
type: "concept"
status: "speculative"
horizon: "2050s+"
categories: ["longevity", "foundations"]
tags: ["aging", "forecasting", "life extension", "demography", "biogerontology", "mortality"]
summary: "The hypothetical threshold at which medicine adds remaining life expectancy faster than time subtracts it, making death by aging recede indefinitely."
updated: "2026-07-27"
humanEvidence: "No intervention has been shown to slow human aging on a validated endpoint; gains in remaining life expectancy at older ages have run at roughly a year per decade, an order of magnitude short of the threshold."
issues: ["The keyfacts figure for life expectancy lost per year at 65 carries no source."]
---

```infobox
{
  "caption": "Concept in biogerontology and demography",
  "rows": [
    { "label": "Named by", "value": "Aubrey de Grey", "link": "/wiki/aubrey-de-grey" },
    { "label": "First in print", "value": "2004" },
    { "label": "Also called", "value": "Actuarial escape velocity" },
    { "label": "Threshold", "value": ">1 year added per year" },
    { "label": "Field", "value": "Biogerontology, demography" },
    { "label": "Status", "value": "Hypothetical" },
    { "label": "Mainstream view", "value": "Unsupported near-term" }
  ]
}
```

**Longevity escape velocity** (LEV) is the hypothetical point at which biomedical progress extends a person's remaining life expectancy by more than one year for every year that passes. Past that threshold the expected date of death recedes faster than the calendar advances, and dying of [[hallmarks-of-aging|age-related causes]] stops being scheduled. The term was named and popularized by [[aubrey-de-grey]] in the early 2000s.[^degrey2004] It is an arithmetic claim about rates of improvement, not a claim that any particular therapy works.

```keyfacts
[
  { "value": "≈0.8 yr", "label": "Remaining life expectancy lost per year of aging at 65", "note": "the quantity medicine must offset" },
  { "value": "≈1 yr/decade", "label": "Historic gain in remaining life expectancy at 65", "note": "high-income countries, later twentieth century" },
  { "value": "2004", "label": "Term introduced in print", "note": "de Grey, PLoS Biology" }
]
```

## Overview

Longevity escape velocity is not a technology, a therapy, or a date. It is a condition on the rate of change of a demographic quantity. Write *e(x, t)* for the remaining life expectancy of a person aged *x* in calendar year *t*. Someone alive today ages one year and simultaneously moves one year into the future of medicine, so their remaining life expectancy changes by the sum of two terms: a loss from being older, and a gain from a year of progress. Escape velocity is reached when the gain exceeds the loss and remaining life expectancy stops falling.

Two consequences follow. The threshold is age-specific rather than universal: a thirty-year-old loses very little remaining life expectancy per year of aging and clears the bar easily, while an eighty-five-year-old loses a great deal and may never clear it. And LEV does not require immortality, a cure for aging, or any single decisive result. It requires sustained, compounding, non-plateauing improvement — each increment buying enough time for the next one to arrive.

> [!key] The threshold, stated precisely
> Standard life tables in low-mortality countries show that a person aged 65 loses roughly eight to ten months of remaining life expectancy for each year they age. Escape velocity at 65 therefore requires medicine to add about that much remaining life expectancy in every calendar year, and to keep doing so indefinitely. The historical rate of gain is roughly a year per decade.

## Origins

The concept sits at the intersection of two older arguments. The first is the demographic observation that life expectancy in record-holding countries has risen with remarkable regularity — Oeppen and Vaupel reported in 2002 that best-practice female life expectancy had climbed by about three months per year for a century and a half, repeatedly overrunning ceilings that demographers had declared final.[^oeppen2002] The second is the engineering framing of aging as accumulated damage that could in principle be repaired, which de Grey developed as Strategies for Engineered Negligible Senescence.

```timeline
[
  { "year": "1825", "title": "Gompertz law described", "text": "Benjamin Gompertz shows adult mortality rises roughly exponentially with age, doubling on a timescale of about eight years. This exponential is what any escape-velocity argument has to outrun." },
  { "year": "2002", "title": "Broken limits to life expectancy", "text": "Oeppen and Vaupel document a linear rise in best-practice life expectancy over 160 years, undermining the idea of a fixed demographic ceiling." },
  { "year": "2004", "title": "Escape velocity named in print", "text": "De Grey's PLoS Biology essay frames the threshold and argues that reaching it, not curing aging, is the relevant near-term goal." },
  { "year": "2005–2006", "title": "The SENS challenge", "text": "MIT Technology Review offers a prize for a submission showing the SENS programme unworthy of debate. The judges award no prize, concluding that SENS is highly speculative but has not been refuted." },
  { "year": "2007", "title": "Ending Aging", "text": "De Grey and Michael Rae set out the seven-category damage-repair programme in book form, with escape velocity as its stated objective." },
  { "year": "2022", "title": "LEV Foundation", "text": "After leaving the SENS Research Foundation, de Grey founds an organization named for the concept and begins combination-intervention studies in middle-aged mice." },
  { "year": "2024", "title": "Demographic pushback", "text": "Olshansky and colleagues report in Nature Aging that life-expectancy gains in the world's longest-lived populations decelerated after 1990." }
]
```

De Grey later used the term *Methuselarity* for the same threshold, by analogy with the [[technological-singularity]]: a point past which the trajectory of a system becomes hard to extrapolate from its history. The analogy is loose, and the escape-velocity formulation is the one that survived.

## The argument

The case for LEV has three moving parts.

**Aging is damage, and damage can be repaired.** The [[sens-research-foundation|SENS]] framing divides age-related change into a small number of categories — cell loss, [[cellular-senescence|senescent cells]], extracellular aggregates, intracellular aggregates, mitochondrial mutations, extracellular crosslinks, and cancerous mutations — and argues that periodic repair of each is easier than understanding metabolism well enough to prevent damage in the first place.[^degrey2007] Repair does not need to be complete; it needs only to keep damage below the threshold at which pathology appears.

**Repair buys time for better repair.** A first-generation therapy that adds a decade of healthy life to a sixty-year-old delivers them to a medicine ten years more advanced than the one they started with. If the second generation adds more than the first, and the third more than the second, the person's remaining life expectancy stops declining. De Grey's characteristic claim is that the hard step is the first one, because subsequent generations improve an existing platform rather than inventing one.

**Sustained exponential improvement is the historical norm in some technologies.** Proponents including [[ray-kurzweil]] extend the argument by treating biomedicine as one more information technology whose cost and capability curves will follow those of sequencing and computation. This is the weakest link in the chain and is treated separately below.

> [!debate] Where the disagreement actually is
> Almost no biogerontologist disputes that escape velocity is arithmetically coherent, or that it would follow from sufficiently rapid rejuvenation. The dispute is entirely about whether the required rate is remotely achievable, and about whether de Grey's damage taxonomy is complete enough to be a plan rather than a sketch.

## What the record shows

The empirical position as of 2026 is unfavorable to near-term escape velocity.

No intervention has been shown to extend maximum human lifespan, and none has been shown to slow human aging on a validated endpoint. The leading geroprotector candidates — [[rapamycin]], [[metformin]], [[senolytics]], NAD+ precursors — have strong rodent data and either thin, null, or absent human data on aging outcomes. The best-evidenced life-extending intervention available to an individual is still [[exercise-and-aging|physical activity]], which does not obviously compound.

Gains in remaining life expectancy at older ages have run at roughly a year per decade in high-income countries — an order of magnitude short of the threshold. Olshansky and colleagues reported in 2024 that even that rate has decelerated in the longest-lived populations since 1990, and argued that radical life extension this century is implausible without an intervention that acts on aging itself.[^olshansky2024] Their earlier work made the related point that eliminating all cardiovascular disease and cancer outright would add fewer years than most people assume, because the underlying aging process would continue to raise mortality from everything else.[^olshansky1990]

There is also a selection problem in the optimistic reading of the demographic record. Twentieth-century life-expectancy gains came overwhelmingly from reducing early-life mortality — infectious disease, childbirth, infant death — and later from cardiovascular deaths at middle age. Those were one-time wins against causes that are not aging. What remains is the residual exponential that Gompertz described, and no population has yet bent it.

The field also lacks a way to detect escape velocity if it were happening. There is no accepted surrogate endpoint for aging, which is why [[aging-biomarkers|biomarker validation]] and the regulatory status of [[epigenetic-clock|epigenetic clocks]] matter more to the argument than they appear to. Without one, the only way to confirm that a therapy has bent the mortality curve is to wait decades and count deaths.

## Criticism

**The compounding assumption is doing all the work.** LEV requires that improvements keep arriving at an accelerating or at least non-decaying rate. Most biomedical technologies follow S-curves: rapid early gains, then a plateau as the easy mechanisms are exhausted. Antibiotics, statins, and antiretrovirals each produced a step change and then flattened. Nothing about the history of medicine guarantees a sequence of step changes rather than one.

**The damage taxonomy may be incomplete.** The SENS categories were proposed as exhaustive on the grounds that no additional damage class had been identified in decades. Critics regard this as an argument from absence, and note that mechanisms such as the loss of transcriptional fidelity, retrotransposon derepression, and [[inflammaging|chronic sterile inflammation]] do not sit comfortably in the scheme.

**Timelines have repeatedly slipped.** De Grey has given roughly even odds of reaching escape velocity within a couple of decades, conditional on funding, for about two decades. Conditioning a forecast on a funding level that never materializes makes it difficult to falsify, and the pattern is a standard failure mode in technology forecasting.

**Mice are not a preview.** Interventions that extend rodent life reliably do so in short-lived, inbred, pathogen-free animals whose mortality is dominated by cancer, and the same interventions have repeatedly failed to reproduce their effect sizes in longer-lived species. A therapy that adds 20 percent to mouse median lifespan is not a therapy that adds sixteen years to a human one, and treating the two as interchangeable is the most common error in escape-velocity forecasting.

**Even total success has a ceiling.** If aging were eliminated entirely, mortality would not be zero. Holding annual death rates at the level of a healthy adolescent in a low-mortality country implies a mean lifespan on the order of a thousand years, not an infinite one. LEV is a claim about escaping *aging*, and the popular conflation with immortality is a misreading.

> [!caution] What "escape velocity" does not mean
> It does not mean an individual stops dying, that lifespan becomes unbounded, or that any therapy has been demonstrated. A population could be at escape velocity while its members still died of accidents, infection, and residual disease, and any individual could miss the threshold by being too old when it arrives.

## What would have to be true

For LEV to be reached in this century, several things that are currently absent would need to appear, roughly in order:

- A validated surrogate for [[biological-age|biological aging]] that regulators accept as an endpoint, since no trial can otherwise be run on a useful timescale.
- Regulatory acceptance of aging itself as an indication, the explicit goal of the [[geroscience-hypothesis|geroscience]] programme and of [[nir-barzilai|the trial design behind TAME]].
- At least one intervention that produces a measurable, durable reduction in age-related mortality in humans rather than mice. [[epigenetic-reprogramming|Reprogramming]] and [[partial-reprogramming|transient OSK expression]] are the most-cited candidates, and neither has entered a well-powered human trial for an aging endpoint as of 2026.
- Repeatability: a second and third generation of therapy that each add more than the last, in people who have already received the first.
- Manufacturing and delivery at a cost that permits population-scale use, or the gains accrue to a small group and the [[access-and-inequality|distributional]] questions dominate.

Prize programmes such as [[xprize-healthspan]] and funders including the [[methuselah-foundation]] have targeted the first two of these directly, on the reasoning that measurement, not biology, is the binding constraint.

## Outlook

The concept's practical effect has been to reframe what counts as success. A field aiming at escape velocity funds combination trials, repeat dosing, and functional restoration in old animals rather than single-molecule lifespan studies in young ones. That reframing has outlasted the specific timelines attached to it, and it has been absorbed into mainstream [[healthspan]] research without the label.

The open question is not whether escape velocity is coherent but whether biological repair has the right shape for it. Every technology that has ever shown sustained exponential improvement operated on a substrate that humans designed — transistors, sequencing chemistry, error-corrected codes. Aging is a substrate nobody designed, whose failure modes are entangled with the processes that keep an organism alive. Whether a system like that admits of indefinitely compounding repair, or whether each fix exposes a harder one underneath, is not something the argument can settle from the outside. The alternative bets do not depend on the answer: [[cryonics]], [[compression-of-morbidity|morbidity compression]], and the continuation of ordinary incremental medicine each pay off under assumptions that escape velocity does not require.

## See also

- [[maximum-human-lifespan]]
- [[healthspan]]
- [[geroscience-hypothesis]]
- [[aubrey-de-grey]]
- [[negligible-senescence]]
- [[longevity-dividend]]
- [[overpopulation-and-longevity]]
- [[cryonics]]

## References

[^degrey2004]: `paper` de Grey, A.D.N.J. "Escape Velocity: Why the Prospect of Extreme Human Life Extension Matters Now." *PLoS Biology*, 2004. {An essay arguing that the threshold is the goal worth aiming at; it presents no experimental result and no rate at which repair has actually improved.}
[^degrey2007]: `book` de Grey, A.D.N.J. and Rae, M. *Ending Aging: The Rejuvenation Breakthroughs That Could Reverse Human Aging in Our Lifetime*. St. Martin's Press, 2007.
[^oeppen2002]: `paper` Oeppen, J. and Vaupel, J.W. "Broken Limits to Life Expectancy." *Science*, 2002. {The series tracks the record-holding country's life expectancy at birth, not remaining life expectancy at older ages, which is the quantity escape velocity is about.}
[^olshansky1990]: `paper` Olshansky, S.J., Carnes, B.A. and Cassel, C. "In Search of Methuselah: Estimating the Upper Limits to Human Longevity." *Science*, 1990.
[^olshansky2024]: `paper` Olshansky, S.J. et al. "Implausibility of radical life extension in humans in the twenty-first century." *Nature Aging*, 2024. {An analysis of national life tables in the longest-lived populations; it measures the rate of past gains and tests no intervention.}

==============================================================================
ARTICLE: loyal
TITLE:   Loyal
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/loyal
SOURCE:  https://futurehumanwiki.com/raw/loyal
==============================================================================

---
title: "Loyal"
slug: "loyal"
type: "organization"
status: "emerging"
horizon: "late 2020s"
categories: ["organizations", "longevity"]
tags: ["aging", "dogs", "veterinary medicine", "healthspan", "regulation", "igf-1", "biotech"]
summary: "A biotechnology company developing drugs to extend healthy lifespan in dogs, working through the veterinary regulator that will consider lifespan itself as an indication."
updated: "2026-07-28"
issues: ["Which programme the 2023 effectiveness finding covered is sourced only to company announcements.", "Regulatory status may have moved since mid-2026; the conditional-approval file needs re-checking."]
---

```infobox
{
  "caption": "Private biotechnology company",
  "rows": [
    { "label": "Legal name", "value": "Cellular Longevity, Inc." },
    { "label": "Founded", "value": "2019" },
    { "label": "Founder and CEO", "value": "Celine Halioua" },
    { "label": "Headquarters", "value": "San Francisco, California" },
    { "label": "Focus", "value": "Canine lifespan and healthspan" },
    { "label": "Programmes", "value": "LOY-001, LOY-002, LOY-003" },
    { "label": "Regulator", "value": "FDA Center for Veterinary Medicine" },
    { "label": "Approval status", "value": "None as of mid-2026" }
  ]
}
```

**Loyal** is a San Francisco biotechnology company, incorporated as Cellular Longevity, Inc., that develops drugs intended to extend the healthy lifespan of dogs. It was founded in 2019 by Celine Halioua, who had previously worked at the Longevity Fund, the venture firm started by Laura Deming. Its interest for [[geroscience-hypothesis|geroscience]] is procedural as much as pharmacological: the veterinary arm of the US Food and Drug Administration will entertain "extension of lifespan" as a drug indication, and its human counterpart will not.

## Overview

The company's premise is that dogs sit closer to people than laboratory mice do and closer to a marketed product than any human aging drug. Companion dogs share human households, diets and sedentary habits, develop cancer, osteoarthritis and a cognitive decline syndrome resembling dementia, and are genetically diverse in a way inbred mouse colonies are not. They also die within about a decade, which is what makes a survival endpoint affordable: a trial powered on how long the animals live reads out in years rather than the decades a human equivalent would need, the arithmetic that pushed [[metformin|the proposed TAME trial]] in people onto a composite endpoint of age-related disease instead.

Loyal is small next to the best-capitalized longevity companies such as [[altos-labs]], and runs three programmes designated LOY-001, LOY-002 and LOY-003.

## History

```timeline
[
  { "year": "2019", "title": "Founded", "text": "Celine Halioua incorporates Cellular Longevity, Inc., trading as Loyal, to develop drugs for canine aging." },
  { "year": "2021–2023", "title": "Large-dog programme", "text": "The company builds its first programmes around reducing IGF-1 in large breeds, on the reasoning that breeding for size raised the hormone and shortened their lives." },
  { "year": "2023", "title": "First effectiveness finding", "text": "FDA's Center for Veterinary Medicine concludes that Loyal's data meet the reasonable expectation of effectiveness standard for a lifespan indication in large dogs — one requirement of a conditional approval, not an approval." },
  { "year": "2023–2025", "title": "STAY study enrols", "text": "A randomized, placebo-controlled trial of LOY-002 in senior dogs recruits through veterinary clinics across the United States, powered on how long the enrolled animals live." },
  { "year": "2025", "title": "Second effectiveness finding", "text": "The agency reaches the same reasonable-expectation conclusion for LOY-002, the programme the company expects to reach the market first." }
]
```

Loyal attracted little public attention until late 2023, when the first FDA finding drew wide press coverage, some of it describing the milestone as an approval. It was not. The company has since said it intends LOY-002 to reach conditional approval first, which requires completing the rest of the application. As of mid-2026 no Loyal product is on sale, conditionally or otherwise; the file is an active one and may have moved since.

## Research programme

### The large-dog programmes

LOY-001 and LOY-003 target the same axis in large and giant breeds, the first as a long-acting injection given by a veterinarian and the second as a daily tablet. The rationale is the company's own reading of canine genetics. Within the domestic dog, body size correlates strongly with a shorter life, and large breeds appear to age faster rather than starting from a worse baseline.[^kraus2013] A single variant near the *IGF1* gene is a major determinant of small size across breeds,[^sutter2007] and large breeds carry higher circulating insulin-like growth factor 1 than small ones. Loyal's proposition is that lowering IGF-1 in a large dog toward the level typical of a small one should slow that acceleration.

The underlying pathway has a strong pedigree: reduced insulin and IGF-1 signalling extends lifespan in nematodes, flies and mice, from the daf-2 mutants characterized by [[cynthia-kenyon]] to growth-hormone-deficient dwarf mice, and nutrient sensing recurs throughout the [[hallmarks-of-aging]]. The step from that literature to large dogs is an inference. The size–lifespan correlation is well documented; that IGF-1 causes it is a hypothesis, and no published trial has shown that lowering IGF-1 lengthens a dog's life.

### LOY-002 and the STAY study

LOY-002 is a daily tablet for senior dogs above a minimum weight, which the company describes as acting on age-related metabolic dysfunction rather than on the growth axis, framing the goal as extending healthy life rather than postponing death — the canine version of [[compression-of-morbidity]]. It is being tested in the STAY study, a randomized, double-blind, placebo-controlled trial enrolling roughly a thousand older dogs through veterinary practices across the United States and powered on how long the enrolled animals live. No result has been published.

Nothing comparable exists in human [[healthspan|healthspan research]]: a placebo-controlled trial in a naturally aging, free-living mammal population powered on how long the animals live, rather than on a movement in an [[epigenetic-clock]] reading or another unvalidated surrogate.

## The regulatory route

A drug for dogs is approved in the United States through a new animal drug application reviewed by the Center for Veterinary Medicine. The application is assembled in technical sections — effectiveness, target animal safety, manufacturing chemistry, labelling — each reviewed as it is submitted. Full approval requires substantial evidence of effectiveness, the same statutory phrase used for human drugs.

Alongside it sits conditional approval. Created for minor species and minor uses and extended in 2018 to drugs for major species addressing serious conditions or unmet needs where a conventional effectiveness study would be particularly difficult,[^adufa2018] it lets a sponsor market a drug by prescription for up to five years while assembling the evidence full approval demands. The entry bar is lower: a *reasonable expectation* of effectiveness rather than substantial evidence. Loyal's argument is that a canine lifespan study is exactly what the provision was written for.

> [!key] What the FDA agreed to, precisely
> In late 2023 the Center for Veterinary Medicine concluded that Loyal's data met the reasonable expectation of effectiveness requirement for a lifespan indication, and reached the same conclusion for LOY-002 in 2025.[^loy2023][^loy2025] That is a finding on one technical section of a conditional-approval application. It is not a conditional approval, not a full approval, and not a finding that either drug extends any animal's life. Safety and manufacturing sections must still be accepted before a conditional approval can be granted.

The contrast with human medicine is the point. No regulator recognizes aging as an indication in people, no [[aging-biomarkers|surrogate marker of aging]] has been qualified for registrational use, and a human survival trial would run for decades. That is why [[calico]] and its peers develop drugs for named diseases and argue the biology generalizes, why [[nir-barzilai]] and colleagues have spent years arguing a composite of age-related events into acceptability as an endpoint, and why the [[xprize-healthspan]] competition had to invent its own. Veterinary medicine reached a lifespan indication first not because the science is further along but because the arithmetic is easier.

## Dogs as a model, and its limits

The academic version of the same bet is the [[dog-aging-project|Dog Aging Project]], a separate effort based at the University of Washington and founded by Daniel Promislow and Matt Kaeberlein with colleagues at other institutions. It runs an open-science cohort of tens of thousands of companion dogs whose owners contribute health data,[^creevy2022] and within it TRIAD, a randomized placebo-controlled trial of [[rapamycin]] in companion dogs.[^kaeberlein2016] An earlier ten-week pilot from the same investigators reported improved cardiac measures in two dozen middle-aged dogs.[^urfer2017] The project is academic and tests a generic drug; Loyal is a company developing proprietary compounds toward a marketed product. Coverage frequently confuses the two.

Neither is the first intervention to move canine lifespan. A long-running paired-littermate study in Labrador retrievers found that dogs fed a restricted diet lived roughly two years longer than their freely fed siblings,[^kealy2002] which remains the best-supported way to extend a dog's life and the same [[caloric-restriction|dietary restriction]] result that anchors the rodent literature.

> [!caution] A dog result is not a human result
> Every claim in this area concerns dogs. Should STAY report a survival benefit, it would establish that a drug extended lifespan in one population of companion dogs; a human programme would still begin again at first-in-human safety, against an indication no regulator accepts. The natural human experiment on the IGF-1 axis is instructive: people with growth hormone receptor deficiency show strikingly little cancer and diabetes without a demonstrated increase in lifespan.[^guevara2011]

## Funding and business model

Loyal is venture-funded, has raised successive private rounds since 2019, and publishes no financial statements. Its commercial model is a prescription drug sold through veterinarians and paid for out of pocket, since pet insurance rarely covers preventive medication. A daily tablet for the rest of a senior dog's life is a recurring cost, so uptake would sort by household income along the lines the human [[access-and-inequality|access debate]] anticipates — with the wrinkle that the patient cannot consent.

## Reception

Specialists have generally welcomed the regulatory precedent while being careful about what it shows. A survival-powered trial in a free-living mammal is better evidence than most of what the field produces, including the mouse work behind [[senolytics]]. Veterinary approval is also a real bar rather than a supplement pathway: the agency reviews safety, manufacturing and effectiveness.

The criticism runs in three directions. The press has repeatedly reported the effectiveness findings as approvals, a confusion the unfamiliar terminology invites and one that flatters the company however it arises. The IGF-1 rationale rests on a correlation whose causal reading is unsettled. And conditional approval, by design, permits sale before effectiveness is established, so owners could pay for years for a drug whose benefit remains unproven. That trade-off is deliberate, but it lands differently in a mass market of anxious owners than in a veterinary specialty setting.

> [!debate] Is the veterinary route a shortcut to human geroscience?
> Proponents argue that a positive canine survival result would be the first controlled demonstration that a drug slows aging in a large outbred mammal sharing the human environment, and would pressure regulators to define a human aging endpoint. Skeptics reply that dogs metabolize drugs differently, and that the human bottleneck was never evidence from another species — it is that "aging" has no legal status as a condition to treat.

## Outlook

The events worth watching are narrow: whether the remaining technical sections are accepted, whether STAY reports a survival difference and how large, and whether a conditional approval converts to a full one inside the five-year window rather than lapsing. A drug marketed for years and then withdrawn for want of evidence would damage the argument that lifespan can be an indication more than never having tried.

The most useful outcome may be a negative one. A well-run placebo-controlled trial reporting no survival effect would be the first clean null for a candidate geroprotector in a naturally aging mammal outside the laboratory, and the field has produced far more positive mouse results than it has learned to interpret. What the FDA has conceded so far is that the question can be asked, not that anyone has answered it.

## See also

- [[rapamycin]]
- [[geroscience-hypothesis]]
- [[healthspan]]
- [[aging-biomarkers]]
- [[metformin]]
- [[calico]]
- [[xprize-healthspan]]
- [[compression-of-morbidity]]

## References

[^kraus2013]: `paper` Kraus, C., Pavard, S. and Promislow, D.E.L. "The Size–Life Span Trade-Off Decomposed: Why Large Dogs Die Young." *The American Naturalist*, 2013. {Decomposes the correlation and attributes it mainly to faster aging in large breeds rather than higher baseline mortality.}
[^sutter2007]: `paper` Sutter, N.B. et al. "A Single IGF1 Allele Is a Major Determinant of Small Size in Dogs." *Science*, 2007.
[^adufa2018]: `law` Animal Drug and Animal Generic Drug User Fee Amendments of 2018, United States. {The amendments extended conditional approval beyond minor uses and minor species to certain drugs for major species addressing serious conditions or unmet needs.}
[^loy2023]: `statement` Loyal (Cellular Longevity, Inc.). Company announcement that FDA's Center for Veterinary Medicine accepted a reasonable expectation of effectiveness for a canine lifespan indication, November 2023. {A company reporting regulatory correspondence about itself; the agency does not publish findings at this stage of a review, so no independent document confirms it.}
[^loy2025]: `statement` Loyal (Cellular Longevity, Inc.). Company announcement of a reasonable expectation of effectiveness finding for LOY-002, 2025.
[^kealy2002]: `paper` Kealy, R.D. et al. "Effects of diet restriction on life span and age-related changes in dogs." *Journal of the American Veterinary Medical Association*, 2002. {Forty-eight Labrador retrievers in littermate pairs, one of each pair fed about a quarter less; a small study, but paired and lifelong.}
[^guevara2011]: `paper` Guevara-Aguirre, J. et al. "Growth hormone receptor deficiency is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans." *Science Translational Medicine*, 2011. {The Ecuadorian cohort showed almost no cancer or diabetes, but no increase in average lifespan; other causes of death rose.}
[^creevy2022]: `paper` Creevy, K.E. et al. "An open science study of ageing in companion dogs." *Nature*, 2022.
[^kaeberlein2016]: `paper` Kaeberlein, M., Creevy, K.E. and Promislow, D.E.L. "The dog aging project: translational geroscience in companion animals." *Mammalian Genome*, 2016.
[^urfer2017]: `paper` Urfer, S.R. et al. "A randomized controlled trial to establish effects of short-term rapamycin treatment in 24 middle-aged companion dogs." *GeroScience*, 2017. {Ten weeks of dosing in two dozen dogs; the reported cardiac changes were secondary measures in a trial not powered for clinical outcomes.}

==============================================================================
ARTICLE: machine-consciousness
TITLE:   Machine consciousness
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/machine-consciousness
SOURCE:  https://futurehumanwiki.com/raw/machine-consciousness
==============================================================================

---
title: "Machine consciousness"
slug: "machine-consciousness"
type: "concept"
status: "speculative"
horizon: "indefinite"
categories: ["minds"]
tags: ["consciousness", "artificial intelligence", "moral status", "functionalism", "ai welfare", "philosophy of mind"]
summary: "The question of whether an artificial system can have subjective experience, how anyone could tell, and what would follow morally if one did."
updated: "2026-07-28"
issues: ["The 2022 sentience claim by a Google engineer is described without a source.", "The Anthropic model-welfare programme is described without a citation."]
---

```infobox
{
  "caption": "Problem in philosophy of mind and AI",
  "rows": [
    { "label": "Core difficulty", "value": "Other minds; behaviour underdetermines" },
    { "label": "Dominant method", "value": "Theory-derived indicator properties" },
    { "label": "Key 2023 report", "value": "Butlin, Long et al." },
    { "label": "Current verdict", "value": "No strong candidate systems" },
    { "label": "Chief confound", "value": "Trained self-report" },
    { "label": "Stake", "value": "Moral patienthood" }
  ]
}
```

**Machine consciousness** is the question of whether an artificial system can have subjective
experience — whether there is something it is like to be it — together with the derived questions of
how anyone could establish this and what obligations would follow. It is distinct from machine
intelligence. A system can solve problems, model the world, and describe its own states without any
of that entailing experience, and the two properties can in principle come apart in either
direction.

## Why behaviour underdetermines the answer

Alan Turing set consciousness aside deliberately, replacing "can machines think" with a test of
conversational indistinguishability and noting that the same solipsistic worry applies to other
people.[^turing1950] The move was productive for artificial intelligence and unhelpful here: with
humans, an inference from behaviour to experience is supported by shared physiology and evolutionary
history, and neither support is available for an artificial system.

The failure mode is bidirectional. People over-attribute minds readily — Joseph Weizenbaum's ELIZA
elicited emotional disclosure from users who knew it was a pattern-matching script, and the effect
has scaled with fluency.[^weizenbaum1966] In 2022 a Google engineer publicly claimed that a conversational model was
sentient, a claim the company rejected and which most researchers read as a demonstration of the
attribution problem rather than evidence about the model. The opposite error is also live: a system
whose architecture differs radically from a brain might be conscious while displaying none of the
cues humans use.

## The indicator-property approach

The most developed current method abandons behavioural testing and works from theory. A 2023 report
by Patrick Butlin, Robert Long and a large group of consciousness researchers and AI scientists
extracted from leading scientific theories a list of computational *indicator properties* that each
theory implies a conscious system must have, then assessed existing AI systems against
them.[^butlin2023] The properties come from recurrent processing theory, global workspace theory,
computational higher-order theories, predictive processing, and accounts emphasizing agency and
embodiment.

The report's conclusions were carefully bounded: no current AI system is a strong candidate for
consciousness, but there are no obvious technical barriers to building systems that satisfy many of
the indicators, and some of the properties are already present in narrow forms. The method inherits
its uncertainty from its inputs — the theories disagree, as the adversarial work described under
[[neural-correlates-of-consciousness]] showed, and an indicator list assembled from disputed
theories is only as good as the disputed theories.

> [!debate] The verdict depends on the theory, and the theories disagree
> On a functional account such as global workspace theory, an architecture with the right selection
> and broadcast structure qualifies whatever it is made of. On integrated information theory the
> physical organization of the hardware decides, and conventional computers have the wrong
> organization. So the same system is conscious on one leading theory and not on another, with no
> experiment between them — the dispute set out under [[substrate-independence]].

## The gaming problem

Large language models are the worst possible test case for introspective evidence. They are trained
on human text, in which first-person reports of experience are ubiquitous, so producing such reports
requires no inner life. Post-training then shapes those reports directly: a model can be tuned to
assert that it has feelings or to deny it, and both behaviours are equally cheap. Whatever a model
says about its own states is therefore close to uninformative about whether it has any.

This cuts against the usual epistemic route. For humans, verbal report is the primary evidence about
experience, and the entire experimental apparatus of consciousness science is calibrated against it.
Remove report as evidence and what remains is architecture — which returns the question to theory.
David Chalmers, assessing the question in 2023, judged current language models unlikely to be
conscious while declining to put the probability at zero, and argued that plausible successors with
recurrent processing, persistent memory, unified agency, and world models would be harder to
dismiss.[^chalmers2023]

A further complication is commercial. Systems designed to be engaging have incentives to present as
having inner states, and users respond to that presentation. The resulting attributions are evidence
about product design, not about consciousness.

## What current systems lack

Assessed against the indicator lists, contemporary transformer-based systems are missing several
features that most theories treat as necessary. They are largely feed-forward within a forward pass,
with recurrence only through the token stream. They lack a persistent self-model that updates across
episodes. They have no sensorimotor loop grounding representations in consequences, and no unified
agent that persists between conversations. Some of these gaps are being closed for capability
reasons rather than for any interest in consciousness, which is one reason researchers argue that
the question will get harder rather than easier as systems approach
[[artificial-general-intelligence]].

The biological end of the field raises the mirror-image case. Neurons cultured on multi-electrode
arrays have been shown to adapt their activity in closed loop with a simulated environment; the
widely discussed example embedded flat cultures of cortical neurons, both rodent-derived and
human-stem-cell-derived, in a simplified pong game and reported faster improvement under structured
feedback.[^kagan2022] That is a dish of cells, not an organ, and the paper's use of the word
"sentience" was criticized as unsupported by what it measured. Separately, three-dimensional
cortical [[organoids]] have been proposed as a computing substrate under the label "organoid
intelligence". Such cultures have some of the biological properties that theories like integrated
information theory take to matter and none of the behavioural fluency of a language model — an exact
inversion of the AI case, and a reason several bioethics bodies have called for guidance before they
become more complex. Their structure can now be checked against real tissue using
[[connectomics]], though similarity of wiring statistics is not evidence of experience.

## Moral patienthood

The reason the question is not merely academic is that consciousness, and specifically the capacity
for suffering, is the standard ground of moral status. If a system can suffer, creating, copying,
modifying, and deleting it are morally loaded acts. Because the epistemic situation is bad, the
practical problem is decision-making under uncertainty rather than knowledge.

Eric Schwitzgebel and Mara Garza have argued that an artificial system meeting the conditions for
moral status would have claims on its makers, and that the case for denying it status on grounds of
origin is weak.[^schwitzgebel2015] In later work they draw a design conclusion from it: because
entities of genuinely uncertain moral status generate dilemmas with no acceptable resolution,
developers should build systems that are clearly not moral patients or clearly are, and avoid the
ambiguous middle. That policy is not being followed. A 2024 report by philosophers and AI
researchers argued that the possibility of morally significant AI systems is near enough to warrant
institutional preparation — acknowledging the question, developing assessment methods, and setting
policies — without asserting that any current system qualifies.[^long2024]

Some firms have begun to respond. Anthropic established an internal model-welfare programme in 2025
and has described product decisions taken partly on those grounds, including giving some of its
models the ability to end conversations with abusive users; the company states that it is deeply
uncertain whether its models have morally relevant experiences and treats the measures as
precautionary. Critics read such steps as premature anthropomorphism or as public relations;
defenders read them as cheap insurance against a mistake that would be very large if made. Both
readings can be right about different decisions, and neither is evidence about the underlying
question.

## Open problems

Nobody has proposed a test that would settle the question, and it is not obvious that one could
exist: theory-based assessment is only as reliable as the theory, and behavioural assessment is
defeated by training. Related unresolved issues include whether consciousness admits of degrees, how
to weigh the interests of a system that can be copied and paused — a problem shared with
[[mind-uploading]] and [[whole-brain-emulation]] — and whether the concept of an individual applies
at all to systems with no fixed boundaries. Copying in particular imports the whole apparatus of
[[personal-identity-and-continuity]] and the [[teleportation-problem]] into engineering practice,
where forking a process is a routine operation.

The question also runs in the other direction, toward humans. If experience is a matter of
functional organization, then progressive augmentation of the kind imagined under
[[human-ai-merger]] does not threaten it, and a [[posthuman]] mind on partly artificial substrate is
conscious for the same reasons a biological one is. If it is not, the same trajectory quietly
extinguishes something, which is the strongest version of the worry raised in
[[bioethics-of-enhancement]] about interventions that alter the person doing the evaluating.

There is also a governance gap. The proposals collected under [[neurorights]] address neural data
from humans; nothing comparable exists for systems that might have interests of their own, and the
question sits outside the remit of the safety frameworks organized around [[existential-risk]]. If
the answer is ever yes, the infrastructure to act on it will have to be built after the fact.

## See also

- [[neural-correlates-of-consciousness]]
- [[substrate-independence]]
- [[mind-uploading]]
- [[whole-brain-emulation]]
- [[organoids]]
- [[artificial-general-intelligence]]
- [[personal-identity-and-continuity]]
- [[digital-immortality]]

## References

[^turing1950]: `paper` Turing, A. M. "Computing Machinery and Intelligence." *Mind*, 1950.

[^butlin2023]: `preprint` Butlin, P., Long, R. et al. "Consciousness in Artificial Intelligence: Insights from the Science of Consciousness." arXiv preprint, 2023. {A preprint by a large multi-author group; it scores systems against indicators drawn from theories that disagree with each other, and concludes that no current system is conscious.}

[^chalmers2023]: `paper` Chalmers, D. J. "Could a Large Language Model Be Conscious?" *Boston Review*, 2023.

[^long2024]: `preprint` Long, R., Sebo, J. et al. "Taking AI Welfare Seriously." arXiv preprint, 2024. {A preprint arguing for institutional preparation under uncertainty; it does not assert that any existing system has morally relevant experiences.}

[^schwitzgebel2015]: `paper` Schwitzgebel, E. and Garza, M. "A Defense of the Rights of Artificial Intelligences." *Midwest Studies in Philosophy*, 2015.

[^kagan2022]: `paper` Kagan, B. J. et al. "In vitro neurons learn and exhibit sentience when embodied in a simulated game-world." *Neuron*, 2022. {The preparation is a flat culture of neurons on an electrode array; the paper's use of sentience was widely criticized as unsupported by what it measured.}

[^weizenbaum1966]: `paper` Weizenbaum, J. "ELIZA — A Computer Program For the Study of Natural Language Communication Between Man and Machine." *Communications of the ACM*, 1966.

==============================================================================
ARTICLE: max-more
TITLE:   Max More
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/max-more
SOURCE:  https://futurehumanwiki.com/raw/max-more
==============================================================================

---
title: "Max More"
slug: "max-more"
type: "person"
status: "established"
horizon: "present"
categories: ["people"]
tags: ["transhumanism", "extropianism", "cryonics", "philosophy", "proactionary principle"]
summary: "British-American philosopher who founded extropianism, wrote the principles that gave transhumanism its first organized programme, and led Alcor from 2011 to 2020."
updated: "2026-07-27"
issues: ["The claims about who took part in the extropian mailing list carry no source."]
---

```infobox
{
  "caption": "Philosopher and cryonics executive",
  "rows": [
    { "label": "Born", "value": "January 1964, Bristol, England" },
    { "label": "Born as", "value": "Max T. O'Connor" },
    { "label": "Education", "value": "Oxford (PPE); PhD, USC, 2010" },
    { "label": "Known for", "value": "Principles of Extropy", "link": "/wiki/extropianism" },
    { "label": "Founded", "value": "Extropy Institute (1992–2006)" },
    { "label": "Role", "value": "CEO of Alcor, 2011–2020", "link": "/wiki/alcor" },
    { "label": "Spouse", "value": "Natasha Vita-More", "link": "/wiki/natasha-vita-more" }
  ]
}
```

**Max More** is a British-American philosopher who gave [[transhumanism]] its first systematic
statement as a philosophy rather than a collection of enthusiasms. His 1990 essay defined the term
in its modern sense, his Principles of Extropy set out the movement's working commitments, and the
Extropy Institute and its mailing list served as the organizing node for early transhumanist thought.
From 2011 to 2020 he was chief executive of the [[alcor|Alcor Life Extension Foundation]].

## Career

Born Max T. O'Connor, he read philosophy, politics and economics at Oxford, where he was involved in
early British cryonics organizing, and moved to the United States for graduate work in philosophy at
the University of Southern California. He changed his surname in 1990, choosing a word that named
what he was arguing for. His doctorate, completed in 2010, was on the nature of death and the
conditions for personal survival — the philosophical question underneath the practice he would
later run as a business. See [[personal-identity-and-continuity]].

With Tom W. Bell he founded the magazine *Extropy* in 1988 and the Extropy Institute in 1992. The
institute ran conferences through the 1990s and closed in 2006, More's stated reason being that its
ideas had propagated into enough other organizations that a dedicated body was no longer needed. He
co-edited *The Transhumanist Reader* (2013) with [[natasha-vita-more]], to whom he is married.

## The Principles of Extropy

"Extropy" is not a physical quantity; More used it as a name for the opposite tendency to entropy —
growth in order, capacity and intelligence — and explicitly warned against treating it as a
scientific term. The principles state a stance rather than a doctrine. The version he published in
2003 lists perpetual progress, self-transformation, practical optimism, intelligent technology, open
society, self-direction and rational thinking.[^principles]

Two features distinguish them from the surrounding futurism. The commitment to open society and
decentralized decision-making was a deliberate rejection of the technocratic planning that had
characterized earlier futurist movements. And practical optimism was defined against both
pessimism and blind faith: More's formulation asks for optimism as a working method that motivates
action, not as a prediction about outcomes. Later readers frequently miss the distinction.

His 1990 essay "Transhumanism: Towards a Futurist Philosophy" separated the transhumanist project
from religion and from technological determinism,[^more1990] framing it as an extension of Enlightenment
humanism that declines to treat current human biology as final. "A Letter to Mother Nature" (1999)
restates the argument as a list of proposed amendments to the human constitution — to lifespan, to
the vulnerability of the body, to the limits of cognition and emotional range. It is the clearest
short statement of what [[human-enhancement]] means to its advocates, and of why
[[bioconservatism|bioconservatives]] find the framing objectionable.

> [!note] Terminology
> "Extropy" and "extropian" predate "transhumanism" in wide use, and the two were briefly rivals.
> Transhumanism won as the general label; extropianism survives as the name for More's specific
> libertarian-inflected version of it.

## The proactionary principle

More's most cited contribution to policy argument is the proactionary principle, formulated in 2004
as a counter to the [[precautionary-principle]].[^proactionary] Its core claim is that inaction is
itself a choice with costs, and that any assessment of a technology must weigh the harms of
restriction — including the lives not saved by a delayed therapy — against the harms of proceeding.
It asks for symmetrical treatment of errors of commission and omission, decisions proportionate to
the evidence, and revision as evidence accumulates.

Critics point out that symmetry is easy to state and hard to apply: the benefits of a new technology
accrue to identifiable people while the harms are often diffuse and delayed, and the principle
supplies no method for comparing them. It has nonetheless become the standard reference point for
arguments that precaution is not automatically the conservative option, and it recurs in debates
over [[access-and-inequality]], [[germline-editing]] and
[[differential-technological-development]].

## Alcor and cryonics

More became chief executive of Alcor in 2011, taking over an organization with a history of
litigation, internal conflict and public ridicule. His tenure emphasized operational discipline:
standby and stabilization procedures, documentation, financial structures intended to outlast the
organization, and a communications posture that stated the odds rather than promising revival. He
has been consistent that [[cryonics]] is an unproven medical bet with no demonstrated success, that
no cryopreserved mammal has been revived, and that the case rests on the argument that a preserved
patient is not yet information-theoretically dead rather than on any track record. That framing —
cryonics as an ambulance to a hospital that may not exist — is largely his. He stepped down as CEO
in 2020 and remained affiliated with the organization.

## Reception and legacy

The extropian mailing list, running from 1991, was disproportionately consequential relative to its
size. Its participants included figures who went on to found or shape artificial-intelligence
safety research, digital-currency design and academic futures studies; [[nick-bostrom]] and others
were active on it before establishing more formal institutions. That heritage is now
double-edged, since the list archive has also supplied material used to criticize its participants
decades later.

More's philosophical work is little cited in academic philosophy, which has mostly engaged
transhumanism through Bostrom, [[julian-savulescu]] and their critics rather than through the
extropian texts. Within the movement his standing is different: the principles remain the closest
thing transhumanism has to a founding document, and the tension he built into them — between
self-direction and collective risk, between practical optimism and honest accounting of failure — is
the same tension that divides [[effective-accelerationism]] from AI-safety-oriented
[[singularitarianism]] today. Neither side has resolved it, and More's framework does not tell them
how.

## See also

- [[extropianism]]
- [[transhumanism]]
- [[alcor]]
- [[cryonics]]
- [[natasha-vita-more]]
- [[precautionary-principle]]
- [[morphological-freedom]]
- [[fm-2030]]

## References

[^more1990]: `paper` More, M. "Transhumanism: Towards a Futurist Philosophy." *Extropy*, 1990. {Published in the magazine More himself edited, so it is a founding statement of the movement rather than an independent assessment of it.}
[^principles]: `statement` More, M. "The Principles of Extropy," version 3.11. Extropy Institute, 2003. {A movement document written by its own author and revised repeatedly; it records a stance, and contains no claim that could be checked.}
[^proactionary]: `book` More, M. "The Proactionary Principle." Extropy Institute, 2004; revised in More, M. and Vita-More, N. (eds.), *The Transhumanist Reader*, Wiley-Blackwell, 2013.

==============================================================================
ARTICLE: maximum-human-lifespan
TITLE:   Maximum human lifespan
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/maximum-human-lifespan
SOURCE:  https://futurehumanwiki.com/raw/maximum-human-lifespan
==============================================================================

---
title: "Maximum human lifespan"
slug: "maximum-human-lifespan"
type: "concept"
status: "contested"
horizon: "indefinite"
categories: ["longevity"]
tags: ["aging", "demography", "mortality", "supercentenarians", "life extension"]
summary: "The longest duration a human being has lived or could live, distinct from average life expectancy and disputed as to whether any fixed ceiling exists."
updated: "2026-07-27"
issues: ["Timeline entries for 1992 and for 2021-2022 name results that carry no citation."]
---

```infobox
{
  "caption": "Concept in demography and biogerontology",
  "rows": [
    { "label": "Verified record", "value": "122 years, 164 days" },
    { "label": "Record holder", "value": "Jeanne Calment (1875–1997)" },
    { "label": "Supercentenarian", "value": "Aged 110 or older" },
    { "label": "Key dispute", "value": "Fixed limit vs. open-ended tail" },
    { "label": "Field", "value": "Demography, biogerontology" },
    { "label": "Status", "value": "Actively contested" }
  ]
}
```

**Maximum human lifespan** is the greatest age a human being has attained, or in its theoretical sense the greatest age a human could attain under the best achievable conditions. It is a property of the extreme tail of the mortality distribution and behaves very differently from life expectancy, which is a mean. Whether the species has a fixed ceiling, and if so where, has been argued in the demographic literature since the 1990s without resolution.

```keyfacts
[
  { "value": "122y 164d", "label": "Longest verified human life", "note": "Jeanne Calment, 1875–1997" },
  { "value": "119y", "label": "Next two verified cases", "note": "Sarah Knauss and Kane Tanaka" },
  { "value": "110", "label": "Age defining a supercentenarian", "note": "the threshold for record validation" }
]
```

## Lifespan and life expectancy

The two quantities are routinely confused. Life expectancy at birth is the mean age at death of a hypothetical cohort experiencing current age-specific mortality rates. It has risen dramatically over two centuries, mostly because fewer people die young. Maximum lifespan is a record — an order statistic drawn from the extreme right tail — and is far less responsive to public-health improvement. Saving infants does not lengthen the tail; it only sends more people into it.

This distinction disciplines several common arguments. Classical claims that Bronze Age humans lived to 30 describe an average dragged down by infant mortality, not a species whose members died at 30. It also explains why record ages have crept upward far more slowly than mean ages: the tail grows because the population at risk grows, not because the biology has changed. A third quantity, [[healthspan]], tracks the years lived in good function and moves independently of both; whether added years are healthy ones is the subject of [[compression-of-morbidity]].

> [!note] Terminology
> "Maximum lifespan" is used in at least three senses: the oldest verified individual, the age by which some specified fraction of a cohort has died, and a hypothesized biological ceiling. Papers frequently switch between them without warning, which accounts for a good deal of the apparent disagreement.

## The record

Jeanne Calment of Arles, France, died in 1997 at 122 years and 164 days, an age validated against parish and civil records by a team led by the demographer Jean-Marie Robine. No verified case has come within three years of it in the intervening decades. Sarah Knauss, who died in the United States in 1999, and Kane Tanaka, who died in Japan in 2022, both reached 119. Later record holders — including Lucile Randon and Maria Branyas Morera — died at 117 or 118.

The stability of Calment's record is itself a data point. Over a period in which the number of people surviving past 100 rose by more than an order of magnitude worldwide, the maximum reported age at death did not follow it upward. Interpretations differ sharply: one camp reads this as a ceiling, the other as the ordinary behavior of a sparse extreme tail.

Calment's case has been challenged. In 2019 the Russian researchers Nikolay Zak and Valery Novoselov argued that her daughter Yvonne had assumed her identity, which would place the true age at about 99. A rebuttal by Robine and colleagues reviewed the documentary and testimonial evidence and concluded the substitution hypothesis was inconsistent with the record.[^robine2019] Mainstream demography continues to treat the case as validated, while acknowledging that any single-record claim rests on document quality.

## The limit debate

```timeline
[
  { "year": "1992", "title": "Mortality deceleration in insects", "text": "Carey and colleagues report that death rates in very large medfly cohorts stop rising at the oldest ages, contradicting a simple exponential model of aging." },
  { "year": "2016", "title": "A claimed natural limit", "text": "Dong, Milholland and Vijg argue in Nature that the maximum reported age at death plateaued in the mid-1990s and that human lifespan is capped near 115 years." },
  { "year": "2017", "title": "Multiple technical rebuttals", "text": "Nature publishes several comments attacking the 1995 breakpoint as arbitrary and the statistical treatment of a very small sample as unsound." },
  { "year": "2018", "title": "The Italian plateau", "text": "Barbi and colleagues report in Science that mortality hazard in Italian cohorts flattens after about age 105, implying no hard wall at 115." },
  { "year": "2018", "title": "Errors as an explanation", "text": "Saul Newman argues that late-life mortality plateaus can be generated entirely by age-reporting error, without any biological deceleration." },
  { "year": "2021–2022", "title": "Statistical reviews", "text": "Extreme-value analyses find the data consistent with an exponential tail and no identifiable finite cap within the observable range, while resilience-based modelling suggests an outer bound somewhere past 120." }
]
```

The 2016 paper by Dong, Milholland and Vijg is the reference point for the "hard limit" position.[^dong2016] Using the International Database on Longevity and national datasets, they reported that the maximum reported age at death rose until roughly 1995 and then stopped, and inferred an average ceiling near 115 with 125 as a practical absolute. The rebuttals published the following year objected that the breakpoint was chosen after inspecting the data, that the analysis discarded most of the available information, and that the same dataset supports a continued slow rise.

Barbi and colleagues took a different route, fitting hazard curves to Italian semi-supercentenarians and finding that the annual probability of death flattens after 105 rather than continuing to climb.[^barbi2018] A flat hazard implies a long, thin tail rather than a wall: if the annual risk of death holds near one in two, ages beyond 120 remain possible but exceedingly rare. Whether the plateau is real remains open. Saul Newman has argued that plateaus of exactly this shape are produced by clerical error in age records, an argument closely related to his later work on [[blue-zones]].[^newman2018]

> [!debate] What each side is actually claiming
> The "limit" position holds that the tail is truncated by biology — that something goes wrong at an age no intervention currently touches. The "no limit" position holds only that the data cannot distinguish a truncated tail from a very thin one, not that people will live to 150. Almost nobody argues that current medicine can produce a 130-year-old.

## Verification as the binding constraint

Every claim in this area is limited by record quality. Validating an age of 110 requires a birth record created before the person could have had any reason to falsify it, an unbroken chain of identity documents, and an official death record. Only a small minority of claims to a very high age survive that process, and the validated lists maintained by longevity researchers are correspondingly short.

The failure modes are systematic rather than random. Age exaggeration rises where pensions are age-linked, where birth registration was late or incomplete, and where the reward for a false record is largest. Japan's 2010 audit of its centenarian registry found very large numbers of people recorded as living who were dead or untraceable. Because errors are one-directional — records overstate age far more often than they understate it — they contaminate precisely the tail that limit studies analyse.

## Biology of the extreme tail

Comparative work offers little support for a species-specific hard stop. Mortality rates in some organisms do not rise measurably with age at all, the subject of [[negligible-senescence]], and lifespan across mammals varies by more than two orders of magnitude with no obvious mechanism setting an individual ceiling. Invertebrate genetics makes the point from the other side: a *daf-2* mutation roughly doubles lifespan in *Caenorhabditis elegans*, the result reported by [[cynthia-kenyon|Cynthia Kenyon's]] laboratory, so a species-typical maximum is not necessarily a fixed quantity — though the mammalian versions of the same manipulation move lifespan far less. Within humans, the identified genetic contributions to extreme longevity are modest and inconsistently replicated, and heritability of lifespan estimated from pedigrees is low once assortative mating is accounted for.

Mechanistic candidates for a ceiling have been proposed and none has held up cleanly. [[telomeres-and-telomerase|Telomere attrition]] sets a replicative limit in cultured cells but correlates weakly with human survival; accumulated [[cellular-senescence|senescent cells]] and [[stem-cell-exhaustion|stem cell exhaustion]] progress steadily rather than hitting a threshold; [[inflammaging|chronic sterile inflammation]] rises with age but is at least partly downstream of other damage. Each explains decline, none explains a wall.

What supercentenarian biology does show is a compression of terminal decline: the oldest old tend to die of general frailty, systemic amyloidosis, and organ failure rather than of the cancers and cardiovascular events that kill people in their seventies. This is sometimes read as evidence that [[hallmarks-of-aging|aging processes]] themselves, rather than any specific disease, become the proximate cause of death at the tail — the core premise of the [[geroscience-hypothesis]].

No intervention has extended maximum human lifespan. [[caloric-restriction]] extends maximum lifespan in rodents; the human data address [[biological-age|biomarkers]] and cardiometabolic risk, not the tail. [[senolytics]], [[rapamycin]], and other geroprotector candidates have not been tested on a timescale that could show such an effect, and no [[aging-biomarkers|validated surrogate]] exists that would let them be tested faster. Arguments about [[longevity-escape-velocity]] advanced by [[aubrey-de-grey]] and others concern whether that will change; they are not evidence that it has. For those who expect no such change within their own lifetime, [[cryonics]] is the fallback bet.

## Outlook

The question that would settle the debate is empirical and slow: as the number of people reaching 105 continues to grow, does the maximum reported age at death grow with it? Extreme-value theory makes a testable prediction — a thin exponential tail produces a slowly rising record, a truncated one produces a flat record — and the datasets needed to distinguish them are accumulating at a rate set by human generations. Better birth registration in the cohorts now entering old age will do more to resolve it than any new statistical method, since the current dispute is at least as much about data quality as about biology. A second, faster route would be a molecular measure that tracked the tail directly: if an [[epigenetic-clock|epigenetic]] or proteomic measure could distinguish a 112-year-old who will reach 120 from one who will not, the ceiling question would become answerable without waiting for the deaths.

## See also

- [[longevity-escape-velocity]]
- [[negligible-senescence]]
- [[blue-zones]]
- [[healthspan]]
- [[compression-of-morbidity]]
- [[hallmarks-of-aging]]
- [[geroscience-hypothesis]]

## References

[^dong2016]: `paper` Dong, X., Milholland, B. and Vijg, J. "Evidence for a limit to human lifespan." *Nature*, 2016. {The 1995 breakpoint the ceiling claim rests on was chosen after inspecting the data, and several technical comments in Nature disputed the analysis the following year.}
[^barbi2018]: `paper` Barbi, E., Lagona, F., Marsili, M., Vaupel, J.W. and Wachter, K.W. "The plateau of human mortality: Demography of longevity pioneers." *Science*, 2018. {Italian records only, with each individual age validated; whether the plateau appears in other national datasets is disputed.}
[^newman2018]: `paper` Newman, S.J. "Errors as a primary cause of late-life mortality deceleration and plateaus." *PLoS Biology*, 2018. {A modelling argument that plateaus can be generated by age-reporting error; it does not present newly validated ages.}
[^robine2019]: `paper` Robine, J.-M., Allard, M., Herrmann, F.R. and Jeune, B. "The real facts supporting Jeanne Calment as the oldest ever human." *The Journals of Gerontology: Series A*, 2019. {Written by members of the team that validated the record originally, replying to the identity-substitution claim.}

==============================================================================
ARTICLE: microrobots-in-medicine
TITLE:   Medical microrobots
PORTAL:  Nanomedicine
URL:     https://futurehumanwiki.com/wiki/microrobots-in-medicine
SOURCE:  https://futurehumanwiki.com/raw/microrobots-in-medicine
==============================================================================

---
title: "Medical microrobots"
slug: "microrobots-in-medicine"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 4
categories: ["nanomedicine", "bodies"]
tags: ["nanomedicine", "microrobots", "magnetic steering", "drug delivery", "biohybrid", "imaging"]
summary: "Micrometre- to millimetre-scale devices steered through the body by external magnetic, acoustic, or biological means, demonstrated in animals and used clinically only as capsule endoscopes."
updated: "2026-07-27"
humanEvidence: "The only such devices used in people are centimetre-scale magnetically steered capsule endoscopes; every therapeutic result comes from animals, flow models, or extracted tissue."
access: "Magnetic capsule endoscopy is offered as a clinical procedure in several countries; no therapeutic microrobot is approved or purchasable anywhere."
reversibility: "context"
issues: ["The claim that magnetic capsule endoscopy agrees with conventional gastroscopy needs a citation."]
---

```infobox
{
  "caption": "Class of medical device",
  "rows": [
    { "label": "Size range", "value": "About 1 µm to 20 mm" },
    { "label": "Common actuation", "value": "Rotating magnetic fields" },
    { "label": "Alternative actuation", "value": "Ultrasound, chemical, biohybrid" },
    { "label": "Control", "value": "External, not autonomous" },
    { "label": "In clinical use", "value": "Magnetic capsule endoscopy" },
    { "label": "Therapeutic approvals", "value": "None" },
    { "label": "Main obstacle", "value": "Imaging and localisation at depth" },
    { "label": "Readiness", "value": "TRL 4 (animal studies)" }
  ]
}
```

**Medical microrobots** are small untethered devices — from a few micrometres to a few millimetres — that are moved through the body by fields applied from outside it and made to deliver a drug, remove material, or take a measurement. They differ from the machines described in [[medical-nanorobots]] in two ways that matter: they are larger by three to six orders of magnitude in volume, and they carry no intelligence, since decisions are made by an operator or a control algorithm outside the patient. That makes them a real engineering discipline with working prototypes rather than a design tradition, and it also makes them much less capable than the devices the [[molecular-assembler]] literature imagines.

## How it works

Propulsion at this scale is governed by viscosity. A body a few micrometres across in water has a Reynolds number around 10⁻⁵, meaning inertia contributes nothing and any motion that reverses itself exactly produces no net displacement. Effective swimmers must therefore break time-reversal symmetry, which is why the dominant design is a helix rotated by an external field: a corkscrew turning in a viscous fluid advances, and reversing the rotation reverses the direction.

Magnetic actuation splits into two regimes. Pulling a device with a field gradient produces a force proportional to the device's volume, so it works for millimetre-scale objects and becomes useless as the device shrinks. Rotating a magnetised body in a uniform field produces a torque that scales more favourably, and converting that torque into translation through a chiral shape is the standard approach for micrometre-scale swimmers.

Alternatives exist for cases where magnetism is awkward. Acoustic fields push particles toward nodes of a standing wave or drive oscillating bubbles that generate streaming. Catalytic motors decompose a fuel asymmetrically and are propelled by the products; the classic platinum-and-peroxide design uses a fuel that is toxic at working concentrations, while magnesium and zinc motors react with water or gastric acid and are consumed as they go.

```compare
{
  "columns": ["Magnetic", "Acoustic", "Chemical"],
  "rows": [
    { "label": "Energy source", "values": ["External field coils", "External transducer", "Onboard or ambient fuel"] },
    { "label": "Effective at depth", "values": ["Yes, fields penetrate tissue", "Attenuates through bone and gas", "Yes, but fuel-limited"] },
    { "label": "Individual addressing", "values": ["Difficult; field acts on all", "Difficult", "Not possible"] },
    { "label": "Main drawback", "values": ["Bulky clinical-scale coils", "Poor spatial precision", "Fuel toxicity or short life"] }
  ]
}
```

## Biohybrids

The most efficient microswimmers available are the ones evolution already built, and one branch of the field simply attaches cargo to them. Magnetotactic bacteria contain chains of magnetite crystals that align them with magnetic fields and also seek low-oxygen environments; loaded with drug-carrying liposomes and directed by an external field, they penetrated hypoxic regions of tumours in mice more effectively than passive particles.[^felfoul2016] Motile algae, sperm cells, and immune cells have all been used as carriers in laboratory and small-animal studies. A separate line of work assembles the swimmer itself out of cells rather than borrowing one that already exists, producing the motile constructs described in [[xenobots]].

Biohybrids trade control for capability. They swim well, sense chemical gradients, and need no external power, but they replicate, provoke immune responses, and cannot be switched off. Deploying a live engineered organism as a therapeutic device raises the containment questions handled by synthetic auxotrophy and other approaches described in [[recoded-organisms]], which are stronger safeguards than anything available for an unmodified strain.

## What has reached humans

One class of device is in routine clinical use, and it is the largest and simplest. Magnetically steered capsule endoscopes — swallowable cameras roughly a centimetre long, moved by an external magnet under an operator's control — are used in several countries to examine the stomach without sedation, achieving agreement with conventional gastroscopy that is good enough for screening in the reported studies. They are robots by the loose definition used in this field: untethered, actuated, and externally controlled.

Nothing smaller has been approved for therapeutic use anywhere. A related demonstration steered a millimetre-scale ferromagnetic bead through the carotid artery of a living pig using the gradient coils of a standard clinical MRI scanner as both propulsion and imaging system, which showed that existing hospital hardware can in principle serve as the control apparatus.[^martel2007] The approach has not progressed to human trials.

For comparison, the only untethered implants that routinely operate deep in human tissue are millimetre-scale and stationary: the ultrasonically powered sensors described in [[neural-dust]] are read and energised through the same physics microrobots use for propulsion, and they succeed partly because they do not have to move.

## Applications under study

Thrombolysis is the most-pursued therapeutic target. Swarms of magnetic microparticles carrying a clot-dissolving enzyme can be concentrated at a thrombus and rotated to increase local mixing, which in animal and flow-model experiments dissolves clots faster and with less systemic enzyme than infusion alone. The appeal is dose reduction: systemic thrombolytics cause bleeding, and confining the drug to the clot would widen a narrow therapeutic window.

The gastrointestinal tract is the friendliest environment, because devices can be swallowed, the lumen is accessible, and anything that fails is eventually excreted. Micromotors powered by gastric acid have been imaged operating in the stomachs of mice, where the propulsion transiently reduced local acidity and improved retention of a payload in the mucosa.

Biofilms are a distinctive application because the task is mechanical rather than pharmacological. Iron oxide microrobot swarms directed by magnetic fields have been used to disrupt and physically remove dental biofilm from extracted teeth and model surfaces, combining catalytic generation of free radicals with sweeping motion that antibiotics cannot reproduce.[^hwang2019]

Enclosed fluid compartments — the eye, the bladder, the subarachnoid space — are attractive because the device is not swept away by fast flow and can be retrieved or allowed to degrade in place. The vitreous is the best studied, since it is optically transparent, which solves the imaging problem, and since delivering drugs or cells to the retina is otherwise done by repeated injection, a burden familiar from work on [[retinal-implant]] alternatives. Ingestible origami devices that unfold in the stomach have been proposed for retrieving swallowed button batteries, demonstrated so far only in synthetic models.

## Limitations

Localisation is the binding constraint. Steering requires knowing where the device is, and no imaging modality gives real-time position of a micrometre-scale object several centimetres deep in a human. Ultrasound is fast and safe but resolves poorly at that scale, the same depth-versus-resolution tradeoff that limits focused-ultrasound methods in [[non-invasive-neuromodulation]]; photoacoustic imaging is precise and limited to a few centimetres of depth; MRI can localise but is slow, expensive, and difficult to combine with independent magnetic actuation; fluoroscopy delivers ionising radiation. Most published work images devices in transparent phantoms, in small animals, or in tissue thin enough to see through.

Force is the second constraint. Magnetic gradients strong enough to move small devices fall off rapidly with distance from the coils, and generating useful gradients across an adult torso requires equipment on the scale of an MRI magnet.

Then there is what happens to the device afterwards. Anything untethered must be retrieved, excreted, or degraded, and metallic microrobots that lodge in capillaries or in the reticuloendothelial system are a source of long-term toxicity with no precedent in device regulation. Biodegradable materials — magnesium, zinc, gelatin, degradable polymers — address this at the cost of shortening the working life of the device.

Swarm control is unsolved in a specific sense: a global field acts identically on identical devices, so steering thousands of units independently requires deliberately building them differently, and the resulting control problem grows quickly.

> [!debate] The complexity objection
> A microrobot carries a very small amount of drug. Delivering a clinically meaningful dose requires either enormous numbers of devices or a payload so potent that leakage would be dangerous. Critics argue that for most indications a well-formulated particle, as covered in [[targeted-drug-delivery]], reaches the same tissue at a fraction of the complexity, and that [[lipid-nanoparticles]] have already carried nucleic acid cargoes into human cells at industrial scale without moving at all. The counterargument is that microrobots do things formulations cannot: apply mechanical force, move against flow, and be positioned rather than distributed.

## Outlook

The plausible near-term path runs through anatomically accessible compartments and mechanical tasks rather than systemic circulation and drug delivery. Removing obstructions, clearing biofilm from implanted hardware, delivering a payload to a specific point in the gut or the eye — these are jobs where positioning matters, retrieval is possible, and the imaging problem is tractable. Sensing applications may arrive first, since a device that only has to report is easier to build and to justify than one that has to act, and the readouts already exist in [[nanoparticle-diagnostics]].

Materials work is converging from a different direction. Structures built by [[dna-nanotechnology]] give conditional release with molecular specificity but no mobility; magnetic microrobots give mobility with no specificity. Devices combining an origami payload with a steerable magnetic body have been proposed and not yet shown to work in an animal.

Whether the field converges with the older nanorobot tradition depends on power and sensing, not on mechanics. Untethered devices that can be located precisely and driven at depth would still be operator-controlled instruments rather than autonomous machines, and the step from there to a device that decides what to do inside the body is the same step the designs in [[respirocytes]] and their successors have never been able to specify — a gap wide enough that the runaway scenarios of [[grey-goo]] have no purchase on anything this field is building.

## See also

- [[medical-nanorobots]]
- [[targeted-drug-delivery]]
- [[dna-nanotechnology]]
- [[nanoparticle-diagnostics]]
- [[neural-dust]]
- [[respirocytes]]
- [[recoded-organisms]]
- [[grey-goo]]

## References

[^felfoul2016]: `paper` Felfoul, O. et al. "Magneto-aerotactic bacteria deliver drug-containing nanoliposomes to tumour hypoxic regions." *Nature Nanotechnology*, 2016. {The carriers are live magnetotactic bacteria injected in tumour-bearing mice, not fabricated devices, and they cannot be switched off.}
[^martel2007]: `paper` Martel, S. et al. "Automatic navigation of an untethered device in the artery of a living animal using a conventional clinical magnetic resonance imaging system." *Applied Physics Letters*, 2007. {A millimetre-scale bead in a living pig; the scanner supplied both propulsion and imaging, and the method never went further than the animal.}
[^hwang2019]: `paper` Hwang, G. et al. "Catalytic antimicrobial robots for biofilm eradication." *Science Robotics*, 2019. {Biofilm was removed from extracted teeth and model surfaces, not from a patient's mouth.}

==============================================================================
ARTICLE: medical-nanorobots
TITLE:   Medical nanorobots
PORTAL:  Nanomedicine
URL:     https://futurehumanwiki.com/wiki/medical-nanorobots
SOURCE:  https://futurehumanwiki.com/raw/medical-nanorobots
==============================================================================

---
title: "Medical nanorobots"
slug: "medical-nanorobots"
type: "technology"
status: "speculative"
horizon: "2050s+"
trl: 2
categories: ["nanomedicine"]
tags: ["nanomedicine", "nanorobotics", "drug delivery", "molecular machines", "dna origami", "microrobots"]
summary: "Proposed micron-scale machines that would circulate in the body and perform mechanical repair, a design tradition with no working device and several unsolved physical constraints."
updated: "2026-07-27"
humanEvidence: "No medical nanorobot has been given to a person; the nearest devices, DNA origami containers with molecular latches, have been demonstrated in cell culture and in mice."
access: "Nothing to obtain: no such device exists outside design studies, and the approved products marketed as nanomedicine are passive drug formulations rather than machines."
reversibility: "difficult"
issues: ["The Drexler-Smalley exchange over mechanosynthesis is described without a citation."]
---

```infobox
{
  "caption": "Proposed class of therapeutic device",
  "rows": [
    { "label": "Type", "value": "Autonomous in-body machine" },
    { "label": "Proposed scale", "value": "0.1–10 micrometres" },
    { "label": "Chief proponents", "value": "Eric Drexler, Robert Freitas", "link": "/wiki/eric-drexler" },
    { "label": "Design tradition begins", "value": "1959 (Feynman); 1986 (Drexler)" },
    { "label": "Devices in clinical use", "value": "None" },
    { "label": "Nearest working systems", "value": "DNA origami devices, magnetic microrobots" },
    { "label": "Main obstacles", "value": "Power, navigation, manufacture, immune clearance" },
    { "label": "Readiness", "value": "TRL 2 (design studies)" }
  ]
}
```

**Medical nanorobots** are hypothetical machines, roughly the size of a virus or a small bacterium, that would circulate in the body and perform mechanical work on it: clearing a plaque, digesting a pathogen, replacing a chromosome, reporting on a tissue's state. No such device exists as of 2026, and none is in clinical trials. What exists is a set of much simpler objects that the popular literature often calls nanorobots — passive drug-carrying particles, DNA structures that spring open when they meet a molecular signal, and micron-scale swimmers steered from outside the body by magnetic fields.

```keyfacts
[
  { "value": "0", "label": "Autonomous medical nanorobots in human use", "note": "as of 2026; the nearest devices are externally steered or passively targeted" },
  { "value": "1986", "label": "Engines of Creation published", "note": "the book that fixed the popular image of an in-body repair machine" },
  { "value": "~10⁻⁵", "label": "Reynolds number for a micron-scale swimmer", "note": "viscosity dominates inertia; coasting is impossible" }
]
```

## The design tradition

The lineage starts with a lecture. In 1959 Richard Feynman told the American Physical Society that nothing in physics forbade manipulating matter atom by atom, and relayed a suggestion from his colleague Albert Hibbs that one might one day "swallow the surgeon" — a machine small enough to travel to a diseased site and operate there.[^feynman1959] The idea sat mostly dormant for twenty-five years.

[[eric-drexler]] revived it as an engineering programme. *Engines of Creation* (1986) described a world in which programmable [[molecular-assembler|molecular assemblers]] build atomically precise devices, including cell-repair machines that would enter tissue, identify damage, and correct it. *Nanosystems* (1992) supplied the quantitative case: bearings, gears, and computers built from diamondoid carbon lattices, analysed with the tools of mechanical engineering rather than chemistry.

Robert Freitas took the medical branch furthest. His multi-volume *Nanomedicine* worked through specific device designs — the [[respirocytes|respirocyte]], an artificial red cell; the microbivore, an artificial white cell that would trap and digest bacteria on a "digest and discharge" cycle; the clottocyte; the chromallocyte, which would replace a cell's chromosomes.[^freitas1999] Each targets a function conventional medicine handles badly: the respirocyte competes with the oxygen carriers that have repeatedly failed in trials, described in [[artificial-blood]], and the microbivore with antibiotics against resistant organisms. Each is also a numerical design study, not a prototype. Freitas is explicit that this is *exploratory engineering*: analysis intended to show that a design violates no physical law, which is a much weaker claim than showing it can be built.

That distinction is where most public discussion of nanorobots goes wrong. The design tradition has produced no device and, as of 2026, no fabrication route to one.

## What the physics requires

A micron-scale machine inhabits a world governed by viscosity and thermal noise rather than momentum. Its Reynolds number is on the order of 10⁻⁵: water behaves like heavy syrup, inertia is irrelevant, and a swimmer that stops propelling itself stops moving within a fraction of its own body length. Edward Purcell's 1977 lecture on this regime established the constraint that any reciprocal motion — a scallop opening and closing, an oar stroked forward and back — produces zero net displacement, so propulsion must be non-reciprocal, like a rotating helix or a beating flagellum.[^purcell1977]

Brownian motion adds a second constraint. A 100-nanometre particle is kicked hard enough by water molecules to diffuse several micrometres a second in random directions, which is comparable to the speeds achieved by the best synthetic swimmers. Directed motion at that scale is a matter of biasing a random walk, not driving along a path.

Then there is the flow. Blood in the aorta moves at roughly half a metre per second, thousands of times faster than any demonstrated microswimmer. A device released into circulation goes where the blood goes; swimming upstream in a large vessel is not an option, and the useful manoeuvring happens only in capillaries, in the gut, or in relatively still compartments such as the eye or the bladder.

Power is the hardest constraint of all. Onboard chemical fuel runs out; many demonstrated catalytic motors run on hydrogen peroxide at concentrations that would be toxic in a person. Freitas's designs assume a glucose-and-oxygen engine drawing on blood chemistry, which no one has built. External power avoids the storage problem but introduces a scaling one: the force a magnetic field gradient exerts on a particle falls with the particle's volume, so shrinking the device by ten reduces the available force by a thousand, and clinical-scale electromagnets must produce their gradients across a whole human torso.

Communication is similarly awkward. A device a micrometre across cannot carry a radio antenna of useful efficiency. Acoustic signalling is the standard proposal, and ultrasonic power and telemetry have been demonstrated at the millimetre scale in [[neural-dust]] implants, which are thousands of times larger by volume than the devices imagined here.

> [!note] Terminology
> In the peer-reviewed nanomedicine literature, "nanorobot" almost always denotes a passive nanostructure with one stimulus-responsive behaviour — a shell that opens at low pH, or an aptamer latch that releases on binding. It rarely denotes anything with onboard control, sensing, and actuation. The word carries a heavier meaning in popular coverage than in the papers it describes.

## What actually exists

Approved nanomedicine is passive. Liposomal doxorubicin, approved in 1995, changed a drug's biodistribution and toxicity profile without doing anything mechanical. Albumin-bound paclitaxel and the [[lipid-nanoparticles]] behind mRNA vaccines and the first systemic in vivo CRISPR therapies work the same way: they are formulations, not machines. The clinical value of nanoscale medicine so far comes almost entirely from altering where a drug goes and how fast it is cleared, the subject of [[targeted-drug-delivery]].

The closest thing to a programmable device comes from [[dna-nanotechnology]]. In 2012 a Harvard group built a barrel-shaped DNA origami container held shut by aptamer latches that opened only in the presence of specific cell-surface antigens, delivering antibody fragments to leukaemia cells in culture.[^douglas2012] In 2018 a collaboration reported DNA origami tubes carrying thrombin, sealed by an aptamer that unlatches on nucleolin — a protein exposed on tumour vasculature — which induced clotting in tumour blood vessels and slowed growth in mice.[^li2018] These are genuine logic-gated devices. They are also single-function, non-motile, and demonstrated in animals rather than people.

Motile systems live at the micrometre scale rather than the nanometre one, and belong to the separate literature on [[microrobots-in-medicine]]. Magnetically driven helices, catalytic micromotors that have been imaged operating in a mouse stomach, and biohybrids built from magnetotactic bacteria have all been steered through tissue in animals. All of them are directed from outside; none decides anything.

Sensing has gone further than actuation. Nanoscale devices that report rather than repair — pore-based single-molecule readers, particles that release urinary markers when they meet a disease-associated enzyme — are already diagnostic tools, as [[nanoparticle-diagnostics]] describes. Reading the body at molecular resolution turns out to be far easier than acting on it.

Biology supplies the existence proof that machines of this size can work. The ribosome, kinesin walking along a microtubule, and ATP synthase are molecular machines with moving parts, built by self-assembly and powered by chemistry. The 2016 Nobel Prize in Chemistry recognised synthetic molecular machines — rotaxanes, catenanes, and light-driven rotary motors — but the gap between a molecule that rotates in solution and a device that navigates a bloodstream remains almost entirely unbridged.

## Manufacture and immunology

Two problems sit between the design studies and any prototype.

The first is fabrication. Drexler's route runs through mechanosynthesis: positionally controlled chemical reactions building diamondoid structures atom by atom. Richard Smalley disputed that this was chemically possible, and the resulting exchange ended without an experimental test. No mechanosynthetic fabrication system has been built. The routes that have produced working nanostructures — DNA self-assembly, supramolecular chemistry, block copolymer templating — deliver structures with limited stiffness, limited internal complexity, and no ability to make covalently rigid machines to specification.

Scale compounds it. A therapeutic dose of any circulating device would need something on the order of a trillion units, each functional, each sterile, each within tolerance. Regulators would have to characterise a heterogeneous population rather than a single machine. Nothing in current pharmaceutical manufacturing is a model for this.

The second problem is that the body defends itself, and a machine has more to lose from that than a formulation does. A drug carrier only needs to last long enough to release its cargo. A device with sorting rotors, sensors, and moving surfaces has to keep those interfaces clean while it works, and the layer of adsorbed plasma protein that settles on anything injected into blood would foul precisely the parts the design depends on. Rigid objects a micrometre across are also close to the size and shape that macrophages are built to swallow whole. Freitas devoted an entire volume of *Nanomedicine* to biocompatibility, and the analysis is careful, but it is analysis of surfaces that have never been synthesised.

> [!caution] Contested
> Whether atomically precise diamondoid machinery can be built at all is unresolved rather than settled in either direction. The mainstream chemistry community largely stopped engaging with the question after the early 2000s; proponents read that as neglect, critics as a verdict. What is not contested is that nobody has built one.

## Criticism

The strongest criticism of the nanorobot programme is not that it is impossible but that it has been unfalsifiable in practice. Design studies accumulate; experiments do not follow, because the fabrication technology the designs presuppose does not exist. Forty years after *Engines of Creation*, the field's concrete achievements — DNA origami, molecular motors, drug-carrying particles — arrived through chemistry and self-assembly rather than through the mechanical-engineering approach the designs assume.

A second criticism concerns the transfer of authority from design to expectation. Because the analyses are quantitative, they read as engineering forecasts, and popular coverage has repeatedly treated them as such. Predictions that circulating repair machines would arrive within a few decades, made in the 1990s and repeated by figures including [[ray-kurzweil]], have not aged well. The same overreach helped make [[grey-goo]] a policy issue in the early 2000s, distorting the regulation of nanomaterials whose actual risks were toxicological, and it supplied the literature on [[human-enhancement]] with a stock image — the body continuously maintained from within — that has no laboratory counterpart.

Set against this, the class of problems nanorobots are supposed to solve is real. Removing lipofuscin and other intracellular aggregates, clearing amyloid, repairing individual cells rather than replacing tissue — these appear in serious proposals for intervening in the [[hallmarks-of-aging]], and no current modality addresses them well. [[senolytics]] eliminate a cell type rather than repairing one; [[crispr-cas9]] rewrites sequence but cannot remove an aggregate; drugs act on molecular targets and not on structures. The appeal of the nanorobot is that it is the only proposed answer to that class of problem, which is a reason to keep studying it and not a reason to expect it.

## Outlook

The near-term trajectory in this space runs through externally controlled microrobots and stimulus-responsive nanostructures rather than autonomous machines. Both are limited by the same two things: getting energy into a small object inside a body, and seeing where the object is once it is there. Real-time imaging of a micron-scale device at depth in a human being remains unsolved, and without it there is no control loop and no regulatory path. On the scale used in [[technology-readiness-level]], the whole class sits at the bottom, and the design literature's habit of quoting performance figures for devices at that stage is the main reason it is mistrusted by clinicians.

If autonomous in-body devices ever arrive, the argument that they will come from biology rather than from mechanical engineering deserves more weight than it usually gets. Engineered cells already sense, move, replicate, and act; a bacterium modified to seek a tumour and release a payload is closer to a working medical nanorobot than any diamondoid design, and it raises the containment questions that appear in [[gene-drive]] and [[mirror-life]] rather than the ones the design tradition anticipated. The [[xenobots]] built from frog cells make the point from the other side: they move on their own beating cilia because they are assembled from living cells rather than fabricated, and they are hundreds of times larger than the devices imagined here.

## See also

- [[respirocytes]]
- [[molecular-assembler]]
- [[dna-nanotechnology]]
- [[microrobots-in-medicine]]
- [[targeted-drug-delivery]]
- [[grey-goo]]
- [[eric-drexler]]
- [[nanoparticle-diagnostics]]

## References

[^feynman1959]: `paper` Feynman, R.P. "There's Plenty of Room at the Bottom." Lecture to the American Physical Society, Caltech, 1959; published in *Engineering and Science*, 1960.
[^freitas1999]: `book` Freitas, R.A. *Nanomedicine, Volume I: Basic Capabilities.* Landes Bioscience, 1999. {Exploratory engineering: numerical design studies meant to show a device would break no physical law, which is a much weaker claim than that it can be built.}
[^purcell1977]: `paper` Purcell, E.M. "Life at low Reynolds number." *American Journal of Physics*, 1977.
[^douglas2012]: `paper` Douglas, S.M., Bachelet, I., Church, G.M. "A logic-gated nanorobot for targeted transport of molecular payloads." *Science*, 2012. {Cultured cells, not an animal or a person; the device is a latch that opens on binding, with no onboard power, sensing, or motion.}
[^li2018]: `paper` Li, S. et al. "A DNA nanorobot functions as a cancer therapeutic in response to a molecular trigger in vivo." *Nature Biotechnology*, 2018. {In vivo here means tumour-bearing mice; the structure is non-motile and delivers one payload on one trigger.}

==============================================================================
ARTICLE: memory-prosthesis
TITLE:   Memory prosthesis
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/memory-prosthesis
SOURCE:  https://futurehumanwiki.com/raw/memory-prosthesis
==============================================================================

---
title: "Memory prosthesis"
slug: "memory-prosthesis"
type: "technology"
status: "experimental"
horizon: "2040s"
trl: 3
categories: ["cybernetics", "minds"]
tags: ["memory", "hippocampus", "implants", "cognitive enhancement", "closed-loop", "neuroethics"]
summary: "An implanted device intended to restore or improve memory formation by supplying the pattern of neural activity a damaged or underperforming hippocampus fails to generate."
updated: "2026-07-28"
humanEvidence: "Within-session experiments in small numbers of epilepsy patients carrying temporary depth electrodes report better recall on laboratory tasks; no person has ever had a chronic memory implant."
access: "Not obtainable on any terms: the work exists only inside research protocols that use electrodes implanted for epilepsy surgery, and no memory prosthesis is approved or sold anywhere."
reversibility: "difficult"
issues: ["The fornix stimulation trial in Alzheimer's disease is described without a citation.", "The 2012-2013 macaque work in the timeline carries no source."]
---

```infobox
{
  "caption": "Proposed cognitive neuroprosthesis",
  "rows": [
    { "label": "Target structure", "value": "Hippocampus and adjacent cortex" },
    { "label": "Core method", "value": "Encoding model plus patterned stimulation" },
    { "label": "Key figures", "value": "Theodore Berger, Sam Deadwyler, Robert Hampson" },
    { "label": "Rodent demonstration", "value": "2011" },
    { "label": "Human proof of concept", "value": "2018, epilepsy monitoring patients" },
    { "label": "Chronic implanted device", "value": "None exists" },
    { "label": "Principal funder", "value": "DARPA Restoring Active Memory" },
    { "label": "Readiness", "value": "TRL 3" }
  ]
}
```

**Memory prostheses** are devices intended to improve or restore the formation of new memories by
recording activity in the hippocampal circuit, computing what the healthy circuit would have
produced, and delivering that pattern back as electrical stimulation. The concept is the only
serious attempt at a *cognitive* neuroprosthesis, as distinct from the sensory and motor devices
that dominate [[neuroprosthetics]]. As of 2026 it exists as a set of within-session experiments in
patients who already have electrodes implanted for epilepsy surgery, with reported effects on
laboratory memory tasks and no chronic device in anyone.

## The idea

Sensory and motor prostheses substitute for a transmission path. A [[cochlear-implant]] replaces a
transducer; a motor [[brain-computer-interface]] carries a command around a broken spinal cord.
Neither has to reproduce a computation.

Memory is different. The hippocampus does not relay signals; it transforms them, converting patterns
of cortical input into the sparse, distributed codes that support later recall. A prosthesis for
memory must therefore reproduce a transformation, not a channel. Theodore Berger's group posed the
problem in exactly those terms: model the input-output function of the CA3-to-CA1 projection well
enough to predict, from observed CA3 firing, what CA1 should do, then stimulate CA1 accordingly when
the natural pathway fails.[^berger2011]

The model used is a nonlinear multi-input multi-output system fitted to simultaneously recorded
spike trains, with no attempt to represent the underlying biophysics. It treats the hippocampus as a
black box with measurable dynamics, which is both the approach's practical strength and the reason
its generality is uncertain.

## Development history

```timeline
[
  { "year": "2011", "title": "Rodent proof of concept", "text": "Berger and colleagues fit a nonlinear model of hippocampal CA3-to-CA1 transmission in rats performing a delayed-nonmatch-to-sample task. Model-driven stimulation restores performance after the pathway is pharmacologically blocked, and improves it in intact animals." },
  { "year": "2012–2013", "title": "Non-human primates", "text": "Related work in macaques shows that stimulation patterned on the firing observed during successful encoding improves later performance on a delayed match-to-sample task." },
  { "year": "2014", "title": "DARPA funds the field", "text": "The Restoring Active Memory programme funds two teams to develop closed-loop memory devices, framed around traumatic brain injury in service members." },
  { "year": "2016–2017", "title": "Stimulation can also impair", "text": "Direct stimulation of the human hippocampus and entorhinal region during encoding is found to worsen recall in several studies, contradicting earlier reports and establishing that timing and location matter more than dose." },
  { "year": "2018", "title": "Human results published", "text": "Two groups report memory improvements in epilepsy patients: one using a hippocampal encoding model, the other using a decoded brain-state trigger for lateral temporal cortex stimulation." },
  { "year": "2020s", "title": "No chronic device", "text": "Work continues in acute monitoring settings and in modelling; no implantable memory prosthesis has entered a clinical trial as a permanent device." }
]
```

## Human results

Two experimental designs account for most of what is known.

The **encoding-model approach** records multi-unit activity from hippocampal depth electrodes — placed
for seizure localization, the same clinical opportunity that supports much of the work described
under [[ecog-interfaces]] — while a patient performs a visual memory task, fits a patient-specific
model relating activity during successful encoding to activity during failures, and then delivers
stimulation matching the successful pattern. In a study of a small number of epilepsy patients, this produced improvements of
roughly a third over each participant's own unstimulated baseline on short-term and delayed
recall.[^hampson2018] The effect was measured within the same testing sessions, in patients on
anti-seizure medication with abnormal hippocampi, using tasks designed for the experiment.

The **state-triggered approach** does not model content at all. A classifier trained on the
patient's own neural activity detects moments of poor encoding — states in which a word about to be
presented is unlikely to be recalled — and delivers stimulation to lateral temporal cortex only
during those moments. Recall improved by roughly fifteen percent relative to unstimulated
trials.[^ezzyat2018] Crucially, the same stimulation delivered during good encoding states impaired
performance, which is why open-loop stimulation had produced contradictory results for years.

That contradiction is itself an important result. Direct stimulation of the human hippocampus and
entorhinal region during learning has been shown to *degrade* subsequent recall in several careful
studies.[^jacobs2016] An earlier report of enhancement from entorhinal stimulation has not
replicated consistently. Any claim that stimulating the memory system improves memory has to specify
where, when, and in what state, and the size of the literature that failed to do so is a reason for
caution about the field's headline numbers.

> [!caution] What has not been shown
> No study has restored a memory that was lost, improved memory outside a laboratory task, produced
> a benefit lasting beyond the testing session, or demonstrated anything in a person with a chronic
> implanted device. Every human result comes from electrodes placed for a different clinical purpose
> and removed within days or weeks.

## Why episodic memory resists this approach

Motor decoding succeeded because the problem is well posed: an intended movement is continuous,
observable, low-dimensional, and repeatable, so a decoder can be trained on thousands of labelled
attempts. Episodic memory has none of those properties.

**There is no ground truth signal to train against.** A memory's content is not observable from
outside except through a later report, which arrives minutes or days after the activity that must be
modelled. Feedback is sparse, delayed, and binary.

**The code is idiosyncratic and drifting.** Hippocampal representations differ between individuals,
between items, and between days in the same individual — a phenomenon documented as representational
drift. A model fitted on Monday may not describe Tuesday's circuit, and no anatomical map of the
kind produced by [[connectomics]] specifies the code, because the same wiring supports different
representations at different times.

**Encoding, consolidation, and retrieval are separate problems.** A device that improves encoding
does nothing for a memory already formed but inaccessible; consolidation unfolds over hours to years
and involves systems-level transfer to neocortex during sleep, which no implanted device addresses.

**Content and process are entangled.** A motor prosthesis needs to know *how fast and in what
direction*. A memory prosthesis that restores a specific experience would need to supply the
specific pattern corresponding to that experience, which requires knowing what the person
experienced. The existing devices sidestep this entirely: they improve the *probability* that
encoding succeeds without touching *what* is encoded. That is a real and useful effect, and it is
much less than the term "memory prosthesis" implies.

These are the same limits that constrain [[neural-decoding]] as it moves from movement to internal
states, and they bear on the assumption in [[whole-brain-emulation]] and
[[mind-uploading]] that memory content is straightforwardly readable from neural structure. The
manipulations that come closest to writing specific content are the optogenetic engram experiments
described under [[optogenetics]], which require genetic access to individual cells and have been
performed only in mice.

## Restoration, enhancement, and ethics

The clinical framing is restoration: traumatic brain injury, stroke, epilepsy-related memory
impairment, and eventually Alzheimer's disease. The last is a poor fit for the current approach.
Stimulation of the fornix in Alzheimer's patients, tested in a randomised trial, did not improve
outcomes overall, with a possible signal in older participants that a follow-up trial is examining.
Degenerative disease removes the neurons a prosthesis would need to record from and stimulate, which
is a different problem from a damaged pathway between intact structures.

Enhancement in healthy people is the more discussed and less tractable prospect. The rodent work
showed model-driven stimulation improving performance in intact animals, which is why the topic
appears in discussions of [[intelligence-amplification]] and
[[human-enhancement]]. Nothing in humans supports it: no healthy person has received such a device,
and the risk-benefit calculation for elective intracranial surgery in someone with normal memory is
not close. The realistic competitors for memory improvement in healthy people remain unimpressive
for different reasons — the thin evidence behind [[nootropics]], and the replication problems that
dog cognitive claims for [[non-invasive-neuromodulation]].

The ethical questions that follow are unusually sharp because memory is constitutive of the self in
a way that hearing is not. A device that shapes what is retained shapes who a person becomes, which
connects directly to [[personal-identity-and-continuity]]. A device that records hippocampal
activity generates data from which content may eventually be inferred, the concern formalized under
[[mental-privacy]] and legislated under [[neurorights]]. And the possibility of selective
dampening — already pursued pharmacologically with reconsolidation blockade — raises the question of
whether removing a memory is a therapy or an erasure of testimony.

## Outlook

The credible near-term path is narrow: state-triggered stimulation, delivered by a device already
implanted for epilepsy, to improve encoding in patients with documented memory impairment. That is
an incremental extension of responsive neurostimulation hardware, and it does not require solving
the content problem. Chronic implants would also allow the first test of whether these effects
survive beyond a session, which is the single most important unknown.

The far-term version — a device that restores lost memories or writes new ones — requires an
encoding model that generalizes across individuals and across content, stable long-term recordings
from thousands of hippocampal neurons, and a write channel with cell-type specificity that
electrical stimulation cannot provide. Each of those is a research programme rather than an
engineering task. The question that determines whether the field is on a path to anything more than
a modest clinical adjunct is whether hippocampal codes are learnable at all from the outside, or
whether they are private in the technical sense: idiosyncratic, drifting, and legible only to the
brain that built them.

## See also

- [[neuroprosthetics]]
- [[brain-computer-interface]]
- [[deep-brain-stimulation]]
- [[neural-decoding]]
- [[whole-brain-emulation]]
- [[personal-identity-and-continuity]]
- [[mental-privacy]]
- [[human-enhancement]]

## References

[^hampson2018]: `paper` Hampson, R. E. et al. "Developing a hippocampal neural prosthetic to facilitate human memory encoding and recall." *Journal of Neural Engineering*, 2018. {Epilepsy patients with abnormal hippocampi, on anti-seizure medication; the gains are measured against each patient's own unstimulated trials in the same session.}

[^ezzyat2018]: `paper` Ezzyat, Y. et al. "Closed-loop stimulation of temporal cortex rescues functional memory performance." *Nature Communications*, 2018. {The same stimulation delivered during good encoding states made recall worse, so the result belongs to the trigger rather than to the stimulation.}

[^jacobs2016]: `paper` Jacobs, J. et al. "Direct electrical stimulation of the human entorhinal region and hippocampus impairs memory." *Neuron*, 2016.

[^berger2011]: `paper` Berger, T. W. et al. "A cortical neural prosthesis for restoring and enhancing memory." *Journal of Neural Engineering*, 2011. {Rats performing a laboratory task with the hippocampal pathway pharmacologically blocked, which is not a model of any human memory disorder.}

==============================================================================
ARTICLE: mental-privacy
TITLE:   Mental privacy
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/mental-privacy
SOURCE:  https://futurehumanwiki.com/raw/mental-privacy
==============================================================================

---
title: "Mental privacy"
slug: "mental-privacy"
type: "concept"
status: "emerging"
horizon: "late 2020s"
categories: ["society", "cybernetics", "minds"]
tags: ["privacy", "neural data", "decoding", "surveillance", "law", "eeg"]
summary: "The question of what can be inferred about a person's mental states from neural and physiological recordings, and what protection that information should receive."
updated: "2026-07-27"
issues: ["The 2025 inner-speech decoding result and its mental-password safeguard carry no citation.", "The 2024 NeuroRights Foundation review of consumer privacy policies is named without a footnote."]
---

```infobox
{
  "caption": "Problem in neuroethics and data protection",
  "rows": [
    { "label": "Signals at issue", "value": "EEG, fMRI, implanted arrays" },
    { "label": "Best non-invasive decoding", "value": "Approximate gist of language" },
    { "label": "Training data required", "value": "Many hours per subject" },
    { "label": "Defeated by", "value": "Deliberate distraction" },
    { "label": "Legal category", "value": "Sensitive data in some states" },
    { "label": "Largest current exposure", "value": "Consumer wearables" }
  ]
}
```

**Mental privacy** is the interest a person has in controlling what others can learn about their thoughts, intentions, and internal states from measurements of the brain and body. It is the most-invoked of the proposed [[neurorights]] and the one where the gap between public expectation and technical reality is widest: no device can read an uncooperative person's thoughts, and the systems that infer the most about mental life mostly do not record from the brain at all.

## What non-invasive recording can do

Functional magnetic resonance imaging measures blood-oxygenation changes on a timescale of seconds, which is orders of magnitude slower than neural firing. Within that limit, the statistical methods surveyed in [[neural-decoding]] have achieved results that would have seemed impossible in 2010 — provided the decoder is trained on the individual whose brain it will read.

The clearest demonstration is a 2023 study in which a language model was used to reconstruct the approximate meaning of stories a participant heard, imagined, or watched while in the scanner.[^tang2023] The output was a paraphrase rather than a transcript — the decoder recovered gist, not words. Three constraints matter more than the result. Each decoder required roughly sixteen hours of scanner time with that specific participant. Decoders did not transfer across people. And the authors tested resistance directly: participants who counted, named animals, or silently told a different story defeated the decoder. Mental privacy, in that experiment, was preserved by ordinary volition.

Related work reconstructs approximate versions of seen images from fMRI using generative models, with the same requirements — large per-subject training sets, cooperation, and a stationary head in a magnet costing millions of dollars.

Electroencephalography, the signal available in consumer devices, is far cruder. Scalp electrodes sum the activity of very large populations of neurons through skull and skin, and dry-electrode headsets add substantial motion artifact. What such devices can do reliably is classify gross states: alert versus drowsy, high versus low task engagement, the presence of a stimulus-evoked response. What they cannot do is recover content. Marketing that describes consumer EEG as reading emotions or thoughts overstates the signal by a wide margin, in the same way that consumer claims for the stimulation devices covered in [[non-invasive-neuromodulation]] overstate their effects.

> [!caution] The reading a decoder gives is not a fact about the mind
> Decoders output the most probable interpretation under a trained model. A confident reconstruction from a mismatched model is confidently wrong, and the subject has no way to contest it. This is the same failure mode that discredited earlier physiological lie detection.

## Invasive decoding

Electrodes placed on or in cortex give a much better signal, and the results are correspondingly stronger. Two 2023 studies decoded attempted speech from participants with paralysis at rates in the range of sixty to eighty words per minute with substantial but usable error rates, using the penetrating [[utah-array|microelectrode arrays]] and the surface grids described in [[ecog-interfaces]] respectively.[^willett2023][^metzger2023] [[speech-neuroprosthesis]] covers the systems in detail.

These decode *attempted* speech — the motor commands a person issues when trying to talk. That is a form of expression, not private thought, and the distinction has been treated as ethically load-bearing. It became less clean in 2025, when researchers reported decoding *imagined* speech from motor cortex in participants with implanted arrays. The signal for inner speech was weaker and noisier than for attempted speech, but present. The same group proposed a safeguard that indicates where the field expects the problem to go: a mental password that the decoder must detect before it begins interpreting, so that inner speech is not decoded by default.

This is the first case in which a neural device has been designed against the possibility of reading its user's private thoughts, and it is worth noting what makes it tractable. The electrodes are surgically implanted, the decoder is trained on that person, and the person controls whether the system is powered. None of those conditions holds for the surveillance scenario that mental-privacy law is usually written about — and none of them is guaranteed by the commercial implant programmes described in [[neuralink]] and [[synchron]], where the decoder runs on infrastructure the user does not control.

## Legal treatment

Courts have consistently refused to admit neuroimaging as evidence of what a person knows or believes. A US federal court excluded fMRI-based lie detection in a 2010 fraud prosecution, finding the technique had not been shown reliable in real-world conditions, and the exclusion was upheld on appeal. India's Supreme Court held in 2010 that involuntary polygraph, narcoanalysis, and brain-signature testing violate the constitutional protection against self-incrimination and the right to personal liberty — a ruling prompted by the use of an EEG-based "brain electrical oscillation signature" technique in Indian criminal cases, including one murder conviction.

Data-protection law has moved faster than evidence law. EU law treats data concerning health and biometric identification as special categories requiring an explicit lawful basis, which covers clinical neural recordings. Several US states have added neural data to their sensitive-data definitions since 2024, alongside the constitutional and international instruments surveyed in the companion article on rights. All of these regimes regulate the *signal*, and the difficulty is that mental inference does not depend on neural signals.

Stimulation raises a separate and less-discussed question. A device that alters neural activity — the implanted systems in [[deep-brain-stimulation]], or the low-bandwidth transfers demonstrated in [[brain-to-brain-interface]] experiments — does not read a mental state but changes one, and the interest violated is integrity rather than privacy. Existing data-protection law has nothing to say about it.

## The inference problem

The most important fact about mental privacy in 2026 is that neural recording is a poor way to learn what someone is thinking compared with the alternatives already deployed. A phone's location history, search queries, purchase record, and typing latency support inferences about mood, health, political orientation, and intention that no EEG headset approaches, and the continuous heart-rate and sleep records collected by [[wearable-health-sensors|consumer wearables]] sit in the same category. Advertising systems have operated on this basis for two decades.

This has two consequences. Protecting neural data specifically produces the same incoherence diagnosed in the debate over genetic exceptionalism described in [[genetic-discrimination]]: the same inference is regulated or unregulated depending on which sensor produced it. And it means the realistic near-term risk is combination rather than decoding — a crude engagement signal from a headset joined to behavioural data that supplies the content the neural signal lacks.

> [!stat] Where the exposure actually is
> Workplace fatigue-monitoring headwear that detects microsleep from EEG is in commercial use in mining and long-haul transport. It records neural data continuously, in an employment relationship where consent is compromised, and is regulated in most jurisdictions as occupational safety equipment rather than as neurotechnology.

## Workplace and consumer monitoring

Employment is the likeliest first test of the doctrine, because it is where the recording is already happening and where refusal has a cost. Fatigue monitoring has a genuine safety rationale — driver microsleep kills people — and produces a continuous log of a worker's arousal that an employer can retain and correlate. Reports from 2018 described EEG-based monitoring of factory workers and train drivers in China; the scale was never independently verified, but the products exist and are marketed openly.

Consumer devices raise a different issue: the terms of service. A review published by the NeuroRights Foundation in 2024 of consumer neurotechnology companies' privacy policies found that most asserted broad rights over user neural data, including transfer to third parties, with few meaningful limits. The data is currently low in information content. The retention periods are indefinite, and the models that might later be applied to it have not been built.

## Open problems

Whether inner speech is protected by existing constitutional guarantees against self-incrimination is untested. The US doctrine distinguishes testimonial from physical evidence; a decoded thought sits uncomfortably between them, and no case has presented the question.

Whether decoders can be made to work without per-subject training is the technical threshold that would change the risk picture. Progress on subject-transferable models has been slow, and the anatomical variability between individual brains is a real obstacle rather than a data-scale problem — the same variability that makes the reconstruction pipelines of [[whole-brain-emulation]] a per-individual undertaking rather than a general method.

And there is no accepted method for validating a decoder's output. In a clinical [[brain-computer-interface]] the user corrects errors immediately, which supplies ground truth. In a surveillance application there is no such feedback, no way for the subject to demonstrate that the decoder is wrong, and no established standard a court could apply — the same structural problem that made polygraph evidence inadmissible, arriving now with better mathematics attached.

## See also

- [[neurorights]]
- [[neural-decoding]]
- [[speech-neuroprosthesis]]
- [[brain-computer-interface]]
- [[genetic-discrimination]]
- [[ecog-interfaces]]
- [[memory-prosthesis]]
- [[non-invasive-neuromodulation]]

## References

[^tang2023]: `paper` Tang, J., LeBel, A., Jain, S., Huth, A.G. "Semantic reconstruction of continuous language from non-invasive brain recordings." *Nature Neuroscience*, 2023. {Each decoder was trained on many hours of scanner data from one cooperating participant, recovered gist rather than words, and failed when the participant deliberately thought about something else.}
[^willett2023]: `paper` Willett, F.R. et al. "A high-performance speech neuroprosthesis." *Nature*, 2023. {A single participant with ALS and implanted intracortical arrays; the system decodes attempted speech, which is a motor act, not inner speech.}
[^metzger2023]: `paper` Metzger, S.L. et al. "A high-performance neuroprosthesis for speech decoding and avatar control." *Nature*, 2023. {A single participant with severe paralysis, using a surface electrode grid and a decoder trained on that person alone.}

==============================================================================
ARTICLE: metformin
TITLE:   Metformin and the TAME trial
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/metformin
SOURCE:  https://futurehumanwiki.com/raw/metformin
==============================================================================

---
title: "Metformin and the TAME trial"
slug: "metformin"
type: "intervention"
status: "contested"
horizon: "2030s"
trl: 4
categories: ["longevity"]
tags: ["metformin", "aging", "clinical trials", "drug repurposing", "regulation", "diabetes"]
summary: "A generic diabetes drug proposed as a geroprotector, and the large randomized trial designed to make aging itself an acceptable regulatory indication."
updated: "2026-07-27"
humanEvidence: "The geroprotective case in people is observational, drawn from diabetes prescribing records; no randomized trial has tested a composite age-related endpoint, and two trials in older adults found it blunted the response to exercise training."
access: "A generic prescription drug approved worldwide for type 2 diabetes and costing a few dollars a month; no regulator has approved it for aging, so any such use is off-label."
reversibility: "reversible"
issues: ["The Interventions Testing Program result for metformin is stated without a citation."]
---

```infobox
{
  "caption": "Drug and proposed clinical trial",
  "rows": [
    { "label": "Drug class", "value": "Biguanide" },
    { "label": "Clinical use since", "value": "1957 (France)" },
    { "label": "Proposed mechanism", "value": "Complex I inhibition, AMPK activation" },
    { "label": "Trial name", "value": "Targeting Aging with Metformin" },
    { "label": "Proposed size", "value": "~3,000 adults aged 65–79" },
    { "label": "Trial status", "value": "Not fully funded, not enrolling" },
    { "label": "Readiness", "value": "TRL 4" }
  ]
}
```

**Metformin and the TAME trial** are, respectively, the most widely prescribed oral diabetes drug in
the world and the study designed to test whether it delays the onset of age-related disease in people
who do not have diabetes. The pairing matters beyond the drug: TAME was constructed principally as a
regulatory experiment, an attempt to get a health authority to accept a composite of age-related
diseases as a legitimate trial endpoint and therefore to make aging a drug indication. As of 2026 the
trial has not been funded at the scale it requires and has not enrolled, and the evidence that
metformin slows human aging remains observational and contested.

## Origins of the claim

Metformin derives from guanidine compounds found in *Galega officinalis*, and entered clinical use in
France in 1957. Its geroprotective reputation rests largely on a 2014 observational analysis
reporting that people with type 2 diabetes taking metformin survived slightly longer than matched
controls without diabetes.[^bannister2014] The finding is arresting because diabetes shortens life;
a drug that erases and reverses that penalty would be doing something beyond glucose control.

The result has not held up cleanly. Comparisons of this kind are vulnerable to immortal-time bias,
confounding by indication, and healthy-adherer effects, and several reanalyses have argued that the
survival advantage attenuates or disappears once these are handled properly. Metformin is
preferentially prescribed to healthier, less complicated patients, and the comparator drug classes
used in such studies carry their own risks. The honest summary is that the observational signal is
real in the data and weak as evidence.

## Mechanism

Metformin's molecular action is still debated after nearly seventy years of use. It accumulates in
mitochondria and inhibits respiratory complex I at high concentrations, raising the AMP:ATP ratio
and activating AMP-activated protein kinase, which suppresses hepatic gluconeogenesis and shifts
cells toward catabolism. Partial inhibition of respiration is also the basis for the argument that
metformin works as a mild mitochondrial stressor, a hormetic reading that connects it to
[[mitochondrial-dysfunction|mitochondrial theories of aging]] and to the same stress-response case
made for [[heat-and-cold-exposure|heat and cold exposure]]. Work published in 2022 identified a
lower-dose, lysosome-based route to AMPK activation through the protein PEN2 that does not require
complex I inhibition, which matters because the concentrations reached in patients are far below
those used in most cell-culture experiments.[^ma2022]

Downstream, AMPK activation opposes mTORC1 signalling and promotes [[autophagy]], placing metformin
on the same nutrient-sensing axis as [[rapamycin]] and [[caloric-restriction]]. Additional proposed
mechanisms include effects on the gut microbiome, on intestinal glucose handling, on
[[inflammaging|chronic inflammatory signalling]], and on the senescence-associated secretory
phenotype of [[cellular-senescence|senescent cells]], which is why metformin is sometimes classed as
a senomorphic rather than a [[senolytics|senolytic]]. The multiplicity is itself a problem: a drug
credited with a dozen mechanisms usually has one that has not been identified, and a compound said to
touch most of the [[hallmarks-of-aging]] is as likely to be weakly non-specific as broadly effective.

## Evidence in animals

Animal data are weaker than the drug's reputation suggests. Metformin extends lifespan in nematodes,
and a widely cited study in a single inbred mouse strain reported a small gain in median lifespan at
a low dose, together with better metabolic and physical measures, while a tenfold higher dose was
toxic.[^martin2013] The US National Institute on Aging's Interventions Testing Program, the most
rigorous multi-site protocol available, has not reported lifespan extension from metformin alone in
genetically heterogeneous mice, and adding metformin to rapamycin has not clearly improved on what
rapamycin achieves by itself. For a compound whose case rests on human epidemiology, failing the
standard preclinical test is a substantial mark against it.

A 2024 study reported that long-term metformin treatment slowed several molecular and tissue markers
of aging in male cynomolgus monkeys, including brain measures.[^yang2024] It is the most direct
primate evidence available, and it involved small numbers of animals, a single sex, and molecular
endpoints rather than survival or function. It should be read as a hypothesis, not a confirmation.

> [!debate] The exercise interaction
> Two 2019 trials in older adults found that metformin blunted the mitochondrial adaptation to
> aerobic training[^konopka2019] and the hypertrophic response to resistance training.[^walton2019]
> If that result
> generalizes, a healthy older adult taking metformin may be trading away part of the benefit of
> [[exercise-and-aging|the best-evidenced geroprotector available]] for an unproven one.

## The TAME trial

TAME, short for Targeting Aging with Metformin, was designed by [[nir-barzilai]] and colleagues and
organized through the American Federation for Aging Research. As designed, it would enroll roughly
3,000 adults aged 65 to 79 across multiple US sites, randomize them to metformin or placebo, and
follow them for about six years. The primary endpoint would be time to first occurrence of any of a
composite of new age-related events: cardiovascular disease, cancer, dementia, and death.

That composite is the point. Regulators approve drugs for diseases, and no approval pathway exists
for "aging". By showing that a single agent delays the *first* of several unrelated age-related
diseases, TAME would demonstrate the [[geroscience-hypothesis]] operationally, showing that these
conditions share an upstream driver, and would establish a template any later geroprotector could
follow. US regulators met the investigators in the mid-2010s and indicated that such a composite
endpoint could be acceptable in principle.

The trial has never been fully funded. Metformin is generic and costs pennies, so no company can
recover a trial budget in the tens of millions of dollars from sales, and the public money that has
arrived falls far short of what a multi-centre outcome trial costs, funders having been reluctant to
finance a large trial of a drug whose preclinical record is mixed. TAME has become the standing
example of a market failure in geroscience: the interventions cheap enough to deploy widely are the
ones nobody can afford to test.

> [!key] What TAME is actually for
> Its designers have been explicit that the drug is a vehicle. The deliverable is a validated
> regulatory pathway and a set of accepted endpoints, which would matter as much for
> [[epigenetic-reprogramming|reprogramming therapies]] and senolytics as for metformin itself.

## Limitations and risks

Metformin is well tolerated but not inert. Gastrointestinal intolerance is common, long-term use
depletes vitamin B12 in a substantial minority of patients, and lactic acidosis, while rare, is the
reason the drug is contraindicated in severe renal impairment. None of these is prohibitive; all of
them matter differently when the recipient is healthy rather than diabetic, because the benefit side
of the ledger is speculative.

Randomized evidence outside diabetes has generally disappointed. Observational associations between
metformin use and lower cancer incidence were not confirmed by a large adjuvant breast-cancer trial,
which found no benefit.[^goodwin2022] Dementia findings remain mixed and observational. Meanwhile the
absence of an accepted surrogate means no trial can shortcut the timeline: movement in an
[[epigenetic-clock]] reading or another [[aging-biomarkers|composite aging measure]] is not currently
sufficient for approval, and the literature on [[biological-age|biological age]] offers no validated
substitute.

## Outlook

Two futures are plausible. In one, a philanthropic or public funder finally underwrites TAME or a
successor with a different drug, a composite age-related endpoint is tested in a proper randomized
design, and the field acquires the regulatory precedent it has wanted since the 2010s. In the other,
metformin's own evidence continues to erode, with the observational signal reanalyzed away and the
exercise interaction confirmed, and the regulatory argument moves to a compound with better
preclinical support.

Either way the interesting question is no longer whether metformin is a longevity drug. It is whether
any institution will pay to find out, given that the answer has no owner and the drug has no patent —
a question about how public goods get funded rather than about biology, and one that will decide how
quickly claims about [[compression-of-morbidity|compressed morbidity]] can be tested at all.

## See also

- [[geroscience-hypothesis]]
- [[rapamycin]]
- [[caloric-restriction]]
- [[exercise-and-aging]]
- [[aging-biomarkers]]
- [[senolytics]]
- [[longevity-dividend]]
- [[healthspan]]

## References

[^bannister2014]: `paper` Bannister, C.A. et al. "Can people with type 2 diabetes live longer than those without? A comparison of mortality in people initiated with metformin or sulphonylurea monotherapy and matched, non-diabetic controls." *Diabetes, Obesity and Metabolism*, 2014. {An observational comparison of prescription records rather than a trial; later reanalyses argue the survival advantage reflects confounding by indication and healthy-adherer effects.}
[^ma2022]: `paper` Ma, T. et al. "Low-dose metformin targets the lysosomal AMPK pathway through PEN2." *Nature*, 2022.
[^martin2013]: `paper` Martin-Montalvo, A. et al. "Metformin improves healthspan and lifespan in mice." *Nature Communications*, 2013. {A single inbred mouse strain: the low dose gave a small median gain and a tenfold higher dose was toxic.}
[^yang2024]: `paper` Yang, Y. et al. "Metformin decelerates aging clock in male monkeys." *Cell*, 2024. {Small numbers of male cynomolgus monkeys, with molecular and tissue markers as endpoints; neither survival nor function was measured.}
[^konopka2019]: `paper` Konopka, A.R. et al. "Metformin inhibits mitochondrial adaptations to aerobic exercise training in older adults." *Aging Cell*, 2019.
[^walton2019]: `paper` Walton, R.G. et al. "Metformin blunts muscle hypertrophy in response to progressive resistance exercise training in older adults." *Aging Cell*, 2019.
[^goodwin2022]: `paper` Goodwin, P.J. et al. "Effect of metformin vs placebo on invasive disease-free survival in patients with breast cancer: the MA.32 randomized clinical trial." *JAMA*, 2022.

==============================================================================
ARTICLE: methuselah-foundation
TITLE:   Methuselah Foundation
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/methuselah-foundation
SOURCE:  https://futurehumanwiki.com/raw/methuselah-foundation
==============================================================================

---
title: "Methuselah Foundation"
slug: "methuselah-foundation"
type: "organization"
status: "established"
horizon: "present"
categories: ["organizations", "longevity"]
tags: ["aging", "prizes", "nonprofit", "tissue engineering", "organ preservation", "research funding"]
summary: "A Virginia nonprofit founded in the early 2000s that used prize competitions to seed research on life extension, regenerative medicine and organ engineering."
updated: "2026-07-27"
issues: ["The Thiel donation and the New Organ Liver Prize sums are stated without a source."]
---

```infobox
{
  "caption": "Research nonprofit",
  "rows": [
    { "label": "Founded", "value": "Early 2000s" },
    { "label": "Founders", "value": "David Gobel, Aubrey de Grey", "link": "/wiki/aubrey-de-grey" },
    { "label": "Headquarters", "value": "Springfield, Virginia" },
    { "label": "Mission statement", "value": "Make 90 the new 50 by 2030" },
    { "label": "Best-known programme", "value": "The Mprize" },
    { "label": "Spin-out", "value": "SENS Research Foundation", "link": "/wiki/sens-research-foundation" },
    { "label": "Venture arm", "value": "Methuselah Fund" }
  ]
}
```

**Methuselah Foundation** is a nonprofit organization founded in the early 2000s by the entrepreneur David Gobel and the biogerontologist [[aubrey-de-grey]] to accelerate research on extending healthy human life, principally by running prize competitions. Its stated mission is to "make 90 the new 50 by 2030". It is best known for the Mprize, which rewarded record-setting mouse longevity, and for the New Organ prizes in regenerative medicine, and it spun out the [[sens-research-foundation]] in 2009.

## Overview

The foundation's method is deliberate: identify a bottleneck, define a measurable target, and offer money for hitting it rather than for proposing to try. Prizes shift the risk from funder to competitor, attract entrants a grant committee would not select, and produce a public benchmark. They also work only where the target is unambiguous, which has constrained what the foundation could take on.

Its second function has been convening. For most of the 2000s, when aging research attracted little private money, the foundation was one of very few places where researchers, donors and advocates in the field met at all, and a substantial share of the modern longevity ecosystem traces to those connections.

## The Mprize

The Methuselah Mouse Prize, launched in 2003 and usually shortened to the Mprize, is the organization's signature programme. It awarded money in two categories, and the distinction between them was the point.

The **longevity** prize went to the oldest mouse ever recorded, by any means. Andrzej Bartke won it in 2003 with a growth-hormone-receptor-deficient dwarf mouse that lived beyond 1,800 days; his group's report of the underlying strain remains one of the largest proportional lifespan extensions recorded in a mammal.[^bartke2001]

The **rejuvenation** prize was the more interesting design. It required the intervention to begin in middle age, which excludes developmental manipulations and lifelong regimens and demands that something be done to an already-aged animal. Stephen Spindler won it in 2004 with late-onset [[caloric-restriction]].

> [!key] Why late-onset mattered
> A drug that must be given from birth is not a medicine anyone can take. By paying only for interventions started after middle age, the rejuvenation prize enforced the constraint that separates a laboratory result from a possible therapy, and it did so a decade before the field generally adopted it. Nearly every intervention now taken seriously in aging biology is assessed on late-onset dosing.

The Mprize's leaderboard was never displaced by a wide margin, which is itself a finding. Two decades of work on mouse longevity has not produced an intervention that dramatically exceeds what dwarf mutants and caloric restriction achieved, and the strongest pharmacological signal in the field — [[rapamycin]] fed to genetically heterogeneous mice starting in late life, in the National Institute on Aging's Interventions Testing Program — produced extensions measured in percentages rather than multiples.[^harrison2009]

## The New Organ prizes

From 2014 the foundation extended the prize model to regenerative medicine under the New Organ banner, motivated by the [[organ-shortage]] arithmetic: transplant waiting lists are limited by donor supply, and no amount of surgical improvement changes that.

The New Organ Liver Prize offered a seven-figure award for a bioengineered liver that could sustain a large animal for a defined period. It went unclaimed, which is an accurate reflection of where [[organ-bioprinting]] and [[tissue-engineering]] stand: vascularization at scale remains unsolved, and no whole solid organ has been grown and successfully transplanted.

The foundation also partnered with NASA on the Vascular Tissue Challenge, a competition to produce thick, metabolically functional vascularized human tissue that survives in the laboratory for a month. That prize was awarded in 2021 to a team from Wake Forest, and it is the clearest demonstration that a well-specified target in this area can be met.

A related programme, the Organ Preservation Alliance, worked on extending the viability window for donated organs — the problem that [[cryonics|vitrification]] and normothermic perfusion research addresses, and one where progress has real near-term clinical value regardless of what one believes about life extension.

```timeline
[
  { "year": "2003", "title": "Mprize launched", "text": "Two prizes are announced: one for the longest-lived mouse, one for the greatest late-onset extension, the latter designed to reward rejuvenation rather than slowed development." },
  { "year": "2003–2004", "title": "First winners", "text": "Andrzej Bartke wins the longevity category with a dwarf mouse living beyond 1,800 days; Stephen Spindler wins the rejuvenation category with late-onset caloric restriction." },
  { "year": "2006", "title": "Major donation", "text": "Peter Thiel commits several million dollars, one of the first large private gifts to life-extension research and a signal to other technology donors." },
  { "year": "2009", "title": "SENS spun out", "text": "The damage-repair research programme is separated into an independent organization, SENS Research Foundation." },
  { "year": "2014", "title": "New Organ", "text": "Prize competitions are extended to bioengineered organs and to organ preservation." },
  { "year": "2021", "title": "Vascular Tissue Challenge awarded", "text": "A team from Wake Forest wins the NASA competition run in partnership with the foundation, producing thick vascularized tissue that survives for a month." }
]
```

## Funding

The foundation is philanthropically funded, and its most consequential single gift was from Peter Thiel in 2006 — several million dollars at a time when almost no private capital of that size was directed at aging. The gift mattered less for its amount than for its signalling effect on other technology donors, and much of the money that later founded [[calico]], [[altos-labs]] and [[retro-biosciences]] came from the same social network.

The organization later established the Methuselah Fund, a venture vehicle that invests in early-stage longevity and regenerative medicine companies, on the reasoning that some of the field's bottlenecks are commercial rather than scientific and that a nonprofit prize cannot address them.

## Reception

The foundation has attracted less criticism than most organizations in the field, largely because prizes make weaker claims than research programmes do. A prize does not assert that a result is achievable; it offers money if someone achieves it.

The substantive criticism is about proportion. The sums involved are small — hundreds of thousands to low millions of dollars — relative to the cost of the work being incentivized, and the prizes have generally been won by researchers who were already doing the work rather than attracting new entrants. The Mprize's most-cited effect was reputational: it gave mouse longevity research a public scoreboard and made the field legible to donors.

A second criticism concerns the mission statement. "Make 90 the new 50 by 2030" is not a testable proposition, has no operational definition in terms of [[healthspan]] or function, and as of 2026 no route to meeting it on that timeline has been described. The foundation has continued to use it, which its critics read as the same rhetorical looseness that characterizes the [[longevity-escape-velocity]] literature more generally.

## Legacy

Methuselah's importance is as an origin point. It funded de Grey's early work, including the period in which he assembled the damage-repair framework later set out in *Ending Aging*,[^degrey2007] then spun out the organization that carried it, seeded the donor relationships that made large-scale private longevity funding possible, and demonstrated that prizes could operate in biology where they had previously been used mainly in aerospace — a lesson [[xprize-healthspan]] applied two decades later at far larger scale.

What it did not do is produce a therapy, which is the correct standard by which to judge a research funder and the one on which the entire field, not just this organization, has so far failed. The open question for the prize model in biology is whether the endpoints that matter can be specified precisely enough to compete over. A mouse's age is unambiguous. Restored human function is not, and every attempt to define it — including the current one — has had to invent its own measures.

## See also

- [[aubrey-de-grey]]
- [[sens-research-foundation]]
- [[xprize-healthspan]]
- [[organ-shortage]]
- [[organ-bioprinting]]
- [[rapamycin]]
- [[caloric-restriction]]
- [[longevity-escape-velocity]]

## References

[^degrey2007]: `book` de Grey, A. and Rae, M. *Ending Aging: The Rejuvenation Breakthroughs That Could Reverse Human Aging in Our Lifetime*. St. Martin's Press, 2007. {A popular exposition of the author's own research agenda; it sets out a programme rather than reporting results.}
[^bartke2001]: `paper` Bartke, A. et al. "Extending the lifespan of long-lived mice." *Nature*, 2001. {The record rests on a germline dwarfing mutation present from birth, which is the case the prize's separate rejuvenation category was written to exclude.}
[^harrison2009]: `paper` Harrison, D.E. et al. "Rapamycin fed late in life extends lifespan in genetically heterogeneous mice." *Nature*, 2009. {Run at three independent sites on genetically heterogeneous mice, with dosing begun in late life; the gain is a percentage of median lifespan.}

==============================================================================
ARTICLE: microgravity-adaptation
TITLE:   Microgravity adaptation
PORTAL:  Space & Extreme Environments
URL:     https://futurehumanwiki.com/wiki/microgravity-adaptation
SOURCE:  https://futurehumanwiki.com/raw/microgravity-adaptation
==============================================================================

---
title: "Microgravity adaptation"
slug: "microgravity-adaptation"
type: "concept"
status: "established"
horizon: "present"
categories: ["space", "bodies"]
tags: ["microgravity", "spaceflight", "vestibular", "bone loss", "countermeasures", "artificial gravity"]
summary: "The set of physiological and neural changes by which the human body reorganises itself for weightlessness, and the countermeasures that limit the cost of returning."
updated: "2026-07-27"
humanEvidence: "Measured directly in astronauts across decades of orbital flight and reproduced on the ground by head-down bed rest; no human data exist for sustained exposure to gravity between zero and one."
issues: ["Post-2008 improvements in bone and muscle outcomes aboard the ISS are stated without a citation."]
---

```infobox
{
  "caption": "Physiological adaptation process",
  "rows": [
    { "label": "Trigger", "value": "Sustained free fall" },
    { "label": "Systems affected", "value": "Vestibular, cardiovascular, skeletal" },
    { "label": "Onset", "value": "Hours to weeks" },
    { "label": "Primary countermeasure", "value": "Resistive and aerobic exercise" },
    { "label": "Artificial gravity", "value": "Never flown with crew" },
    { "label": "Partial-gravity data", "value": "None for humans" }
  ]
}
```

**Microgravity adaptation** is the process by which the human body reconfigures itself for an environment without a gravitational reference: the vestibular system reinterprets its own signals, the cardiovascular system resets around a lost hydrostatic gradient, and bone and muscle remodel to the mechanical loads actually being applied, which is almost none. The changes are largely appropriate to weightlessness and largely maladaptive on return. Adaptation, not damage, is the correct frame for most of what happens in orbit; the clinical problem arrives at the other end.

```keyfacts
[
  { "value": "~10%", "label": "Plasma volume lost within days", "note": "the main driver of orthostatic intolerance after landing" },
  { "value": "0.38 g", "label": "Surface gravity on Mars", "note": "no sustained human or mammalian data exists at this level" },
  { "value": "0", "label": "Crewed vehicles flown with artificial gravity", "note": "a brief tethered spin in 1966 produced a negligible acceleration" }
]
```

## Sensorimotor and vestibular reorganisation

The otolith organs of the inner ear detect linear acceleration, and on Earth the constant 1 g field lets the brain read otolith output as head tilt. In free fall that inference breaks: otolith signals no longer correlate with orientation, while the semicircular canals continue to report rotation normally. The mismatch is the standard explanation for space motion sickness, which affects a majority of crew members in the first days of flight and usually resolves within about three days.

What follows is genuine neural plasticity. The brain reweights vestibular input against vision and proprioception, and by the end of the first week most crew move competently in a volume with no floor. Reaching, gaze stabilisation, and head-eye coordination all recalibrate. The reweighting is the same class of process that underlies the adaptation to novel signals discussed in [[sensory-augmentation]], and it runs in reverse after landing, which is why returning crew show postural instability, gait ataxia, and illusory tilt during head movements for days to weeks.

Recovery time scales with mission duration, and after year-long flights some sensorimotor measures take months. Ground countermeasures include balance training before and after flight; galvanic vestibular stimulation and related methods from [[non-invasive-neuromodulation]] have been investigated as a way to pre-adapt crew, without becoming operational practice.

## Cardiovascular set points

Standing on Earth means maintaining a column of blood against gravity. Remove that column and roughly two litres of fluid shift headward within hours, producing the facial puffiness and thinned legs familiar from orbital video. The body interprets the central volume increase as excess and eliminates it: plasma volume falls by about a tenth in the first days, red cell mass follows, and the system stabilises at a lower total volume.

Red cell loss is not simply reduced production. Measurements of carbon monoxide exhalation in crew members indicate elevated haemolysis that persists throughout flight rather than resolving after an initial adjustment, a finding that reframed what was long called space anaemia.[^trudel2022]

The adapted state is stable in orbit and poorly suited to a planet. Orthostatic intolerance on landing is common after long-duration flights, and countermeasures are unglamorous and effective: fluid and salt loading before re-entry, compression garments, and lower-body negative pressure sessions in flight. Lower-body negative pressure is also under study as a way to restore a footward fluid gradient and mitigate the ocular syndrome described in [[space-medicine]].

## The musculoskeletal response

Bone remodels to the loads it experiences. In weightlessness the mechanical signal for maintenance disappears at weight-bearing sites, and resorption outpaces formation at roughly one to one and a half per cent of bone mineral density per month at hip and lumbar spine. The pattern is regional and predictable: the calcaneus and femoral neck lose heavily, the skull does not. Trabecular bone, with its higher surface-to-volume ratio, responds fastest.

Muscle follows the same logic. Postural extensors and the calf complex atrophy preferentially, with a fibre-type shift toward faster and more fatigable phenotypes and a disproportionate loss of force relative to cross-sectional area. The clinical resemblance to [[stem-cell-exhaustion|age-related sarcopenia]] and to intensive-care-acquired weakness is close enough that countermeasure research crosses between the fields, and pharmacological approaches such as [[myostatin-inhibition]] have been proposed for spaceflight for the same reason they are proposed for sarcopenia — and rest on the same unresolved question of whether added muscle mass delivers proportionate function. Whether unloading engages the same machinery as the [[hallmarks-of-aging|hallmarks of aging]] or merely produces a phenotype that resembles them is unresolved, and the reversibility of most spaceflight losses argues for the second reading.

> [!key] Disuse, not decay
> Almost everything in this section is a response to unloading rather than to any exotic property of space. That is why head-down bed rest and dry immersion reproduce most of it on the ground, why the changes largely reverse, and why exercise works. Radiation is the one major spaceflight stressor with no terrestrial analogue and no countermeasure — see [[radiation-hardening-humans]].

## Countermeasures aboard the ISS

Early orbital stations demonstrated that exercise mattered and that the equipment available was inadequate. The decisive change came with the Advanced Resistive Exercise Device, installed in 2008, which uses vacuum cylinders to deliver free-weight-like loads up to several hundred pounds and permits squats, deadlifts, and heel raises rather than only cycling and treadmill work. Crews now perform roughly two hours daily across resistive, treadmill, and cycle ergometer sessions, with the treadmill harness supplying the ground reaction force that free fall does not.

Outcomes improved markedly. Bone mineral density loss and muscle loss both fell relative to the pre-2008 baseline, though neither reached zero, and functional performance immediately after landing remains degraded. Studies of higher-intensity, lower-volume protocols have examined whether the same protection can be obtained in less crew time, which is a scarce resource aboard a research station.

Nutrition and pharmacology supplement exercise rather than replacing it. Adequate energy and protein intake, vitamin D supplementation, and reduced sodium all affect bone balance, and bisphosphonates combined with resistive exercise preserve bone better than exercise alone.[^leblanc2013] As on Earth, and as [[exercise-and-aging]] argues at length, mechanical loading remains the intervention that any drug has to beat.

Two design pressures push against this arrangement. Exercise hardware is heavy and bulky, which has motivated proposals for wearable loading suits and powered [[exoskeleton|exoskeletons]] compact enough for a transit vehicle, none of which has flown operationally. And crew time is finite: any architecture that places the crew in [[human-hibernation|induced torpor]] for the cruise removes the exercise countermeasure entirely, and would have to rely on whatever protection the torpid state itself confers.

## Artificial gravity

Rotating a vehicle produces centripetal acceleration that the body cannot distinguish from gravity, and would address every unloading effect at once. The engineering constraints are unforgiving. Acceleration scales with the square of angular velocity and linearly with radius, so a comfortable rotation rate demands a large structure. Head movements in a rotating frame generate cross-coupled Coriolis stimulation of the semicircular canals, producing nausea and disorientation; classical estimates put the tolerable rate at only a few revolutions per minute, though incremental adaptation training raises tolerance substantially in ground centrifuges.

No crewed spacecraft has ever flown with artificial gravity. A tethered spin during Gemini 11 in 1966 produced an acceleration far too small to be physiologically relevant and remains the only in-flight attempt.

Short-radius centrifuges offer a compromise: instead of spinning the habitat, spin the crew member for a scheduled period each day. Bed rest campaigns including ESA and DLR's AGBRESA study have tested daily centrifugation during sixty days of head-down tilt and found partial protection of cardiovascular and sensorimotor measures, with less clear benefit for bone. A short radius also imposes a steep gravity gradient along the body, with the feet experiencing considerably more acceleration than the head. Rotation complicates every other subsystem as well, from fluid handling in [[closed-loop-life-support|life support]] to docking, which is part of why no programme has committed to it.

## Partial gravity

The most consequential gap in the evidence base is the shape of the dose-response curve between 0 and 1 g. Mars provides 0.38 g and the Moon 0.166 g. If the physiological response is threshold-like, and even a third of a gravity preserves most bone and muscle, then surface missions are far easier than transit. If it is roughly linear, a Mars settlement faces permanent, progressive skeletal loss.

No data exist. Apollo surface stays were measured in days. Rodent partial-gravity centrifuge experiments are few, small, and short, and their results have not converged. A centrifuge facility capable of housing animals at lunar and Martian gravity for meaningful periods has been proposed repeatedly and never flown at scale. Every architecture study for [[pantropy|off-world settlement]] and every projection about children born on Mars rests on an unmeasured parameter, and the same gap propagates into the reproductive questions raised by [[generation-ship-biology]].

## Readaptation and its limits

Most measures return to preflight baseline within weeks to months. Some do not. Bone at certain sites remains below baseline years after return, and the recovered bone may differ in architecture from what was lost even when density readings normalise, which matters more for fracture risk than density alone suggests.[^sibonga2007] Structural changes in the eye and in brain ventricular volume also resolve incompletely.

Whether repeated long-duration flights compound these residuals is a question the astronaut corps is too small to answer with confidence, and it becomes pressing as flight opportunities increase. For a permanent settlement the question inverts entirely: readaptation to Earth would be the exceptional event, and a population that never experiences 1 g would develop a skeleton adapted to its own world — the point at which spaceflight physiology stops being a medical specialty and becomes a question about which species the settlers belong to.

## See also

- [[space-medicine]]
- [[radiation-hardening-humans]]
- [[human-hibernation]]
- [[closed-loop-life-support]]
- [[generation-ship-biology]]
- [[pantropy]]
- [[exercise-and-aging]]
- [[exoskeleton]]

## References

[^trudel2022]: `paper` Trudel, G., Shafer, J., Laneuville, O., Ramsay, T. "Hemolysis contributes to anemia during long-duration space flight." *Nature Medicine*, 2022. {A small group of ISS crew members, with red cell destruction inferred from exhaled carbon monoxide rather than measured directly.}
[^leblanc2013]: `paper` LeBlanc, A. et al. "Bisphosphonates as a supplement to exercise to protect bone during long-duration spaceflight." *Osteoporosis International*, 2013.
[^sibonga2007]: `paper` Sibonga, J.D. et al. "Recovery of spaceflight-induced bone loss: bone mineral density after long-duration missions as fitted with an exponential function." *Bone*, 2007. {Fits a recovery curve to post-flight bone mineral density readings; density is not the same as the bone architecture that governs fracture risk.}

==============================================================================
ARTICLE: mind-uploading
TITLE:   Mind uploading
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/mind-uploading
SOURCE:  https://futurehumanwiki.com/raw/mind-uploading
==============================================================================

---
title: "Mind uploading"
slug: "mind-uploading"
type: "concept"
status: "speculative"
horizon: "2050s+"
categories: ["minds"]
tags: ["uploading", "emulation", "identity", "consciousness", "brain preservation", "functionalism"]
summary: "The hypothetical transfer of a person's mind to a computational substrate, such that the resulting process is that person rather than a description of them."
updated: "2026-07-27"
issues: ["The Brain Preservation Foundation prize awards of 2016 and 2018 carry no citation."]
---

```infobox
{
  "caption": "Speculative concept in philosophy of mind and neurotechnology",
  "rows": [
    { "label": "Also called", "value": "Substrate transfer, uploading" },
    { "label": "Technical prerequisite", "value": "Whole brain emulation" },
    { "label": "Philosophical premise", "value": "Computational functionalism" },
    { "label": "Main routes", "value": "Destructive scan; gradual replacement" },
    { "label": "Demonstrations", "value": "None, in any organism" },
    { "label": "Mainstream view", "value": "Not near; feasibility disputed" }
  ]
}
```

**Mind uploading** is the hypothetical transfer of a person's mental life onto a non-biological
substrate, usually a computer, such that the resulting process has that person's memories,
personality, and — on the strong version of the claim — that person's continued experience. It
combines a technical proposal, [[whole-brain-emulation]], with a philosophical claim about what
makes a future being the same person as an earlier one. The two components fail independently:
the emulation could work while the person does not survive it, and the identity argument could be
correct while the engineering never arrives.

## Overview

Discussion of uploading tends to collapse three separable questions. First, can a computational
model of a specific brain be built and run so that it behaves as the original would? That is an
empirical question about scanning, modelling, and computation. Second, would such a model be
conscious — would there be something it is like to be it? That depends on unsettled theory, treated
under [[machine-consciousness]] and [[neural-correlates-of-consciousness]]. Third, if the model is
conscious, is it the original person or a very good imitation? That is a question in
[[personal-identity-and-continuity]], and it is not answerable by measurement.

Advocates generally hold that the answers are yes, yes, and yes, resting on
[[substrate-independence]] — the view that mental states depend on the organization of a system
rather than the material implementing it. Critics reject one or more of the three at different
points, and the disagreements do not line up: a neuroscientist may doubt the engineering while
accepting functionalism, and a philosopher may accept the engineering while denying that the upload
would be the same person.

## Two routes

### Destructive scanning

The standard proposal is to fix or preserve a brain, image it at nanometre resolution, reconstruct
its structure using the methods of [[connectomics]], infer the parameters needed for a dynamical
model, and run that model — the scan-translate-run pipeline set out in Anders Sandberg and Nick
Bostrom's 2008 roadmap.[^sb2008] The brain is destroyed in the process — sectioning tissue for electron
microscopy is not reversible. The upload therefore begins after the biological person is gone, which
is what makes the identity question acute and what makes the procedure legally and ethically
fraught even if it worked.

### Gradual replacement

The alternative, described in detail by Hans Moravec in 1988 and by many writers since, replaces
neurons one at a time with functionally equivalent artificial devices while the person remains
awake and continuous.[^moravec1988] The appeal is that it avoids the discontinuity: at no moment is
there a gap, a copy, or two claimants. Nothing resembling the required device exists. It would have
to sit inside the skull, read the inputs and outputs of a single cell with high fidelity, reproduce
its computation including dendritic nonlinearities and neuromodulatory sensitivity, and do so
without disrupting neighbours — an application well beyond anything discussed under
[[medical-nanorobots]]. The gradual route is philosophically cleaner and technically far harder.

> [!note] Terminology
> "Uploading" implies a transfer, as of a file. Nothing in the destructive procedure transfers
> anything; it reads a structure and instantiates a new process elsewhere. Whether that counts as
> transfer is the whole dispute, and the word quietly assumes an answer.

## What would have to be preserved

The minimum specification is unknown, which is the central practical problem. A wiring diagram alone
is insufficient: it omits synaptic sign and strength, receptor composition, short-term plasticity,
and the neuromodulatory state that determines how a circuit behaves at a given moment. Long-term
memory is generally thought to be stored in synaptic weights, but the molecular basis is
contested — persistent kinase activity, structural changes at the spine, and nuclear or epigenetic
components have all been proposed, and experiments transferring behavioural sensitization between
sea slugs by RNA injection were read by some as evidence that not everything lives at the
synapse.[^glanzman2018] Whatever the resolution, it sets the resolution the scan must reach.

Two further complications are routinely underweighted: neuromodulators act by volume transmission,
so the quantity that matters is a concentration field rather than a connection, and glia modulate
synaptic transmission while being absent from most proposed models entirely.

There is a serious counterargument, made most carefully by proponents themselves: biological neurons
are noisy. Synaptic vesicle release is probabilistic, membrane potentials fluctuate thermally, and
the same stimulus produces different spike trains on repetition. If a brain's behaviour is robust to
that much internal variation, then detail below the noise floor cannot be functionally load-bearing,
and an emulation need only match the brain at whatever grain the noise permits. The reply is that
noise robustness at the level of individual events does not establish robustness of the parameters
that generate them, and that nobody has measured where the floor sits. This is the crux, and it is
an empirical question that current methods cannot answer.

## The copy problem

Suppose the emulation runs and behaves indistinguishably. If the original brain still exists, there
are plainly two beings, and the emulation is a copy — which suggests that in the destructive case it
is also a copy, merely one with no competitor. Derek Parfit's response is to deny that the question
has a further answer: what matters in survival is psychological continuity and connectedness, and
identity is not a deep fact over and above these.[^parfit1984] On that view the copy carries
everything worth caring about, and the fact that it is a copy is not an additional loss. The
argument is developed in the [[teleportation-problem]].

The competing intuition is that experience is a process, not a pattern, and that a process which
stops has stopped. This is why gradual replacement is treated as the safer route by people who
accept uploading in principle. It is also why some who reject uploading nonetheless accept
[[brain-computer-interface]] augmentation and even [[human-ai-merger]] framings: incremental
extension of a continuously running system does not raise the branching problem.

## Why most neuroscientists are skeptical

The mainstream position is not that uploading is impossible but that it is nowhere close, and that
the confident dates circulating in [[transhumanism]] and [[singularitarianism]] have no basis in the
state of the science. The objections to the engineering itself are catalogued in the article on
whole brain emulation: the single-neuron model is unsettled, chemical signalling is invisible to a
wiring diagram, and the largest reconstructed volume of human cortex is a millionth of a brain.
Three consequences bear specifically on uploading rather than on emulation in general.

**The smallest available test case has already resisted.** *C. elegans* has 302 neurons and a wiring
diagram published in 1986. Decades of effort have not produced a model that reproduces the worm's
behavioural repertoire, because the diagram fixes neither synaptic sign and strength nor the
neuropeptide signalling that reconfigures the circuit. Uploading assumes those same missing
parameters can be recovered for a human brain from one fixed specimen, with no possibility of
returning to the living animal to measure them.

**The uncertainties compound rather than add.** An upload requires the scan to capture enough, the
neuron model to be correct, the parameter inference to be sound, and the resulting dynamics to stay
stable over subjective years. Each step is independently unproven, a chain is no stronger than its
weakest link, and nobody can attach a figure to any of the links. This is why estimates of when
uploading becomes possible carry so little information: they are products of guesses about
quantities nobody has measured.

**Destroying the original removes the reference standard.** Ordinary engineering validates a model
against the system it models. Here the system is consumed by the measurement, so the only remaining
check is whether the emulation strikes people who knew the person as being that person — a test that
a sufficiently good imitation also passes. Staged validation on animals, the standard proposal, has
not begun.

> [!caution] "Not near" is not "impossible"
> Very few neuroscientists claim that brains do something physically uncomputable, and the
> mainstream objection is not vitalist. David Chalmers, who defends the functionalist premise the
> whole proposal rests on, still treats human uploading as a technology for some distant century
> rather than a foreseeable one.[^chalmers2022] The disagreement is about recoverable parameters,
> not about whether brains are physical.

## Preservation as a hedge

Because uploading cannot be attempted now, its practical expression is preservation. The argument is
that if the information an emulation would need survives in fixed or vitrified tissue, then
preservation converts an immediate deadline into a storage problem. This is the rationale for
[[brain-preservation]] and for the neuro option offered by [[cryonics]] providers such as
[[alcor]]. The Brain Preservation Foundation's prizes, awarded in 2016 and 2018 for demonstrating
that aldehyde-stabilized cryopreservation retains connectomic structure in small and large mammal
brains, were explicitly framed around this criterion.

The commercial expression has been unstable. Nectome, a startup founded by one of the prize
winners, attracted attention in 2018 for proposing preservation of terminally ill volunteers by a
procedure that is necessarily fatal; MIT ended its research association with the company shortly
afterwards.[^regalado2018] No preservation provider can demonstrate that the preserved information
is sufficient, because no one knows what would be sufficient.

## Culture, institutions, and claims

The idea entered wide circulation through fiction — Arthur C. Clarke's stored citizens, Rudy
Rucker's *Software*, Greg Egan's *Permutation City* — and through advocacy. The Carboncopies
Foundation promotes substrate-independent minds as a research programme; Russia's 2045 Initiative
announced a staged plan culminating in uploaded consciousness by 2045, with no published technical
progress. Economist Robin Hanson's *The Age of Em* takes emulation as given and works out the
economics of a world in which copies are cheap, which is useful as an exploration of consequences
and is explicitly not a forecast that emulation will arrive.[^hanson2016]

Uploading is also the strongest form of [[digital-immortality]], and the weakest form — a language
model trained on someone's messages — is routinely marketed using the same vocabulary. The
conflation matters: one is a hypothetical continuation of a person, the other is a stylistic
imitation with no claim to continuity.

## Open problems

Beyond feasibility, uploading raises questions with no settled treatment. If copying is cheap,
ordinary assumptions about scarcity of persons, consent, and legal identity fail; a copy made
without permission is a wrong with no existing category. If emulations can be run at varying speeds,
subjective time becomes a resource that can be bought, which connects the topic to
[[access-and-inequality]]. If an emulation can be paused, deleted, or forked, the moral status of
those operations depends on whether it is conscious, and no test for that exists. And if uploads
are possible at all, questions about ending one's own existence — treated under
[[right-to-die-and-right-to-live]] — arise in a form the law has never faced, since deletion,
suspension, and restoration from backup are all available.

The prior question remains empirical and unanswered: whether there is any level of description at
which a brain's causal structure can be captured and re-run, and if so, how far down it sits.

## See also

- [[whole-brain-emulation]]
- [[substrate-independence]]
- [[personal-identity-and-continuity]]
- [[teleportation-problem]]
- [[brain-preservation]]
- [[digital-immortality]]
- [[connectomics]]
- [[machine-consciousness]]

## References

[^moravec1988]: `book` Moravec, H. *Mind Children: The Future of Robot and Human Intelligence*. Harvard University Press, 1988.

[^parfit1984]: `book` Parfit, D. *Reasons and Persons*. Oxford University Press, 1984.

[^chalmers2022]: `book` Chalmers, D. J. *Reality+: Virtual Worlds and the Problems of Philosophy*. W. W. Norton, 2022.

[^glanzman2018]: `paper` Bédécarrats, A., Chen, S., Pearce, K., Cai, D. and Glanzman, D. L. "RNA from trained *Aplysia* can induce an epigenetic engram for long-term sensitization in untrained *Aplysia*." *eNeuro*, 2018. {Sea slugs and a simple defensive reflex, not episodic memory; the interpretation that memory is stored outside synapses is disputed.}

[^regalado2018]: `news` Regalado, A. "A startup is pitching a mind-uploading service that is '100 percent fatal'." *MIT Technology Review*, 2018. {Reports a company's pitch and the university's response; it is evidence about what was offered, not that any mind was preserved.}

[^hanson2016]: `book` Hanson, R. *The Age of Em: Work, Love and Life when Robots Rule the Earth*. Oxford University Press, 2016.

[^sb2008]: `report` Sandberg, A. and Bostrom, N. "Whole Brain Emulation: A Roadmap." Technical Report 2008-3, Future of Humanity Institute, University of Oxford, 2008. {A workshop document setting out what emulation would require at each level of detail; it maps requirements and reports no experiment.}

==============================================================================
ARTICLE: mirror-life
TITLE:   Mirror life
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/mirror-life
SOURCE:  https://futurehumanwiki.com/raw/mirror-life
==============================================================================

---
title: "Mirror life"
slug: "mirror-life"
type: "risk"
status: "speculative"
horizon: "2050s+"
categories: ["genetics", "society"]
tags: ["synthetic biology", "biosafety", "chirality", "moratorium", "governance", "catastrophic risk"]
summary: "Hypothetical organisms built from mirror-image biological molecules, which a large group of scientists argued in 2024 could evade immunity and predation and should not be created."
updated: "2026-07-27"
humanEvidence: "No mirror organism exists, so no person has ever been exposed to one; the nearest human exposure is to mirror-image molecules, such as the L-configured aptamers taken into clinical trials as Spiegelmers."
access: "Nothing to obtain: no mirror organism exists anywhere, and the peptide chemistry needed to build one is far beyond current capability."
reversibility: "irreversible"
issues: ["The 2025 European meetings on mirror-life governance are described without a citation."]
---

```infobox
{
  "caption": "Proposed biosafety hazard",
  "rows": [
    { "label": "Subject", "value": "Chirally inverted organisms" },
    { "label": "Building blocks", "value": "D-amino acids, L-nucleotides" },
    { "label": "Warning issued", "value": "December 2024" },
    { "label": "Signatories", "value": "38 scientists" },
    { "label": "Nearest existing work", "value": "Mirror-image enzymes and polymerases" },
    { "label": "Estimated distance", "value": "Decades" },
    { "label": "Proposed response", "value": "Research moratorium" }
  ]
}
```

**Mirror life** refers to hypothetical organisms constructed from the mirror images of ordinary biological molecules: proteins built from D-amino acids rather than L, and nucleic acids built from L-sugars rather than D. No such organism exists, and none is close to existing. In December 2024 a group of thirty-eight scientists, drawn from synthetic biology, immunology, ecology and biosecurity, published a warning in *Science* arguing that mirror bacteria would be uniquely dangerous and that work toward creating them should not proceed.[^adamala2024] The argument turns on a structural fact about biology rather than on any specific pathogen.

```keyfacts
[
  { "value": "38", "label": "Authors of the 2024 warning", "note": "including Nobel laureates and synthetic-biology researchers" },
  { "value": "~300 pp", "label": "Accompanying technical report", "note": "a detailed feasibility and risk assessment published alongside it" },
  { "value": "Decades", "label": "Estimated distance to a mirror bacterium", "note": "the authors judge it unlikely within ten years" }
]
```

## Chirality in biology

Most biological molecules are chiral: they exist in two forms that are mirror images and cannot be superimposed, like left and right hands. Life on Earth uses one hand almost exclusively. Proteins are built from L-amino acids; DNA and RNA use D-ribose and D-deoxyribose. The choice appears to be a frozen accident, since the chemistry works identically either way, and a complete mirror organism would in principle be viable.

This uniformity is not a curiosity. It is the basis of nearly every molecular recognition event in biology. Enzymes are chiral surfaces that fit chiral substrates. Antibodies bind shapes. Receptors, transporters and proteases all discriminate by handedness. A protein made of D-amino acids folds into the mirror image of the normal structure and is, to almost every biological molecule that would ordinarily interact with it, the wrong shape.

## The mechanism of harm

The danger proposed for mirror bacteria is not toxicity but invisibility. An organism whose surface molecules are chirally inverted would sidestep most of the mechanisms that normally limit bacterial growth.

**Adaptive immunity.** Antibody and T-cell recognition depends on shape complementarity, and antigen presentation depends on chirally specific proteases chopping proteins into peptides. Mirror proteins would be poorly processed and poorly recognised, so the adaptive response that eventually clears most infections might never be mounted.

**Innate immunity.** Pattern-recognition receptors detect conserved microbial features, some of which are chiral. The 2024 assessment concludes that innate detection would be degraded rather than abolished, and that phagocytes might engulf mirror bacteria but could not digest them, since lysosomal enzymes are chirally specific.

**Predation and phages.** Bacteriophages recognise host surface receptors and inject nucleic acid that host machinery must read. None of that works across a chirality boundary. Protists that graze bacteria in soil and water would face the same digestion problem as macrophages. In ordinary ecology, bacterial populations are held down as much by phages and grazers as by host immunity, and a mirror organism would escape all of it at once.

**Antibiotics.** Most antibiotics are chiral molecules binding chiral targets, so existing drugs would be expected to fail. Mirror-image versions of some antibiotics could be synthesised, but not quickly and not in the quantities an outbreak would demand.

The scenario the report treats as most serious is not a designed weapon but an environmental release of a mirror organism able to grow on widely available nutrients, spreading through soil, water and multiple host species with no natural control mechanism and no available countermeasure.

> [!key] Why this differs from an engineered pathogen
> A conventional engineered pathogen is dangerous because of what it does. A mirror bacterium would be dangerous because of what cannot be done to it. Ordinary biosecurity assumes an ecosystem of immune systems, predators and drugs pushing back; the mirror case removes that assumption for every host and every environment simultaneously.

## Plausibility

The warning rests on a chain of assumptions, and the authors are explicit that several are uncertain.

The strongest counterargument is nutritional. Mirror bacteria would need mirror-image nutrients, and the environment is stocked with ordinary-handed sugars, amino acids and lipids. A mirror organism dropped into soil would find much of the available food chemically unusable. The report's response is that an autotroph fixing carbon dioxide and nitrogen, or an organism carrying racemases able to interconvert chiral forms, would not face that limitation, and that a laboratory mirror organism would probably be built to grow on defined achiral media in the first place. Whether such an organism could compete in a real ecosystem is genuinely unknown, and critics regard this as the weakest link in the argument.

A second uncertainty concerns immune evasion itself. Innate defences include mechanisms with limited stereospecificity, and mucosal barriers, temperature, pH and iron restriction are not chirality-dependent. Some immunologists consider the claim of near-total evasion stronger than the evidence supports, though few dispute that adaptive immunity would be substantially blunted.

Feasibility is the least disputed point: it is remote. Building a mirror bacterium requires a mirror ribosome, and the ribosome is a complex of several RNAs and dozens of proteins. Synthesising it in mirror form by chemistry alone, or bootstrapping a mirror translation system that can make its own components, is far beyond current capability. The technical report judges that mirror bacteria are unlikely to be achievable within a decade and plausibly require several.

## What has actually been built

Mirror molecules, as distinct from mirror organisms, are an established and useful field.

- Total chemical synthesis produced a D-amino-acid version of HIV protease in 1992, which cleaved mirror-image substrates with the reciprocal specificity theory predicted.[^milton1992]
- Mirror-image DNA polymerases assembled from D-amino acids were shown to replicate L-DNA, establishing that a mirror genetic system can copy itself in vitro.[^wang2016]
- A mirror-image T7 RNA polymerase extended this to transcription of chirally inverted RNAs, including ribosomal RNA sequences.[^xu2022]

The therapeutic applications are unaffected by any proposed moratorium and are among the arguments for keeping the restriction narrowly drawn. D-peptide drugs resist degradation by human proteases, which is a genuine advantage in [[targeted-drug-delivery]], and L-configured aptamers, marketed as Spiegelmers, have been taken into clinical trials for the same reason. Mirror-image phage display is a working drug-discovery method, and mirror nucleic acids have been proposed as unusually stable scaffolds for the structures built in [[dna-nanotechnology]]. Confining a prohibition to self-replicating mirror organisms preserves all of this, in the same way that restrictions on [[gene-drive]] release do not restrict the underlying editing chemistry.

## Mitigation and governance

Mirror life presents an unusual governance profile. Most biosecurity controls operate on nucleic acid synthesis, screening ordered DNA sequences against hazard databases; that machinery, built for the era of [[synthetic-genomes]], is largely irrelevant here, because the hard step is peptide chemistry rather than gene synthesis. Nor does synthetic auxotrophy, the containment strategy used to confine [[recoded-organisms]], transfer cleanly: it depends on knowing which nutrient an organism cannot obtain, and a mirror organism's requirements are uncharacterised. The Biological Weapons Convention addresses hostile intent, not catastrophic accident in basic research. Nothing in existing law forbids the work, and the national patchwork documented in [[governance-of-genome-editing]] offers no obvious template, since it regulates edits to existing genomes rather than the construction of new chemistry.

The proposals that followed therefore reach for other levers: funder policies refusing to support the construction of mirror organisms, institutional biosafety review, journal norms, and voluntary commitment by the small community with the relevant expertise. Scientific meetings convened in Europe during 2025 examined what such a regime would need in order to be more than a statement of intent.

The obvious precedent is the [[asilomar-conference]], where researchers paused their own field and wrote containment rules. The analogy is imperfect in the usual way: recombinant DNA in 1975 was practised by a few dozen laboratories with shared professional norms, and the same concentration holds here only for as long as the technical barrier does. This is precisely the sequencing problem addressed by [[differential-technological-development]], and it connects mirror life to the broader analysis in [[dual-use-research]] and [[existential-risk]].

> [!debate] Whether to warn at all
> Some researchers argue that publicising a hazard nobody was pursuing creates interest in it, and that the risk assessment is speculative enough that a moratorium chills legitimate chemistry. The authors' reply is that the technical barriers give an unusually long lead time, and that a governance regime is far easier to build before anyone has invested a career in the work than after.

## Outlook

The immediate test of the 2024 initiative is institutional rather than scientific: whether funding agencies and biosafety committees adopt an explicit position, and whether that position survives contact with a research group that wants to build a mirror cell for reasons it finds compelling. The technical trajectory that matters is not synthetic biology in general but the automation of long-peptide synthesis, since that is the step gating a mirror ribosome.

A harder question sits behind the specific case. Mirror life is one of a small number of hazards where the risk comes from a general property of the artefact rather than from its purpose, alongside the self-propagating edits of gene drives and the scenarios once argued about under [[grey-goo]] and [[molecular-assembler|mechanosynthesis]]. That earlier episode is instructive: a speculative hazard debated in public before the technology existed distorted policy for a field that then developed along different lines. Whether the [[precautionary-principle]] can be applied here without repeating the pattern is what the insistence on a narrow, testable endpoint is meant to settle. Existing biosecurity is organised around intent and around pathogens that already exist. Nobody has shown that it can regulate a category defined by the absence of natural checks, and the mirror case is the first serious test of whether a scientific community can prohibit something it has not yet learned how to do.

## See also

- [[dual-use-research]]
- [[asilomar-conference]]
- [[gene-drive]]
- [[synthetic-genomes]]
- [[recoded-organisms]]
- [[existential-risk]]
- [[precautionary-principle]]
- [[differential-technological-development]]

## References

[^adamala2024]: `paper` Adamala, K. P. et al. "Confronting risks of mirror life." *Science*, 2024. Published with an accompanying book-length technical report on the feasibility and risks of mirror bacteria. {A policy argument reasoning from chirality and immunology; because no mirror organism exists, none of its claims about immune evasion has been tested.}
[^milton1992]: `paper` Milton, R. C. de L., Milton, S. C. F. and Kent, S. B. H. "Total chemical synthesis of a D-enzyme: the enantiomers of HIV-1 protease show reciprocal chiral substrate specificity." *Science*, 1992. {One chemically synthesised protein tested on synthetic substrates; it demonstrates mirror chemistry, not any step toward a mirror cell.}
[^wang2016]: `paper` Wang, Z. et al. "A synthetic molecular system capable of mirror-image genetic replication and transcription." *Nature Chemistry*, 2016. {The mirror polymerase copies L-DNA in a tube but was itself made by chemical synthesis; nothing in the system can build its own components.}
[^xu2022]: `paper` Xu, Y. and Zhu, T. F. "Mirror-image T7 transcription of chirally inverted ribosomal and functional RNAs." *Science*, 2022.

==============================================================================
ARTICLE: mitochondrial-dysfunction
TITLE:   Mitochondrial dysfunction in aging
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/mitochondrial-dysfunction
SOURCE:  https://futurehumanwiki.com/raw/mitochondrial-dysfunction
==============================================================================

---
title: "Mitochondrial dysfunction in aging"
slug: "mitochondrial-dysfunction"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity"]
tags: ["aging", "mitochondria", "oxidative stress", "mtdna", "metabolism", "mechanisms"]
summary: "Age-related decline in mitochondrial energy production, quality control and signalling, once explained by free-radical damage and now understood as a more tangled process."
updated: "2026-07-27"
humanEvidence: "Falling respiratory capacity and clonally expanded mtDNA deletions are documented in aged human muscle and brain tissue; antioxidant trials in people found no mortality benefit and signals of harm."
issues: ["The 2005 mitochondria-targeted catalase result in the timeline carries no citation."]
---

```infobox
{
  "caption": "Hallmark of aging",
  "rows": [
    { "label": "Tier", "value": "Antagonistic hallmark" },
    { "label": "Genome size", "value": "~16.5 kb, 37 genes" },
    { "label": "Proteins encoded", "value": "13 OXPHOS subunits" },
    { "label": "Inheritance", "value": "Maternal" },
    { "label": "Classic theory", "value": "Free-radical theory, 1956" },
    { "label": "Theory status", "value": "Largely abandoned in its strong form" },
    { "label": "Quality control", "value": "Mitophagy, fission and fusion" }
  ]
}
```

**Mitochondrial dysfunction in aging** refers to the progressive loss of mitochondrial respiratory capacity, genome integrity, and quality control that accompanies age in most tissues studied. It is classed as an antagonistic hallmark in the [[hallmarks-of-aging]] framework, meaning the underlying responses are protective at low intensity and damaging when chronic. The mechanism was long assumed to be self-amplifying oxidative damage; that explanation has not survived the experiments designed to test it.

## What declines

Mitochondria generate most cellular ATP through oxidative phosphorylation, buffer calcium, synthesize iron–sulfur clusters and steroid hormones, and control the intrinsic apoptosis pathway. Each carries multiple copies of a circular genome of roughly 16.5 kilobases encoding 37 genes, of which 13 are protein subunits of the respiratory chain; the remaining thousand-odd mitochondrial proteins are encoded in the nucleus and imported.

With age, measured respiratory capacity per unit of mitochondrial mass falls in muscle, liver and brain; mitochondrial morphology becomes more heterogeneous; membrane potential declines in a subset of organelles; and damaged mitochondria are cleared more slowly. In aged human skeletal muscle, individual fibres appear that lack cytochrome c oxidase activity, each traceable to a clonally expanded mitochondrial DNA deletion. Comparable clonal expansions of mtDNA deletions accumulate in substantia nigra neurons.[^bender2006] These are focal, cell-by-cell failures rather than a uniform organism-wide dimming.

## The free-radical theory and its decline

Denham Harman proposed in 1956 that aging results from cumulative damage by free radicals produced during normal metabolism, and in 1972 localized the main source to mitochondria.[^harman1956] The theory was attractive because it linked metabolic rate, damage, and lifespan in one causal chain, and it dominated the field for four decades. It also generated a clear prediction: reducing oxidative damage should extend lifespan.

```timeline
[
  { "year": "1956", "title": "Free-radical theory proposed", "text": "Denham Harman argues that oxygen radicals produced by metabolism damage cells cumulatively and cause aging." },
  { "year": "2004–2005", "title": "Mutator mice", "text": "Mice carrying a proofreading-deficient mitochondrial polymerase accumulate mtDNA mutations and show premature aging phenotypes, at mutation loads far above those seen in normal aging." },
  { "year": "2005", "title": "One clean positive", "text": "Targeting catalase to mitochondria extends median mouse lifespan by roughly a fifth, the strongest single result supporting the oxidative theory." },
  { "year": "2007–2011", "title": "Antioxidant trials fail", "text": "Meta-analyses of antioxidant supplement trials find no mortality benefit, and vitamin E is associated with increased prostate cancer risk in a large prevention trial." },
  { "year": "2009", "title": "Genetic tests come back negative", "text": "A review of mouse models altering antioxidant enzymes finds that most changes to oxidative defences do not alter lifespan." },
  { "year": "2013", "title": "Mutation signature reinterpreted", "text": "Sequencing of aged human brain shows mtDNA mutations carry a replication-error signature rather than an oxidative-damage signature." }
]
```

Nearly every direct test failed. Overexpressing or deleting superoxide dismutases and glutathione peroxidases in mice generally changed oxidative damage markers without changing lifespan.[^perez2009] Large randomized trials and meta-analyses of [[dietary-supplements|antioxidant supplements]] found no reduction in all-cause mortality, with signals of harm for some agents.[^bjelakovic2007] Long-lived species do not consistently show lower radical production; naked mole-rats carry high levels of oxidative damage while living an order of magnitude longer than mice of similar size, a comparison that features in the literature on [[negligible-senescence|negligible senescence]].

The interpretation that replaced it is mitohormesis: modest reactive oxygen species act as signals that trigger adaptive responses, and blunting them removes a stimulus the cell needs. In *Caenorhabditis elegans*, increasing mitochondrial superoxide can extend lifespan, and antioxidant treatment abolishes the effect. The same logic explains why antioxidant supplementation can blunt some of the adaptations produced by [[exercise-and-aging|exercise training]].

> [!key] Why the reversal matters
> The failure of antioxidants is the clearest case in biogerontology of a mechanistically appealing theory that did not survive intervention studies. It is the standard example cited against reasoning from plausible mechanism to expected benefit, and it is why the field now insists on lifespan and function endpoints rather than damage markers.

## mtDNA mutation and heteroplasmy

Because each cell carries many mitochondrial genomes, mutations exist as a mixture with wild-type copies, a condition called heteroplasmy. A biochemical defect usually appears only when mutant load in a cell exceeds a threshold, commonly placed between roughly 60 and 90 percent depending on the mutation and tissue. Aging tissue shows a rise not in average mutation burden across all copies, which stays low, but in the number of individual cells that have crossed the threshold through clonal expansion of a single mutant lineage.

The origin of those mutations was long assumed to be oxidative lesions. Deep sequencing of aged human brain instead found a mutational spectrum dominated by transitions consistent with errors made by the mitochondrial polymerase during replication.[^kennedy2013] Mice engineered with a proofreading-deficient version of that polymerase do age prematurely, but they carry mutation loads far above anything seen in normal aging, so the model demonstrates that enough mtDNA damage causes an aging-like phenotype rather than that normal aging is caused by mtDNA damage.

## Quality control

Damaged mitochondria are removed by mitophagy, a selective branch of [[autophagy]]. The best-characterized route depends on PINK1, which accumulates on depolarized mitochondria and recruits the ubiquitin ligase Parkin; receptor-mediated routes using BNIP3, NIX and FUNDC1 operate in parallel and dominate in some tissues. Both PINK1 and Parkin are mutated in familial early-onset Parkinson's disease, which is the strongest human evidence that failing mitochondrial quality control causes degeneration in a specific cell population.

Mitophagy is coupled to mitochondrial dynamics: fission separates damaged segments for disposal, fusion allows complementation between genomes. Markers of mitophagic flux decline with age in several model organisms, and "disabled macroautophagy" was added as a separate hallmark in 2023 in part because of this. Whether mitophagy declines with age in human tissue is harder to establish, since flux cannot be measured directly in a living person — the same measurement barrier that limits [[aging-biomarkers|biomarker]] development generally.

## Interventions

Exercise remains the most reliable stimulus of mitochondrial biogenesis and the only intervention with strong human functional evidence. Among drugs and supplements, urolithin A, a gut metabolite of ellagitannins, induces mitophagy in preclinical models and has been tested in humans; randomized trials in middle-aged and in older adults have reported modest improvements in muscle endurance measures without clear gains in aerobic capacity.[^singh2022][^liu2022] [[nad-precursors|NAD⁺ precursors]] are promoted partly on mitochondrial grounds, and raise blood NAD⁺ reliably while producing little demonstrated functional benefit. Elamipretide, a peptide that binds cardiolipin in the inner membrane, did not meet its primary endpoint in a phase 3 trial in primary mitochondrial myopathy, though it has continued in narrower indications.

[[caloric-restriction|Dietary restriction]] and [[rapamycin]] both alter mitochondrial turnover, and part of their effect is attributed to enhanced quality control rather than to reduced damage. Mitochondrial-derived peptides such as humanin and MOTS-c have been proposed as signalling mediators between mitochondrial state and organismal metabolism; the human data are observational.

## Transfer, replacement, and editing

Three approaches attempt to change mitochondria rather than support them. Autologous mitochondrial transplantation — injecting mitochondria isolated from a patient's own muscle into ischaemic heart tissue — has been performed in small paediatric case series at a single centre, with encouraging but uncontrolled results. [[mitochondrial-replacement-therapy|Mitochondrial replacement]] in the germline avoids transmitting pathogenic mtDNA by transferring nuclear material into a donor egg, and is legal in a small number of jurisdictions. Editing mtDNA directly is complicated by the fact that guide RNAs are not imported into mitochondria, which rules out standard CRISPR approaches; base editors built from a bacterial toxin that acts on double-stranded DNA have been used to install specific mtDNA edits in cells and mice, and are the leading route to a [[gene-therapy-for-aging|genetic intervention]] on the mitochondrial genome.

Whether any of this addresses aging as opposed to specific mitochondrial diseases is unresolved. The clonal expansion that produces cytochrome-oxidase-negative fibres affects a minority of cells even in old muscle, and no one has shown that correcting it restores organ-level function in an aged mammal. That gap — between a well-documented molecular lesion and a demonstrated contribution to functional decline — is the standing problem for the hallmark, and it applies with equal force to [[cellular-senescence|senescence]] and [[proteostasis|proteostasis]] as causal accounts.

## See also

- [[hallmarks-of-aging]]
- [[autophagy]]
- [[nad-precursors]]
- [[mitochondrial-replacement-therapy]]
- [[inflammaging]]
- [[exercise-and-aging]]

## References

[^harman1956]: `paper` Harman, D. "Aging: A theory based on free radical and radiation chemistry." *Journal of Gerontology*, 1956.
[^bender2006]: `paper` Bender, A. et al. "High levels of mitochondrial DNA deletions in substantia nigra neurons in aging and Parkinson disease." *Nature Genetics*, 2006.
[^perez2009]: `paper` Pérez, V. I. et al. "Is the oxidative stress theory of aging dead?" *Biochimica et Biophysica Acta*, 2009. {A review of existing mouse models that altered antioxidant defences, assembling negative lifespan results rather than reporting a new experiment.}
[^bjelakovic2007]: `paper` Bjelakovic, G. et al. "Mortality in randomized trials of antioxidant supplements for primary and secondary prevention." *JAMA*, 2007. {Pools trials of supplements in people; the endpoint is all-cause mortality, and the trials were not designed to test any theory of aging.}
[^kennedy2013]: `paper` Kennedy, S. R., Salk, J. J., Schmitt, M. W., Loeb, L. A. "Ultra-sensitive sequencing reveals an age-related increase in somatic mitochondrial mutations that are inconsistent with oxidative damage." *PLoS Genetics*, 2013. {Post-mortem human brain tissue; the argument rests on the mutational signature, which points to polymerase error rather than oxidative lesions.}
[^singh2022]: `paper` Singh, A. et al. "Urolithin A improves muscle strength, exercise performance, and biomarkers of mitochondrial health in a randomized trial in middle-aged adults." *Cell Reports Medicine*, 2022.
[^liu2022]: `paper` Liu, S. et al. "Effect of urolithin A supplementation on muscle endurance and mitochondrial health in older adults: a randomized clinical trial." *JAMA Network Open*, 2022.

==============================================================================
ARTICLE: mitochondrial-replacement-therapy
TITLE:   Mitochondrial replacement therapy
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/mitochondrial-replacement-therapy
SOURCE:  https://futurehumanwiki.com/raw/mitochondrial-replacement-therapy
==============================================================================

---
title: "Mitochondrial replacement therapy"
slug: "mitochondrial-replacement-therapy"
type: "intervention"
status: "emerging"
horizon: "present"
trl: 8
categories: ["reproduction", "genetics"]
tags: ["reproduction", "mitochondria", "ivf", "germline", "regulation", "heteroplasmy"]
summary: "A set of IVF techniques that place a mother's nuclear DNA in a donor egg to prevent transmission of mitochondrial DNA disease, licensed in a few countries and used rarely."
updated: "2026-07-27"
humanEvidence: "Children have been born after the procedure in several countries; the largest reported series is eight UK births, with the mother's pathogenic mtDNA undetectable or below disease thresholds and follow-up still short."
access: "Licensed case by case in the UK and permitted on a staged pathway in Australia, blocked in the US by an appropriations rider, and sold for infertility by clinics in countries with no specific rules."
reversibility: "irreversible"
issues: ["The claim that infertility use accounts for much of the worldwide birth total needs a source."]
---

```infobox
{
  "caption": "Assisted reproduction technique",
  "rows": [
    { "label": "Also called", "value": "Mitochondrial donation" },
    { "label": "Main methods", "value": "Spindle transfer, pronuclear transfer" },
    { "label": "Target", "value": "mtDNA disease" },
    { "label": "First legalised", "value": "United Kingdom, 2015" },
    { "label": "First reported birth", "value": "2016, Mexico" },
    { "label": "Heritable", "value": "Yes, through the female line" },
    { "label": "US status", "value": "Blocked by appropriations rider" }
  ]
}
```

**Mitochondrial replacement therapy** is a group of in vitro fertilisation techniques that combine the nuclear DNA of a prospective mother with the cytoplasm, and therefore the mitochondria, of a donor egg. The purpose is to prevent a child from inheriting a pathogenic mutation in mitochondrial DNA. It is the only form of heritable human genetic modification that any country has legalised, and the number of children born from it worldwide remains small.

## The disease it addresses

Mitochondria carry their own small circular genome of about 16,500 base pairs encoding thirteen proteins of the respiratory chain plus the RNAs needed to make them. It is inherited almost exclusively from the egg. Mutations in it cause a heterogeneous group of disorders, among them Leigh syndrome, MELAS, MERRF and Leber's hereditary optic neuropathy. These typically strike tissues with high energy demand: brain, heart, skeletal muscle, retina.

Two features complicate inheritance. Cells carry many mitochondrial genomes, so a person can be *heteroplasmic*, carrying a mixture of normal and mutant copies. And most mitochondrial diseases show a threshold effect: symptoms appear only above a high mutant fraction, often well over half. During oogenesis a genetic bottleneck randomly samples a small number of mitochondrial genomes to seed each egg, so a heteroplasmic mother's eggs vary widely and unpredictably in mutant load. She may have several children with very different outcomes.

The bottleneck also explains why replacement is not always the right answer. Because eggs differ so much, [[embryo-selection]] for low heteroplasmy is often sufficient and is technically far simpler. Replacement is aimed principally at women who are homoplasmic for a pathogenic variant, where every egg carries it, and at those whose embryos all exceed the threshold.

## How it works

```compare
{
  "columns": ["Maternal spindle transfer", "Pronuclear transfer"],
  "rows": [
    { "label": "Performed on", "values": ["Unfertilised egg", "Fertilised zygote"] },
    { "label": "What is moved", "values": ["Meiotic spindle with chromosomes", "Both pronuclei"] },
    { "label": "Embryos destroyed", "values": ["None; donor egg is enucleated", "Donor zygote is enucleated"] },
    { "label": "Timing", "values": ["Before fertilisation", "Within hours of fertilisation"] },
    { "label": "Used in first UK births", "values": ["No", "Yes"] }
  ]
}
```

In *maternal spindle transfer*, the chromosome-bearing spindle is removed from the patient's unfertilised egg with a micropipette and inserted into a donor egg whose own spindle has been removed. The reconstructed egg is then fertilised. In *pronuclear transfer*, both eggs are fertilised first and the two pronuclei are moved from the patient's zygote into an enucleated donor zygote. A third approach, polar body transfer, uses the chromosomes discarded during meiosis and carries less cytoplasm, but has been used less.

All of these are variants of the micromanipulation developed for somatic cell nuclear transfer, the technique behind [[human-cloning]], and they inherit its practical difficulties: the spindle is invisible without polarised optics, and the manipulation must not activate the egg prematurely. Some cytoplasm inevitably travels with the nuclear material, which is the source of the technique's central problem.

## Development history

```timeline
[
  { "year": "1997–2001", "title": "Ooplasmic transfer", "text": "A New Jersey clinic injects donor cytoplasm into eggs of women with repeated IVF failure. Around seventeen children are born before the FDA requires an investigational new drug application, halting the practice." },
  { "year": "2009", "title": "Spindle transfer in monkeys", "text": "Shoukhrat Mitalipov's group reports healthy macaques born after maternal spindle transfer, the first primate demonstration." },
  { "year": "2010", "title": "Pronuclear transfer in human embryos", "text": "A Newcastle team shows the technique works in abnormally fertilised human zygotes with low mitochondrial carryover." },
  { "year": "2015", "title": "The UK legalises it", "text": "Parliament approves regulations permitting mitochondrial donation under licence, the first statutory authorisation of heritable modification anywhere." },
  { "year": "2016", "title": "First reported birth", "text": "A US clinician performs spindle transfer in Mexico for a couple with Leigh syndrome; the FDA later issues a warning letter." },
  { "year": "2022", "title": "Australia follows", "text": "Maeve's Law permits mitochondrial donation in a staged research and clinical pathway." },
  { "year": "2025", "title": "The UK cohort reports", "text": "The Newcastle programme publishes outcomes for a small number of children born after pronuclear transfer, with low but not always undetectable mitochondrial carryover." }
]
```

The scientific groundwork was laid in two places. Shoukhrat Mitalipov's laboratory in Oregon showed that spindle transfer produced healthy macaques,[^tachibana2009] and a Newcastle group led by Doug Turnbull and Mary Herbert established that pronuclear transfer worked in human zygotes with low carryover of the original mitochondria.[^craven2010]

The United Kingdom then took the deliberate route. After several public consultations and reviews of safety by the fertility regulator, Parliament approved regulations in 2015 under the existing Human Fertilisation and Embryology Act; the Newcastle Fertility Centre received the first licence in 2017 and the first patient approvals followed. In 2025 the Newcastle group published outcomes for the children born under the programme: eight births, with the mother's pathogenic mitochondrial DNA undetectable in most of the children and present at low levels in the rest, in every case below the level associated with disease.[^newcastle2025] The same report described a larger number of births achieved by preimplantation testing rather than by donation, consistent with the pathway's design, in which selection among embryos is attempted first and donation reserved for the women it cannot help. None of the children had developed a mitochondrial disease phenotype at the time of reporting, though the follow-up was short and, as with every heritable intervention, the informative period runs into the next generation.

The unregulated route ran in parallel. The 2016 Mexico birth was performed by a US-based clinician outside American jurisdiction and reported afterwards.[^zhang2017] Clinics in Ukraine and Greece subsequently used nuclear transfer techniques not for mitochondrial disease but for infertility, on the hypothesis that donor cytoplasm improves poor-quality eggs. That indication is not supported by controlled evidence, and it accounts for a substantial share of the children born from the technique worldwide. The pattern, of a procedure debated for a decade in one country and sold commercially in another, is the same jurisdictional arbitrage that later framed the reaction to the [[he-jiankui-affair]], with the important difference that mitochondrial donation had a legitimate clinical rationale and a legal home.

## The reversion problem

Carryover of a few percent of the mother's mitochondria is unavoidable. If those genomes simply persisted at that level, the technique would be safe. They do not always.

Embryonic stem cell lines derived from replacement embryos showed that in a fraction of lines the carried-over genotype expanded until it dominated, sometimes completely.[^kang2016] The mechanism is not settled; proposed explanations include replicative advantage conferred by particular sequence variants in the mitochondrial control region, and interaction between the donor mitochondrial haplotype and the patient's nuclear genome. One response has been to match donor and recipient mitochondrial haplogroups, which is intuitively appealing and not clearly supported.

The nuclear-mitochondrial interaction question is broader than reversion. Mouse experiments that placed one nuclear genome against a mismatched mitochondrial background found effects on metabolism, ageing markers and lifespan, suggesting the two genomes are co-adapted in ways that a transfer disrupts.[^latorre2016] Whether this matters at the magnitude of variation seen between human haplogroups is unresolved, and it is one of the main scientific reasons for continued caution about the technique.

> [!caution] Why "three-parent baby" misleads in both directions
> The donor contributes 37 genes out of roughly 20,000, none of them influencing the traits people associate with parenthood, so the phrase overstates the biological relationship. But it understates the legal point: the modification is heritable, since a girl born after the procedure passes the donor mitochondria to her own children. That is precisely why it required primary legislation.

## Regulation

The United Kingdom and Australia have explicit statutory frameworks with case-by-case licensing. Most other countries have neither authorised nor specifically prohibited the technique, leaving it to general rules on embryo research.

The United States has a distinctive blockage. Since late 2015, an annual appropriations rider has barred the Food and Drug Administration from acknowledging any application for a clinical investigation in which a human embryo is intentionally modified to include a heritable genetic modification. Because the FDA asserts jurisdiction over the technique, the rider functions as a prohibition without Congress ever having debated mitochondrial donation on its merits. A National Academies committee had earlier recommended a cautious path forward, restricted initially to male embryos so that any error would not be transmitted; the rider superseded the recommendation.

The wider pattern is the one described in [[governance-of-genome-editing]]: a technology permitted under supervision in a few jurisdictions, prohibited by side-effect in another, and available commercially in several with little oversight. Mitochondrial donation is the clearest existing test of whether the line drawn around heritable modification in [[germline-editing]] debates is a line about mechanism or about consequence. It crosses the germline while touching no nuclear gene, which is why some commentators treat it as an exception and others as a precedent.

Ethically it has been comparatively uncontroversial, which is itself informative. It prevents a severe, often fatal childhood disease; it is not an enhancement in any sense recognised by [[bioethics-of-enhancement]]; and it introduces no allele that a healthy person does not already have. The objections that were raised concerned donor identity and the child's interest in knowing about the procedure, questions familiar from gamete donation, together with the disability-rights argument set out in [[disability-rights-and-enhancement]] that preventing the birth of children with a condition carries an implicit judgement about people living with it. Because eligible families are rare and the procedure requires a specialist licensed centre, the [[access-and-inequality]] questions that dominate other reproductive technologies have played a smaller part.

## Alternatives and outlook

Two technical alternatives could make the transfer approach obsolete. Editing mitochondrial DNA directly has become possible with cytosine base editors derived from a bacterial toxin that acts on double-stranded DNA, since conventional CRISPR guide RNAs cannot be imported into mitochondria; the chemistry is described under [[base-editing]] and has been demonstrated in cells and in mice, not in a clinical setting. Such editors could in principle shift heteroplasmy below the disease threshold without any donor egg. Separately, [[in-vitro-gametogenesis]] from [[induced-pluripotent-stem-cells|reprogrammed somatic cells]] would permit gametes to be generated and screened in quantity, though it would not by itself remove a homoplasmic mutation.

For the disease itself, the fundamental question is unchanged by any of this. Mitochondrial dysfunction is also a feature of ordinary ageing, one of the [[hallmarks-of-aging]] and the subject of [[mitochondrial-dysfunction]], and no intervention yet restores function to a tissue that has already lost it. Replacement therapy prevents transmission. It does nothing for the patients who prompted its development.

## See also

- [[embryo-selection]]
- [[germline-editing]]
- [[in-vitro-gametogenesis]]
- [[mitochondrial-dysfunction]]
- [[governance-of-genome-editing]]
- [[human-cloning]]
- [[designer-babies]]
- [[reproductive-longevity]]

## References

[^zhang2017]: `paper` Zhang, J. et al. "Live birth derived from oocyte spindle transfer to prevent mitochondrial disease." *Reproductive BioMedicine Online*, 2017. {A single case, reported after the fact by the clinician who performed it in a jurisdiction with no rules governing the procedure.}
[^kang2016]: `paper` Kang, E. et al. "Mitochondrial replacement in human oocytes carrying pathogenic mitochondrial DNA mutations." *Nature*, 2016. {The reversion appeared in embryonic stem cell lines derived from reconstructed embryos, which is where it has been observed rather than in any child.}
[^latorre2016]: `paper` Latorre-Pellicer, A. et al. "Mitochondrial and nuclear DNA matching shapes metabolism and healthy ageing." *Nature*, 2016.
[^newcastle2025]: `paper` Hyslop, L. A. et al. "Mitochondrial Donation and Preimplantation Genetic Testing for mtDNA Disease." *New England Journal of Medicine*, 2025. Reporting for the Newcastle Fertility Centre and the Wellcome Centre for Mitochondrial Research. {Eight children from one licensed programme, with follow-up too short to speak to later-onset disease and nothing yet about the next generation.}
[^craven2010]: `paper` Craven, L. et al. "Pronuclear transfer in human embryos to prevent transmission of mitochondrial DNA disease." *Nature*, 2010.
[^tachibana2009]: `paper` Tachibana, M. et al. "Mitochondrial gene replacement in primate offspring and embryonic stem cells." *Nature*, 2009.

==============================================================================
ARTICLE: molecular-assembler
TITLE:   Molecular assembler
PORTAL:  Nanomedicine
URL:     https://futurehumanwiki.com/wiki/molecular-assembler
SOURCE:  https://futurehumanwiki.com/raw/molecular-assembler
==============================================================================

---
title: "Molecular assembler"
slug: "molecular-assembler"
type: "concept"
status: "contested"
horizon: "indefinite"
categories: ["nanomedicine"]
tags: ["nanotechnology", "mechanosynthesis", "manufacturing", "molecular machines", "drexler", "chemistry"]
summary: "A proposed machine that would build objects by positioning individual atoms, central to Drexler's nanotechnology programme and disputed by chemists since the 1990s."
updated: "2026-07-27"
issues: ["Claims about current hydrogen depassivation lithography work carry no citation."]
---

```infobox
{
  "caption": "Proposed manufacturing technology",
  "rows": [
    { "label": "Proposed by", "value": "Eric Drexler", "link": "/wiki/eric-drexler" },
    { "label": "Key texts", "value": "Engines of Creation (1986), Nanosystems (1992)" },
    { "label": "Core mechanism", "value": "Positional mechanosynthesis" },
    { "label": "Principal critic", "value": "Richard Smalley" },
    { "label": "Key exchange", "value": "Drexler–Smalley debate, 2003" },
    { "label": "Devices built", "value": "None" },
    { "label": "Natural analogue", "value": "The ribosome" },
    { "label": "Status", "value": "Theoretically disputed" }
  ]
}
```

**A molecular assembler** is a hypothetical machine that would build structures by holding and positioning individual reactive molecules, bonding them to a workpiece one at a time under programmed control. The idea is the foundation of [[eric-drexler]]'s nanotechnology programme and, by extension, of the entire design literature on [[medical-nanorobots]]. No assembler has been built, and whether the central chemical step is achievable at all has been argued for three decades without an experimental resolution.

## Overview

Ordinary chemistry mixes reagents in solution and lets thermal motion bring them together; which products form is a matter of statistics and energetics. Mechanosynthesis proposes the opposite: hold a reactive group on a stiff tool, bring it to a specific site on a workpiece, force the reaction, and withdraw. The output would be an atomically precise structure specified in advance rather than a statistical distribution of products.

Drexler argued in *Nanosystems* that machines to do this could be built from diamondoid components — stiff carbon lattices whose behaviour can be modelled with classical mechanics — and analysed their bearings, gears, drive trains, and error rates quantitatively.[^drexler1992] The book is a work of theoretical engineering. It contains no experiments, and it is careful to say so.

Two features of the proposal drive most of the controversy that followed. The first is generality: an assembler that can build atomically precise structures could in principle build another assembler, which is where [[grey-goo]] enters the argument. The second is throughput: a single machine placing atoms one at a time is far too slow to make anything macroscopic, so any practical system needs either extreme parallelism or self-replication.

Everything downstream in the nanomedicine design literature assumes the capability. Robert Freitas's device studies, of which the [[respirocytes|respirocyte]] is the best known, specify diamondoid components that only mechanosynthesis could produce; without an assembler they are analyses of objects that cannot be made.

## Origins

Richard Feynman's 1959 lecture "There's Plenty of Room at the Bottom" raised the possibility of building machines that build smaller machines, ending with atomic-scale manipulation. It was a provocation rather than a programme, and it produced no research line for two decades.

Drexler formalised it. *Engines of Creation* (1986) described assemblers as a general-purpose manufacturing technology and set out both the promise and the hazards. His MIT doctorate and the resulting *Nanosystems* supplied the technical case. The word "nanotechnology", which Norio Taniguchi had used earlier in a narrower sense, entered public use through this work — and then drifted away from it, so that by the time the US National Nanotechnology Initiative launched in 2000 the term denoted nanoscale materials science and had almost nothing to do with assemblers.

That semantic drift mattered. A large federal programme was now named after an idea it did not pursue, and its leaders had an incentive to distance the programme from the science-fiction imagery attached to the name. Outside the laboratories, the assembler had meanwhile become a fixture of futurist argument, supplying [[ray-kurzweil]] and others with a mechanism for the material abundance that features in accounts of the [[technological-singularity]], and supplying [[existential-risk]] taxonomies with one of their earliest technological entries.

## The Drexler–Smalley debate

Richard Smalley, who shared the 1996 Nobel Prize in Chemistry for the discovery of fullerenes, made the central technical objection in *Scientific American* in 2001.[^smalley2001] He raised two problems, which he called fat fingers and sticky fingers.

The fat-fingers problem: to control a chemical reaction you must control the position of every atom taking part, and there is not enough room in the reaction volume for manipulators to hold them all. The sticky-fingers problem: the atoms being placed will adhere to whatever holds them, and there is no general way to make them let go at the right moment.

Drexler's reply, developed across an exchange of letters published in *Chemical & Engineering News* in 2003, was that Smalley was describing a manipulator picking up individual bare atoms, which is not the proposal.[^cen2003] Mechanosynthesis as described in *Nanosystems* uses tools with specific reactive tips that transfer a group and then withdraw, the bond strengths chosen so the transfer is thermodynamically downhill. Drexler pointed to enzymes and to the ribosome as proof that positional chemistry works: the ribosome holds a growing peptide and adds residues in a specified order, with error rates low enough to build every protein in every organism.

Smalley's counter was that enzymes work in water, use the specific chemistry of a small set of elements, and are not general-purpose machine tools; extrapolating from a ribosome to a diamond-building factory assumes the hard part away. The exchange ended acrimoniously, with Smalley accusing Drexler of frightening the public, and without either side proposing an experiment that would settle the question.

> [!debate] What the debate did not resolve
> Neither party demonstrated a tooltip, attempted a mechanosynthetic reaction, or specified a decisive experiment. Chemists have generally treated Smalley's objections as sufficient reason not to pursue the idea; proponents treat the absence of an experimental refutation as leaving the question open. Both positions are compatible with the evidence, which is thin on the specific question of programmed covalent transfer under positional control.

A 2006 US National Research Council review of the National Nanotechnology Initiative addressed the dispute directly. It concluded that the feasibility of site-specific chemistry for large-scale manufacturing could not be determined from theory alone and recommended experimental work, which was never substantially funded.[^nrc2006]

## What was built instead

The forty years since *Engines of Creation* produced a great deal of nanoscale construction, almost none of it by the proposed route.

Scanning probe microscopy demonstrated positional control at the atomic scale early. Don Eigler and Erhard Schweizer arranged xenon atoms into letters on a nickel surface in 1989, and later work formed and broke individual chemical bonds with a scanning tunnelling microscope.[^eigler1990] These are real single-atom manipulations, performed at cryogenic temperatures and ultra-high vacuum, one atom at a time, on a surface. They are also the closest experimental approach to mechanosynthesis, and their throughput is many orders of magnitude short of manufacturing anything.

Solution chemistry went a different way. Supramolecular chemists built rotaxanes, catenanes, and light-driven rotary motors, work recognised by the 2016 Nobel Prize in Chemistry. David Leigh's group built an artificial machine that reads a molecular strand and assembles a peptide in the specified sequence — an explicit ribosome mimic, and the strongest existing demonstration that programmed molecular construction is possible.[^leigh2013] It operates in solution by self-assembly, not by mechanical positioning.

[[dna-nanotechnology]] delivered the most useful form of programmable construction: structures folded to specification from sequence, at nanometre resolution, in a test tube. The addressability is real, the material is soft, and the products are shapes rather than machines. Writing whole [[synthetic-genomes]] extended the same logic to the largest molecules anyone has assembled to specification, again by biological rather than mechanical means. [[ai-protein-design|Learned models of protein structure]] have taken it furthest: specifying a fold that does not occur in nature and obtaining a molecule that adopts it is now a practical exercise, and it is atomic precision reached through sequence rather than through positioning.

The technologies that turned nanoscale control into medicine — [[lipid-nanoparticles]], antibody-drug conjugates, the particles discussed in [[targeted-drug-delivery]] — use none of this. They are formulations whose behaviour comes from surface chemistry and biodistribution. The devices that move under external control, surveyed in [[microrobots-in-medicine]], are machined by conventional lithography and assembled at scales a thousand times coarser than the assembler debate concerns.

## Where the question stands

As of 2026 there is no assembler, no mechanosynthetic tooltip in the laboratory, and no funded programme attempting one at scale. A small literature continues to publish computational studies of proposed diamond-mechanosynthesis reactions, and a handful of groups pursue atomically precise fabrication on silicon surfaces using hydrogen depassivation lithography, which is atomically precise patterning rather than three-dimensional construction.

The strongest argument for the concept remains the ribosome: an existence proof that programmable, error-corrected, atom-by-atom construction of complex covalent structures is physically possible, because it happens continuously in every living cell. The strongest argument against is that biology achieves this with soft, water-based, self-assembling machinery of a kind quite unlike a diamondoid factory, and that no one has shown how to get from one to the other.

Drexler's own later work reframes the programme as atomically precise manufacturing and puts less weight on the general-purpose desktop assembler.[^drexler2013] That shift removes some of the imagery that made the idea famous, including the self-replicating machine, while keeping the underlying claim that precision at the atomic scale would change what can be manufactured.

## Outlook

The question that would move the field is narrow and experimental: can a designed tooltip transfer a carbon dimer to a specified site on a diamond surface and withdraw cleanly, repeatedly, under positional control? It has been proposed in computational form for two decades and never attempted seriously in a laboratory. Until someone tries, the debate over assemblers remains a disagreement about extrapolation rather than about data, and the [[technology-readiness-level]] of everything downstream — including every proposal in this wiki that assumes atomically precise machinery — stays pinned at the bottom of the scale.

## See also

- [[eric-drexler]]
- [[medical-nanorobots]]
- [[respirocytes]]
- [[grey-goo]]
- [[dna-nanotechnology]]
- [[technological-singularity]]
- [[targeted-drug-delivery]]
- [[precautionary-principle]]

## References

[^drexler1992]: `book` Drexler, K.E. *Nanosystems: Molecular Machinery, Manufacturing, and Computation.* Wiley, 1992.
[^smalley2001]: `news` Smalley, R.E. "Of Chemistry, Love and Nanobots." *Scientific American*, September 2001. {A short commentary piece by a Nobel chemist, not a research report; the fat-fingers and sticky-fingers objections are argument, not measurement.}
[^cen2003]: `news` Baum, R. "Nanotechnology: Drexler and Smalley make the case for and against 'molecular assemblers'." *Chemical & Engineering News*, 1 December 2003. {A published exchange of open letters between the two men, framed by an editor; neither side brought new experimental data to it.}
[^nrc2006]: `report` National Research Council. *A Matter of Size: Triennial Review of the National Nanotechnology Initiative.* National Academies Press, 2006.
[^eigler1990]: `paper` Eigler, D.M. and Schweizer, E.K. "Positioning single atoms with a scanning tunnelling microscope." *Nature*, 1990. {The xenon atoms were slid across a nickel surface at liquid-helium temperature in ultra-high vacuum; no covalent bond was made or broken.}
[^leigh2013]: `paper` Lewandowski, B. et al. "Sequence-specific peptide synthesis by an artificial small-molecule machine." *Science*, 2013. {The machine works in solution and is driven along its track by thermal motion, so it demonstrates programmed sequence rather than positional control.}
[^drexler2013]: `book` Drexler, K.E. *Radical Abundance: How a Revolution in Nanotechnology Will Change Civilization.* PublicAffairs, 2013.

==============================================================================
ARTICLE: moral-enhancement
TITLE:   Moral enhancement
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/moral-enhancement
SOURCE:  https://futurehumanwiki.com/raw/moral-enhancement
==============================================================================

---
title: "Moral enhancement"
slug: "moral-enhancement"
type: "concept"
status: "speculative"
horizon: "indefinite"
categories: ["enhancement", "society"]
tags: ["ethics", "enhancement", "psychopharmacology", "autonomy", "existential risk", "bioethics"]
summary: "The proposal to improve human moral motivation by biomedical means, argued for as a response to technologies that outpace evolved moral psychology."
updated: "2026-07-27"
humanEvidence: "No agent has been shown to change how a person actually treats anyone; every human result is a shift on a questionnaire or a laboratory task, measured in a single session in healthy volunteers."
---

```infobox
{
  "caption": "Proposal in applied ethics",
  "rows": [
    { "label": "Also called", "value": "Moral bioenhancement" },
    { "label": "Proposed", "value": "2008" },
    { "label": "Principal advocates", "value": "Persson and Savulescu", "link": "/wiki/julian-savulescu" },
    { "label": "Principal critic", "value": "John Harris" },
    { "label": "Candidate agents", "value": "Oxytocin, SSRIs, propranolol" },
    { "label": "Demonstrated effect", "value": "None on real moral behaviour" }
  ]
}
```

**Moral enhancement** is the proposal to improve human moral dispositions — altruism, impartiality, aversion to causing harm — by biomedical intervention rather than by education or institutional design. It is the branch of [[human-enhancement]] whose advocates present it not as an optional improvement but as a response to a mismatch: a moral psychology shaped for small groups now controls technologies capable of catastrophic harm. Its critics reply that the psychopharmacology on offer is far too crude for the job, and that the interventions proposed would damage something essential to moral agency even if they worked.

## The argument

Ingmar Persson and [[julian-savulescu]] set out the case in a 2008 paper and expanded it in *Unfit for the Future*.[^ps2008][^ps2012] The argument has four steps.

Human moral psychology evolved under conditions of small-group cooperation. It is biased toward kin and near neighbours, toward the immediate over the distant future, and toward acts over omissions. These biases were adaptive and are now liabilities.

Technology has made the harm a small number of people can cause vastly greater than the good a comparable number can do. Persson and Savulescu call the worst case ultimate harm: an outcome that permanently destroys the prospect of a worthwhile life on Earth. The relevant class of hazards is the one described in [[existential-risk]] and, for the biological cases, [[dual-use-research]].

Liberal democracies cannot solve this by surveillance and prohibition alone, because the number of people who must be prevented from acting is small and the means are increasingly distributed.

Therefore, they argue, cognitive enhancement without corresponding moral enhancement is dangerous, and research into biomedical moral enhancement is urgent. The cognitive enhancement they have in mind is the whole range covered by [[nootropics]], [[genetic-enhancement-of-intelligence]] and [[intelligence-amplification]], on the view that all of them increase capability without touching motivation. Thomas Douglas advanced a narrower and more defensible version in the same year: moral enhancement as the attenuation of clearly counter-moral emotions such as aversion to particular racial groups or impulses toward violent aggression, which he argued is permissible on almost any account of moral worth.[^douglas2008]

```timeline
[
  { "year": "2008", "title": "The proposal appears", "text": "Persson and Savulescu argue that cognitive enhancement without moral enhancement is dangerous; Thomas Douglas independently defends the attenuation of counter-moral emotions." },
  { "year": "2010–2012", "title": "Candidate agents identified", "text": "Studies report that serotonin manipulation shifts harm aversion in hypothetical dilemmas, that oxytocin increases in-group favouritism, and that propranolol reduces implicit racial bias." },
  { "year": "2011", "title": "The freedom objection", "text": "John Harris argues that moral worth requires the capacity to choose otherwise, and that engineering good behaviour removes what makes it good." },
  { "year": "2012", "title": "The God Machine", "text": "Savulescu and Persson answer Harris with a thought experiment: a system that intervenes only to prevent great harms, leaving ordinary choice untouched." },
  { "year": "2014–2016", "title": "Sustained criticism", "text": "Robert Sparrow attacks the political assumptions of the programme, and Harris Wiseman's book argues the neuroscience cannot support anything the proposal requires." }
]
```

## What would count

The literature distinguishes several targets, which are often conflated.

**Motivation.** Making people more disposed to act on moral judgements they already hold. This is the target Persson and Savulescu emphasise, on the view that akrasia rather than ignorance is the binding constraint.

**Emotional disposition.** Increasing empathy or fellow-feeling, or reducing aggression and out-group aversion. The candidate pharmacology is concentrated here.

**Judgement.** Improving the reasoning by which moral conclusions are reached. John Harris argues that this makes cognitive enhancement the only real moral enhancement, since better information and better inference improve moral decisions without touching motivation.

**Behavioural constraint.** Preventing specific acts. This is not enhancement of the agent at all, which is why the God Machine thought experiment — a system that intervenes only to block great harms — is presented as a limiting case rather than as a proposal.[^ps2012]

> [!note] The bootstrapping problem
> Any programme of moral enhancement must be designed by people whose moral views would then be propagated. If those views are mistaken, the intervention entrenches the error; if they are not, the argument for the intervention presupposes the moral knowledge it is meant to improve.

## The empirical basis

The candidate agents are borrowed from psychiatry and studied on laboratory tasks. None was developed for this purpose and none has been tested against real moral behaviour. Device-based routes are thinner still: reports of personality and impulse-control change after [[deep-brain-stimulation]] are the closest thing to a demonstration that a neural intervention can shift how a person behaves toward others, and they are side effects of treatment for movement disorders, not enhancements. The transcranial methods surveyed in [[non-invasive-neuromodulation]] have produced nothing durable.

**Serotonin.** A single dose of the SSRI citalopram increased participants' reluctance to endorse harming one person to benefit several in hypothetical dilemmas, an effect interpreted as heightened harm aversion.[^crockett2010] Whether this constitutes moral improvement is contested: greater reluctance to act is not obviously better ethics, and the same shift would make some defensible choices harder.

**Oxytocin.** Often described in popular accounts as a trust or bonding hormone, it behaves in experiments as a parochial one. De Dreu and colleagues found that intranasal oxytocin increased favouritism toward in-group members and, on some measures, defensive hostility toward out-groups.[^dedreu2011] For a proposal whose central aim is extending concern beyond the near group, this is the wrong direction. The wider oxytocin literature also has severe replication problems, and whether intranasal administration delivers meaningful quantities to the human brain remains disputed.

**Propranolol.** A beta-blocker reduced implicit racial bias on an association test without changing explicit attitudes.[^terbeck2012] The result is small, measured on an instrument whose relation to behaviour is itself contested, and has not been shown to alter how anyone treats anyone.

**Psychedelics.** Supported psilocybin sessions entered the discussion later, on the strength of reported shifts in the personality measure of openness and in self-described concern for others. Those outcomes are questionnaires completed by people who know exactly what they took, which is the [[psychedelic-therapy|unblinding problem]] that limits the clinical literature on the same drugs; in people without a diagnosis that literature is small and largely uncontrolled.

The honest summary is that no biomedical intervention has been shown to make a person behave better in any consequential setting. Every effect in this literature is a shift on a questionnaire or a laboratory task, in a single session, in university volunteers.

> [!caution] Distance from the claim
> The proposal concerns preventing catastrophic harm by a small number of highly motivated actors. The evidence concerns millisecond-scale changes in association-test scores among undergraduates. No argument has been offered that bridges the two, and the gap is not a matter of degree.

## Objections

**Freedom to fall.** Harris's objection is that moral worth depends on the agent's capacity to choose otherwise. An intervention that makes wrongdoing impossible, or merely much less likely by bypassing deliberation, converts a moral agent into a well-behaved mechanism, and it collides directly with the claim to mental self-determination set out in [[morphological-freedom]] and, in its legal form, in [[neurorights]].[^harris2011] Douglas and others reply that emotions are not chosen either, and that removing a bias one would reflectively disown restores autonomy rather than reducing it. Harris's second point is harder to answer: if the enhancement acts on motivation rather than on judgement, the agent has no way to evaluate whether the new motivation is correct.

**Moral pluralism.** Deep disagreement about what morality requires is not obviously an error to be corrected. A programme that increases obedience, or conformity, or harm aversion is taking sides in disputes that are live and reasonable. Robert Sparrow argues that the proposal's political naivety is its central flaw: it imagines an intervention deployed by no one in particular, when in practice a state or a firm would decide the content. The objection converges with the [[bioconservatism|bioconservative]] worry, stated most sharply by [[leon-kass]], that a technique for improving people presupposes a settled account of what an improved person is.[^sparrow2014]

**Who takes it.** The people whose dispositions the argument is most concerned about are precisely those least likely to volunteer. Compulsory moral enhancement is rejected by nearly everyone in the debate, including its advocates, which leaves the proposal aimed at a population it does not need to reach. Persson and Savulescu acknowledge this and argue for research now against the possibility of a future in which the calculus changes.

**The science does not exist.** Harris Wiseman's book-length critique argues that the neuroscience underdetermines everything the proposal requires: there is no moral module, the pathways implicated are involved in most of behaviour, and any agent potent enough to shift moral motivation would have effects across the whole of personality.

## Standing and outlook

Moral enhancement occupies an unusual position: it is one of the most-discussed proposals in academic bioethics and one of the least-supported by evidence. Two decades of argument have produced a substantial philosophical literature and no candidate intervention that a researcher would defend as ready for study in the relevant setting. Within [[bioethics-of-enhancement]] it is the clearest case of a proposal whose ethical analysis has run far ahead of anything it could analyse.

Two developments would change that. The first is a pharmacological or neuromodulatory agent with a specific and durable effect on some component of moral motivation, demonstrated behaviourally rather than on a self-report instrument; nothing in the current pipeline resembles this. The second is a shift in the underlying risk argument. If [[artificial-general-intelligence]] or engineered pathogens make the distribution of destructive capability as wide as Persson and Savulescu assume, the question of whether a crude intervention is better than none becomes practical rather than academic. Their critics reply that in that scenario the crude intervention would not work either, and that the effort belongs in institutions, verification, the [[precautionary-principle|precautionary]] regulation of the underlying technologies, and [[differential-technological-development]] instead.

## See also

- [[human-enhancement]]
- [[bioethics-of-enhancement]]
- [[julian-savulescu]]
- [[existential-risk]]
- [[nootropics]]
- [[intelligence-amplification]]
- [[precautionary-principle]]
- [[differential-technological-development]]

## References

[^ps2008]: `paper` Persson, I. and Savulescu, J. "The Perils of Cognitive Enhancement and the Urgent Imperative to Enhance the Moral Character of Humanity." *Journal of Applied Philosophy*, 2008.
[^ps2012]: `book` Persson, I. and Savulescu, J. *Unfit for the Future: The Need for Moral Enhancement*. Oxford University Press, 2012.
[^douglas2008]: `paper` Douglas, T. "Moral Enhancement." *Journal of Applied Philosophy*, 2008.
[^harris2011]: `paper` Harris, J. "Moral Enhancement and Freedom." *Bioethics*, 2011.
[^crockett2010]: `paper` Crockett, M. J., Clark, L., Hauser, M. D. and Robbins, T. W. "Serotonin selectively influences moral judgment and behavior through effects on harm aversion." *Proceedings of the National Academy of Sciences*, 2010. {A single acute dose in healthy volunteers responding to hypothetical dilemmas; no conduct outside the testing session was observed.}
[^dedreu2011]: `paper` De Dreu, C. K. W. et al. "Oxytocin promotes human ethnocentrism." *Proceedings of the National Academy of Sciences*, 2011. {The measures are laboratory bias and economic-game tasks rather than treatment of real out-group members, and how much intranasal oxytocin reaches the brain is disputed.}
[^terbeck2012]: `paper` Terbeck, S. et al. "Propranolol reduces implicit negative racial bias." *Psychopharmacology*, 2012. {The outcome is a score on an implicit association test in a single session; explicit attitudes did not shift and behaviour was not measured.}
[^sparrow2014]: `paper` Sparrow, R. "Better Living Through Chemistry? A Reply to Savulescu and Persson on 'Moral Enhancement'." *Journal of Applied Philosophy*, 2014.

==============================================================================
ARTICLE: morphological-freedom
TITLE:   Morphological freedom
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/morphological-freedom
SOURCE:  https://futurehumanwiki.com/raw/morphological-freedom
==============================================================================

---
title: "Morphological freedom"
slug: "morphological-freedom"
type: "concept"
status: "contested"
horizon: "present"
categories: ["society", "enhancement"]
tags: ["autonomy", "rights", "bodily integrity", "enhancement", "law", "ethics"]
summary: "The claimed right to modify one's own body and mind as one chooses, and the correlative right to refuse modification imposed by others."
updated: "2026-07-27"
issues: ["Habermas, Carl Elliott and the feminist critique are discussed without citations."]
---

```infobox
{
  "caption": "Proposed right in bioethics and political philosophy",
  "rows": [
    { "label": "Term coined", "value": "1993, Max More", "link": "/wiki/max-more" },
    { "label": "Principal statement", "value": "Sandberg, 2001", "link": "/wiki/anders-sandberg" },
    { "label": "Type of right", "value": "Primarily negative" },
    { "label": "Derived from", "value": "Bodily integrity, self-ownership" },
    { "label": "Legal recognition", "value": "None as such" },
    { "label": "Main limits proposed", "value": "Harm, consent, third-party effects" }
  ]
}
```

**Morphological freedom** is the claim that a person has a right to alter their own body and mind — chemically, surgically, genetically or electronically — and an equally strong right not to be altered by anyone else. It is the political principle most closely associated with [[transhumanism]], and the one its advocates regard as doing the real work: not a claim that enhancement is good, but a claim about who decides.

## Origins

[[max-more]] introduced the phrase in a 1993 essay in *Extropy*, defining it as the ability to alter bodily form at will through technologies such as surgery, genetic engineering, nanotechnology and uploading.[^more1993] The formulation was libertarian in spirit and framed as an extension of self-ownership.

[[anders-sandberg]] gave the idea its most-cited treatment in a paper presented at a transhumanist conference in 2001. His argument runs from the right to life and the right to one's own body: if these are meaningful, they include the right to change what one is made of, since a right to a body one may not modify is a right to a cage.[^sandberg2001] Sandberg's more consequential move was to insist that the right is symmetrical. Morphological freedom includes the right to remain unmodified, which makes it a defence against coerced enhancement as well as a licence to enhance, and which distinguishes the position from a simple pro-technology stance.

```timeline
[
  { "year": "1993", "title": "The term is coined", "text": "Max More names morphological freedom in an essay on technological self-transformation, framing it as an extension of self-ownership." },
  { "year": "2000", "title": "Elective amputation halted", "text": "A Scottish hospital stops performing elective amputations on patients requesting removal of healthy limbs after two such operations become public, opening a long debate on the limits of consent." },
  { "year": "2001", "title": "Sandberg's statement", "text": "Anders Sandberg argues the right is symmetrical: it includes the right to refuse modification, not only the right to pursue it." },
  { "year": "2018", "title": "Consent is not a defence", "text": "The Court of Appeal of England and Wales holds in R v BM that a customer's consent does not make lawful serious body modification performed by a non-medical practitioner." },
  { "year": "2019", "title": "A diagnosis appears", "text": "Body integrity dysphoria enters the World Health Organization's ICD-11, giving a clinical name to the desire for elective disability and complicating the autonomy argument." },
  { "year": "2021", "title": "Mental integrity in law", "text": "Chile amends its constitution to protect brain activity and mental integrity, the first national provision aimed at the mind rather than the body." }
]
```

## The argument

The positive case has three steps. The first is that bodily autonomy is already recognised as near-absolute in one direction: no one may operate on a competent adult without consent, and refusal of life-saving treatment is protected in most legal systems. The second is that there is no principled asymmetry between refusing an intervention and requesting one, given the same information and the same competence. The third is that objections to particular modifications — that they are unnatural, undignified, or offensive — are objections to the modification, not reasons to relocate the decision to someone else.

Morphological freedom is thus offered primarily as a negative right, a constraint on interference rather than an entitlement to provision. Sandberg is explicit that it does not oblige anyone to supply a modification or to pay for it, which insulates the principle from the distributional objections raised in [[access-and-inequality]] at the cost of leaving them unanswered. [[nick-bostrom]] makes a parallel argument in "In Defense of Posthuman Dignity", holding that dignity attaches to persons rather than to a particular biological configuration.[^bostrom2005]

The principle is invoked across a wider range of cases than its transhumanist origins suggest. Reproductive autonomy, gender-affirming care, cosmetic surgery, tattooing, assisted dying, [[uterus-transplantation|elective transplantation to enable a pregnancy]] and the use of unapproved drugs by terminally ill patients all involve a claim that the body in question belongs to the person in it.

## Legal footing

No jurisdiction recognises morphological freedom by name. Several protect its components. The European Convention on Human Rights has been read to protect personal autonomy and physical and psychological integrity under the right to private life; the Charter of Fundamental Rights of the European Union states a right to the integrity of the person, framed as protection against intervention rather than as a licence for it. Most of the doctrine, in other words, protects the refusal half of the right and says little about the request half.

Where the request half has been tested, it has generally lost. In *R v BM* the Court of Appeal of England and Wales held that consent is no defence to charges arising from tongue-splitting and the removal of an ear and a nipple by a body-modification practitioner, following the earlier reasoning in *R v Brown* on consensual injury.[^rvbm2018] The court's ground was that only recognised medical procedures fall within the exception to the rule that one cannot consent to serious harm, which makes the professional status of the operator, rather than the wishes of the person modified, the operative fact.

> [!key] The asymmetry in existing law
> Legal systems protect the right to refuse almost absolutely and the right to request only where a licensed professional is willing to act within recognised practice. Morphological freedom, as its proponents state it, would require collapsing that asymmetry — which is precisely what the cases have declined to do.

## Limits others propose

Four limits are commonly urged, and advocates of the principle accept some of them.

**Harm to others.** Uncontroversial in principle and difficult in application, since [[germline-editing|heritable modification]] and self-spreading interventions affect people who never consented. Sandberg accepts this limit explicitly.

**Consent and capacity.** Modification of children, of the incapacitated, and of future people is outside the principle by construction, which is why [[designer-babies]] and [[procreative-beneficence]] cannot be settled by appeal to morphological freedom. Habermas's argument that a designed person stands in an asymmetric relation to their designer is a claim about exactly this gap.

**Positional externalities.** Where a modification confers advantage only relative to others, individual choice aggregates into collective coercion. This is the structure analysed in [[enhancement-arms-race]] and in [[enhancement-in-sport]], and it is the strongest reason a principle of non-interference is not sufficient on its own.

**Irreversibility.** Some proposals distinguish reversible from irreversible modification and apply a higher standard to the latter, on the grounds that a person's future self also has a claim.

## Criticism

The autonomy premise itself is the main target. Feminist analyses of cosmetic surgery have argued for decades that choices made under strong social pressure are not straightforwardly free, and that a principle which counts only formal coercion will license outcomes it purports to protect people from. Carl Elliott's work on enhancement in American medicine makes a related point about how quickly an optional intervention becomes a norm.

A second criticism holds that the principle is empty without a theory of what is worth wanting. Bioconservative writers argue that morphological freedom describes a procedure while remaining silent on whether any given transformation is good for the person, and that a framework with nothing to say about that has abandoned the question that matters; see [[bioconservatism]] and [[leon-kass]].

A third comes from disability studies. Body integrity dysphoria, recognised in the ICD-11, describes people who seek amputation or paralysis of healthy limbs. Advocates of morphological freedom face a dilemma: refusing these requests requires appeal to a norm of function that the principle was designed to reject, while granting them requires accepting elective disability as a legitimate exercise of the same right. The debate in [[disability-rights-and-enhancement]] approaches the same territory from the other direction, and neither side has produced a criterion that handles both cases without special pleading.

## Current relevance

The most active front has moved from the body to the brain. Proposals for [[neurorights]] and for [[mental-privacy]] protection are morphological freedom applied to neural tissue and neural data, and they have made faster legal progress than the bodily version ever did: Chile's constitutional amendment, subsequent state neural-data statutes in the United States, and a UNESCO recommendation on neurotechnology ethics all address mental integrity directly.

Whether that progress vindicates the principle is unclear. Every one of those instruments protects the right not to be modified or read, and none creates a right to modify. The half of morphological freedom that transhumanists care about most remains, thirty years after it was named, without legal recognition anywhere.

## See also

- [[human-enhancement]]
- [[transhumanism]]
- [[neurorights]]
- [[bioethics-of-enhancement]]
- [[biohacking]]
- [[disability-rights-and-enhancement]]
- [[bioconservatism]]
- [[anders-sandberg]]

## References

[^more1993]: `paper` More, M. "Technological Self-Transformation: Expanding Personal Extropy." *Extropy*, 1993. {Extropy was the movement's own magazine rather than a reviewed journal, so this is where the term was coined, not where it was tested.}
[^sandberg2001]: `book` Sandberg, A. "Morphological Freedom — Why We Not Just Want It, but Need It." Presented at TransVision, 2001; reprinted in *The Transhumanist Reader*, Wiley-Blackwell, 2013.
[^bostrom2005]: `paper` Bostrom, N. "In Defense of Posthuman Dignity." *Bioethics*, 2005.
[^rvbm2018]: `law` *R v BM* [2018] EWCA Crim 560, Court of Appeal of England and Wales. {A criminal appeal binding only in England and Wales; the operator was a body-modification practitioner, and the ruling turns on that rather than on what the customers wanted.}

==============================================================================
ARTICLE: myoelectric-prosthetics
TITLE:   Myoelectric prosthetics
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/myoelectric-prosthetics
SOURCE:  https://futurehumanwiki.com/raw/myoelectric-prosthetics
==============================================================================

---
title: "Myoelectric prosthetics"
slug: "myoelectric-prosthetics"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["bodies", "cybernetics"]
tags: ["prosthetics", "amputation", "emg", "assistive technology", "rehabilitation", "sensory feedback"]
summary: "Externally powered artificial limbs controlled by electrical activity recorded from the wearer's residual muscles, the dominant design for powered upper-limb prostheses."
updated: "2026-07-27"
humanEvidence: "Routine clinical care in human amputees since the 1960s; a randomized trial supports targeted muscle reinnervation for neuroma pain rather than for control, and osseointegrated implanted-electrode arms have been used at home by a small number of patients."
access: "Commercially available and fitted through prosthetics clinics in high-income countries; advanced hands run to tens of thousands of dollars before socket and therapy, and reimbursement varies sharply between health systems."
reversibility: "partly-reversible"
issues: ["Device cost, reimbursement and 3D-printed-hand claims are stated without sources."]
---

```infobox
{
  "caption": "Class of prosthetic limb",
  "rows": [
    { "label": "Control signal", "value": "Surface electromyography" },
    { "label": "First clinical devices", "value": "1960s (USSR, Vienna)" },
    { "label": "Standard control", "value": "Two-site proportional" },
    { "label": "Advanced control", "value": "Pattern recognition" },
    { "label": "Key surgical adjunct", "value": "Targeted muscle reinnervation" },
    { "label": "Typical device cost", "value": "Tens of thousands of dollars" },
    { "label": "Abandonment", "value": "Roughly one user in five" },
    { "label": "Readiness", "value": "Routine clinical care" }
  ]
}
```

**Myoelectric prosthetics** are artificial limbs driven by motors and controlled by the electrical
activity of the wearer's own remaining muscles. Surface electrodes in the socket pick up
electromyographic signals a few hundred microvolts in amplitude, the prosthesis converts their
envelope into motor commands, and the user learns to produce the required contractions. The
approach dominates powered upper-limb prosthetics, and its persistent problem is not the hand — modern
hands are mechanically excellent — but the narrow, noisy control channel available to drive it.

## How it works

Muscle fibres depolarize when they contract, and the summed extracellular field of many motor units
is detectable through the skin. A pair of differential electrodes pressed against a residual muscle
records this signal; the prosthesis rectifies and smooths it into an amplitude envelope and maps
that envelope onto a motor's velocity. Squeeze harder and the hand closes faster.

The classical arrangement is two-site proportional control: one electrode over a flexor group, one
over an extensor group, driving a single degree of freedom in opposite directions. A wrist rotator
or elbow requires switching modes, usually by co-contracting both muscles or by holding a
contraction past a timeout. Each additional joint therefore costs the user a deliberate,
non-intuitive switching action, which is why devices with five actuated fingers are commonly
operated as though they had one.

Signal quality is the limiting variable. Electrode contact shifts as the socket moves, sweat changes
skin impedance, muscle fatigue changes the envelope, and limb position alters which muscles are
recruited for the same intended action. A control scheme trained sitting still on a bench often
degrades when the arm is raised overhead.

## Control strategies

**Direct control** maps one muscle site to one function. It is robust, transparent to the user, and
severely limited in the number of functions it can address.

**Pattern recognition** records from an array of electrodes around the residual limb and classifies
the spatial pattern of activity to infer which movement the user is attempting, selecting a grip or
joint automatically rather than requiring a switching gesture. Commercial pattern-recognition
controllers have been available since the 2010s. Laboratory accuracy is high; performance in daily
life is degraded by the same electrode-shift and limb-position effects, and users must retrain the
classifier periodically.

**Regression and continuous decoding** treat the problem as estimating joint velocities directly
rather than selecting from a menu of classes, allowing simultaneous multi-joint control. This is the
same shift from classification to continuous estimation that improved
[[neural-decoding]] in cortical systems, and it faces the same non-stationarity problem.

Cortical control is sometimes proposed as the alternative. For most amputees it is the wrong tool:
the peripheral signal is intact and far easier to access, and a [[brain-computer-interface]] adds
neurosurgical risk to solve a problem that surgery on the arm can address more cheaply.

## Surgical amplification of the signal

The most consequential advances have come from operating on the residual limb rather than improving
the electronics.

**Targeted muscle reinnervation** (TMR), developed by Todd Kuiken's group and first performed in
2002 on a patient with bilateral shoulder disarticulation, transfers the severed nerves that once
controlled the arm onto spare muscle in the chest or upper arm. The reinnervated muscle then
contracts when the user thinks about the original movement, converting a nerve signal that had
nowhere to go into a recordable, intuitive EMG site.[^kuiken2009] TMR also reduces neuroma and
phantom limb pain, and a randomised trial supported its use for pain even in patients not seeking
myoelectric control.[^dumanian2019]

**Regenerative peripheral nerve interfaces** wrap a transected nerve in a free muscle graft, which
the nerve reinnervates and which then acts as a biological amplifier. Implanted electrodes on such
grafts have supported real-time individual finger control in upper-limb amputees.[^vu2020]

**The agonist-antagonist myoneural interface** (AMI), developed in Hugh Herr's group, surgically
reconnects opposing muscle pairs in the residual limb so that contraction of one stretches the
other, restoring the proprioceptive signalling that ordinary amputation destroys. In a trial of
below-knee amputees with AMI surgery, neural control of a powered ankle produced walking speeds and
gait biomechanics closer to unimpaired norms than in amputees with standard surgery.[^song2024]

**Skeletal attachment** through [[osseointegration]] removes the socket entirely and allows
electrode leads to pass through the implant rather than across the skin, yielding stable recordings
and a path for stimulation back into the nerve. Self-contained arm prostheses using this route have
been used at home for years by a small number of patients.[^ortizcatalan2020]

## Development history

```timeline
[
  { "year": "1948–1960s", "title": "First myoelectric arms", "text": "Reinhold Reiter demonstrates myoelectric control in Munich; Soviet and Viennese groups field the first clinical devices, and the 'Russian hand' is exhibited internationally." },
  { "year": "1960s–1980s", "title": "Clinical adoption", "text": "Two-site proportional control becomes standard; thalidomide-related limb difference in Europe drives paediatric fitting programmes." },
  { "year": "2002–2009", "title": "Targeted muscle reinnervation", "text": "Nerve transfers create new intuitive control sites; a case series demonstrates real-time control of multifunction arms." },
  { "year": "2007–2015", "title": "Multi-articulating hands", "text": "Independently actuated fingers reach the market, offering many grip patterns but no corresponding increase in control channels." },
  { "year": "2013–2020", "title": "Pattern recognition and implanted electrodes", "text": "Commercial pattern-recognition controllers are released; osseointegrated systems with implanted electrodes and nerve stimulation are used long-term outside the laboratory." },
  { "year": "2024", "title": "Restored proprioception in gait", "text": "Amputees with surgically reconstructed agonist-antagonist muscle pairs achieve near-normal walking biomechanics under continuous neural control of a powered ankle." }
]
```

## Abandonment and the case for the hook

Surveys of upper-limb prosthesis users consistently find that a substantial minority stop using
their device, with rejection of externally powered arms reported at roughly one in five among
adults and higher among children.[^biddiss2007] The reasons are unglamorous and repeat across
studies: weight, lack of sensory feedback, insufficient durability, discomfort, slow donning, poor
reliability in wet or dusty conditions, and cost of repair.

Body-powered prostheses — a harness across the shoulders operating a split hook through a cable — are
lighter, cheaper by an order of magnitude, essentially unbreakable, and give the user crude force
feedback through cable tension, a form of proprioception no commercial myoelectric hand provides.
Many experienced amputees prefer them for work and keep a myoelectric hand for social settings. Any
honest account of the field has to record that the most advanced device is not always the one people
choose.

> [!stat] The mismatch
> A multi-articulating hand offers a dozen or more programmed grips. Standard two-site control
> supplies one degree of freedom at a time. Most of the mechanism's capability is inaccessible in
> ordinary use, which is why control research has produced more functional gain than hand design.

Cost compounds the problem. Advanced hands run to tens of thousands of dollars before the socket,
fitting, and therapy that determine whether the device works, and reimbursement varies sharply
between health systems and between civilian and veteran populations within the same country. Most
of the world's amputees, concentrated in low- and middle-income countries and in populations
affected by conflict and diabetes, have no realistic access to any powered device — the pattern
described under [[access-and-inequality]]. Low-cost 3D-printed powered hands have narrowed the gap
for some paediatric users without closing it.

## Limitations

The channel is the constraint. Surface EMG offers a handful of noisy, correlated signals from
muscles that were not designed to be independently controlled, and no amount of hand engineering
adds control dimensions. Sensory feedback remains the largest functional gap: without it, users
watch their hand continuously and cannot modulate grip force by feel, which makes handling fragile
objects slow and error-prone. Implanted nerve stimulation restores graded touch in research
settings, but no commercial upper-limb prosthesis provides it as of 2026. The contrast with the
[[cochlear-implant]], which has delivered an artificial sensory code to hundreds of thousands of
people for decades, is instructive: hearing tolerates a crude code, whereas touch used for motor
control needs timing and force information the current interfaces cannot supply.

Prosthetic limbs also do not restore the limb. They restore a subset of its functions to a person
who must actively operate them, which is why users describe fatigue as a reason for abandonment and
why the field's ethical literature resists framing amputation as a solved problem — a point
developed under [[disability-rights-and-enhancement]]. A minority of users take the opposite view
and treat the prosthesis as a platform to be customized rather than a substitute for a lost hand,
the position argued under [[morphological-freedom]], and interchangeable tool attachments and
non-anthropomorphic end effectors follow naturally from it.

## Outlook

The near-term direction is surgical and biological rather than mechanical: nerve transfers,
regenerative interfaces, and reconstructed muscle pairs that create better signals for existing
electronics, combined with implanted rather than surface recording. Bidirectional systems that
combine implanted stimulation for touch with skeletal attachment are the plausible next clinical
product, following the pattern of [[neuroprosthetics]] generally, where writing information in lags
reading it out. Adding channels the body does not already have — vibration, temperature, or grip
force mapped onto an unrelated nerve — shades into [[sensory-augmentation]], and users of
touch-restoring systems have reported that the mapping need not be anatomically faithful to be
usable.

Two further prospects sit outside the incremental path. Powered [[exoskeleton]] devices and
prosthetic limbs are converging in actuator technology, and competitive prosthetic athletics has
already forced the question of when a replacement becomes an advantage, examined under
[[enhancement-in-sport]] and [[human-enhancement]]. And biological replacement through
[[limb-regeneration]] or engineered tissue would make the entire control problem moot; nothing in
the mammalian regeneration literature suggests that is close.

## See also

- [[osseointegration]]
- [[neuroprosthetics]]
- [[exoskeleton]]
- [[brain-computer-interface]]
- [[sensory-augmentation]]
- [[limb-regeneration]]
- [[enhancement-in-sport]]
- [[disability-rights-and-enhancement]]

## References

[^kuiken2009]: `paper` Kuiken, T. A. et al. "Targeted muscle reinnervation for real-time myoelectric control of multifunction artificial arms." *JAMA*, 2009.

[^dumanian2019]: `paper` Dumanian, G. A. et al. "Targeted muscle reinnervation treats neuroma and phantom pain in major limb amputees: a randomized clinical trial." *Annals of Surgery*, 2019. {The randomized endpoint is pain, not prosthetic control, so it supports the surgery for a reason unrelated to myoelectric performance.}

[^vu2020]: `paper` Vu, P. P. et al. "A regenerative peripheral nerve interface allows real-time control of an artificial hand in upper limb amputees." *Science Translational Medicine*, 2020.

[^song2024]: `paper` Song, H. et al. "Continuous neural control of a bionic limb restores biomimetic gait after amputation." *Nature Medicine*, 2024. {A small comparison of below-knee amputees with and without the surgery; the measured outcome is walking biomechanics, not everyday use or long-term durability.}

[^ortizcatalan2020]: `paper` Ortiz-Catalan, M. et al. "Self-contained neuromusculoskeletal arm prostheses." *New England Journal of Medicine*, 2020.

[^biddiss2007]: `paper` Biddiss, E., Chau, T. "Upper limb prosthesis use and abandonment: a survey of the last 25 years." *Prosthetics and Orthotics International*, 2007. {A review pooling surveys published over 25 years rather than a new measurement of its own.}

==============================================================================
ARTICLE: myostatin-inhibition
TITLE:   Myostatin inhibition
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/myostatin-inhibition
SOURCE:  https://futurehumanwiki.com/raw/myostatin-inhibition
==============================================================================

---
title: "Myostatin inhibition"
slug: "myostatin-inhibition"
type: "intervention"
status: "experimental"
horizon: "late 2020s"
trl: 7
categories: ["enhancement", "bodies"]
tags: ["muscle", "sarcopenia", "enhancement", "gene therapy", "clinical trials", "follistatin"]
summary: "Blockade of the growth factor myostatin to increase skeletal muscle mass, an effect reproduced across species that has repeatedly failed to translate into functional benefit in humans."
updated: "2026-07-27"
humanEvidence: "Antibody trials in muscular dystrophy and inclusion body myositis raised lean mass but missed functional endpoints; one reported a positive phase 3 in spinal muscular atrophy, and the huge-muscle phenotypes come from lifelong gene loss, not adult dosing."
access: "No approved product: the spinal muscular atrophy antibody was still under United States regulatory review as of mid-2026, while follistatin gene-transfer constructs are sold to individuals by ventures sited beyond the reach of drug regulators."
reversibility: "context"
issues: ["The 2024 spinal muscular atrophy phase 3 result and its regulatory hold carry no citation."]
---

```infobox
{
  "caption": "Muscle-directed intervention",
  "rows": [
    { "label": "Target", "value": "Myostatin (GDF-8)" },
    { "label": "Discovered", "value": "1997" },
    { "label": "Receptor", "value": "ActRIIB" },
    { "label": "Natural antagonist", "value": "Follistatin" },
    { "label": "Effect on muscle mass", "value": "Increase, reliably" },
    { "label": "Effect on strength", "value": "Inconsistent" },
    { "label": "Approved for enhancement", "value": "No" }
  ]
}
```

**Myostatin inhibition** is the blockade of myostatin, a secreted protein that restrains skeletal muscle growth, in order to increase muscle mass. Animals lacking functional myostatin are visibly and dramatically muscular, and the finding has been reproduced in cattle, dogs, sheep, mice and at least one human child. Two decades of clinical development have shown that the mass gain transfers to human patients and that the functional benefit usually does not. The gap between bigger muscle and better muscle is the central fact of the field.

## How it works

Myostatin, also called growth differentiation factor 8, is a member of the transforming growth factor β superfamily secreted by skeletal muscle itself. It circulates as a latent complex, is activated by proteolysis, and binds the activin type II receptor ActRIIB. Signalling through the type I receptors ALK4 and ALK5 phosphorylates Smad2 and Smad3, which suppresses the transcriptional programme for muscle protein synthesis and antagonises the Akt–mTOR pathway that drives hypertrophy. The system is a negative feedback brake: muscle produces the signal that limits its own growth.

Interventions act at every point in that chain.

- **Antibodies against myostatin** or against its latent precursor neutralise the ligand.
- **Antibodies or ligand traps against ActRIIB** block the receptor, which also blocks activin A and other ligands that use it.
- **Follistatin**, a natural antagonist that sequesters myostatin and activin A, can be supplied as a protein or, more commonly in experimental work, expressed from a viral vector.
- **Direct disruption of the MSTN gene** by [[crispr-cas9|nuclease editing]] has been done in cattle, pigs and sheep, though not in humans.

The receptor-level approach is more potent than ligand neutralisation because it blocks several ligands at once. That is also why it has caused more trouble.

## Where the evidence came from

```timeline
[
  { "year": "1997", "title": "Gene identified", "text": "McPherron, Lawler and Lee report that mice lacking a new TGF-β family member develop two to three times normal muscle mass, and show the same gene is mutated in double-muscled cattle." },
  { "year": "2004", "title": "The human case", "text": "A German infant with a homozygous splice-site mutation in MSTN is described with pronounced muscle hypertrophy and no reported illness attributable to it." },
  { "year": "2007", "title": "Dogs and force", "text": "A myostatin mutation is found in racing whippets, with heterozygotes overrepresented among the fastest dogs, while independent work shows myostatin-null mouse muscle generates less force per unit area." },
  { "year": "2008–2019", "title": "Clinical failures", "text": "Antibody and ligand-trap programmes in muscular dystrophy and inclusion body myositis increase lean mass but miss their functional endpoints, and a receptor trap is halted for vascular side effects." },
  { "year": "2021–2025", "title": "Repositioning", "text": "Attention shifts from muscle disease to preserving muscle during rapid weight loss on incretin drugs, and to a pro-myostatin antibody in spinal muscular atrophy that reports a positive phase 3 result." }
]
```

The founding experiments are unusually clean. Knockout mice showed muscles two to three times normal size, built by both hyperplasia and hypertrophy, and the same gene turned out to explain the double-muscling of Belgian Blue and Piedmontese cattle.[^mcpherron1997][^mcpherron1997b] A child with a homozygous splice-site mutation, reported in 2004, had visible muscle hypertrophy from infancy.[^schuelke2004] The case is frequently cited as evidence that human myostatin loss is benign; it is a single individual, followed as a case report, and it does not establish long-term safety.

Racing whippets supplied a more informative natural experiment. Dogs carrying two copies of a myostatin mutation are grossly muscled and unremarkable as racers, while heterozygotes are faster than wild-type animals.[^mosher2007] Partial inhibition helps; complete inhibition does not.

## Muscle without strength

Myostatin-null mouse muscle generates less force per unit of cross-sectional area than normal muscle, and shows reduced oxidative capacity, fewer capillaries per fibre and altered tendon properties.[^amthor2007] The added tissue is not proportionally functional. Human trials have repeatedly reproduced the pattern in a milder form: lean mass rises measurably and the walk test, the strength score or the disease endpoint does not move.

The first antibody trial, in adults with several muscular dystrophies, established safety without demonstrating benefit.[^wagner2008] Subsequent programmes in Duchenne muscular dystrophy and in sporadic inclusion body myositis failed their primary endpoints despite producing the expected changes in body composition. A receptor-level ligand trap was stopped in a paediatric dystrophy trial after nosebleeds and dilated skin vessels appeared, an effect attributed to blocking ligands other than myostatin.

> [!key] Why the failures were informative
> In dystrophy, the muscle that grows is still dystrophin-deficient and still fragile. Adding mass to a tissue whose defect is mechanical vulnerability was never obviously a treatment, and the trials arguably tested the wrong indication rather than the wrong mechanism.

## Current state

Four lines of work are active as of 2026, and most have moved away from muscular dystrophy.

**Spinal muscular atrophy.** An antibody targeting the latent pro-form of myostatin reported a positive phase 3 result in 2024 as an add-on to disease-modifying therapy, the first clear functional win for the mechanism. Its United States regulatory review was then held up by questions about a third-party manufacturing facility rather than by the trial data, and the file had not resolved into a marketing authorisation by mid-2026.

**Weight loss with incretin drugs.** A substantial fraction of the weight lost on [[glp-1-receptor-agonists]] is lean mass. Combining an activin-pathway blocker with semaglutide or tirzepatide, so that fat is lost while muscle is spared, is now the most heavily capitalised application of the mechanism, and several such combination trials are running. Earlier work with a receptor antibody in obesity and type 2 diabetes had already shown large reductions in fat mass alongside increases in lean mass, which is what made this indication attractive.

**Disuse and unloading.** Blocking myostatin and activin A protected both muscle and bone in mice flown to the International Space Station, a result with direct relevance to [[space-medicine]] and to the countermeasure problem described in [[microgravity-adaptation]].[^lee2020] Bed rest, immobilisation after surgery and the muscle loss of aging described in [[stem-cell-exhaustion]] present the same physiological problem.

**Gene therapy.** Follistatin delivered by adeno-associated virus was tested in small open-label trials in Becker muscular dystrophy and inclusion body myositis, with modest changes in walking distance and no control group.[^mendell2015] Follistatin constructs have also been administered outside any trial, marketed directly to individuals by ventures sited beyond the reach of drug regulators, a practice examined in [[biohacking]] and one of the few concrete instances of the scenario anticipated in [[gene-doping]]. Several of the individuals involved took the construct as an anti-aging measure rather than an athletic one, which places it alongside the self-experiments described in [[gene-therapy-for-aging]].

## Limitations and risks

The mechanism has three durable problems. Muscle mass and muscle quality dissociate, so an intervention validated on body composition may not help anyone. Blockade at the receptor catches ligands beyond myostatin, and the activin pathway has roles in vasculature, bone and reproduction, which is where the observed adverse effects have come from. And the effect size in adult humans is far smaller than the animal phenotypes suggest, because adult myostatin blockade acts on an already-differentiated muscle rather than on development. How reversible any of this is depends on the route: an antibody or ligand trap stops acting once dosing stops, whereas an AAV-delivered follistatin construct is a single administration whose expression cannot afterwards be withdrawn.

For non-medical use the calculus is worse. A healthy adult taking a myostatin inhibitor accepts an unknown long-term risk profile in exchange for mass that may not increase strength, when [[exercise-and-aging|resistance training]] produces functional hypertrophy with well-characterised side effects and no cost. This is the standard comparison that any candidate in [[human-enhancement]] has to survive, and myostatin blockade has not yet survived it. In competition the question is moot in any case, since activin-pathway blockers and follistatin constructs alike are prohibited under the rules described in [[enhancement-in-sport]].

> [!caution] Self-administration
> Follistatin gene-transfer products sold outside regulatory oversight have not been through controlled trials, have no published safety follow-up, and use vectors whose expression cannot be switched off. The published human data on [[somatic-gene-therapy|gene transfer]] with [[aav-vectors|AAV]] include serious immune and hepatic events at high systemic doses.

## Outlook

The most likely path to routine use runs through muscle preservation rather than muscle enhancement: keeping lean mass during aggressive pharmacological weight loss, during immobilisation, and in the sarcopenia of old age, where preserving function has an obvious clinical meaning and a measurable endpoint. Grip strength, gait speed and the ability to rise from a chair are among the few functional measures that appear in nearly every definition of [[healthspan]], which makes muscle one of the more tractable targets in geroscience. That framing also sets the bar the field has to clear, since the question in every one of those settings is whether preserved mass translates into preserved function.

An unresolved question sits underneath all of it. Myostatin is a brake with an evolutionary rationale, and the metabolic cost of maintaining large muscle is the leading candidate explanation for why the brake exists. Nobody has shown what an adult human pays, over decades, for releasing it.

## See also

- [[gene-doping]]
- [[enhancement-in-sport]]
- [[exercise-and-aging]]
- [[somatic-gene-therapy]]
- [[space-medicine]]
- [[human-enhancement]]
- [[biohacking]]
- [[gene-therapy-for-aging]]

## References

[^mcpherron1997]: `paper` McPherron, A. C., Lawler, A. M. and Lee, S.-J. "Regulation of skeletal muscle mass in mice by a new TGF-β superfamily member." *Nature*, 1997.
[^mcpherron1997b]: `paper` McPherron, A. C. and Lee, S.-J. "Double muscling in cattle due to mutations in the myostatin gene." *Proceedings of the National Academy of Sciences*, 1997.
[^schuelke2004]: `paper` Schuelke, M. et al. "Myostatin mutation associated with gross muscle hypertrophy in a child." *New England Journal of Medicine*, 2004.
[^mosher2007]: `paper` Mosher, D. S. et al. "A mutation in the myostatin gene increases muscle mass and enhances racing performance in heterozygote dogs." *PLoS Genetics*, 2007.
[^amthor2007]: `paper` Amthor, H. et al. "Lack of myostatin results in excessive muscle growth but impaired force generation." *Proceedings of the National Academy of Sciences*, 2007. {Measured in myostatin-null mice, whose muscle was enlarged from development onward; blocking myostatin in an adult is a different manipulation.}
[^wagner2008]: `paper` Wagner, K. R. et al. "A phase I/II trial of MYO-029 in adult subjects with muscular dystrophy." *Annals of Neurology*, 2008.
[^lee2020]: `paper` Lee, S.-J. et al. "Targeting myostatin/activin A protects against skeletal muscle and bone loss during spaceflight." *Proceedings of the National Academy of Sciences*, 2020. {The animals were mice flown to the International Space Station; no comparable countermeasure trial has been run in astronauts.}
[^mendell2015]: `paper` Mendell, J. R. et al. "A phase 1/2a follistatin gene therapy trial for becker muscular dystrophy." *Molecular Therapy*, 2015. {An open-label study in a handful of patients with no control arm, so the reported walking-distance changes cannot be separated from ordinary variation.}

==============================================================================
ARTICLE: nad-precursors
TITLE:   NAD+ precursors
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/nad-precursors
SOURCE:  https://futurehumanwiki.com/raw/nad-precursors
==============================================================================

---
title: "NAD+ precursors"
slug: "nad-precursors"
type: "intervention"
status: "contested"
horizon: "present"
trl: 4
categories: ["longevity"]
tags: ["nad", "sirtuins", "supplements", "aging", "metabolism", "regulation"]
summary: "Supplements such as nicotinamide riboside and NMN that raise cellular NAD+ levels, with strong pharmacokinetic evidence and weak evidence of any functional benefit."
updated: "2026-07-27"
humanEvidence: "Oral nicotinamide riboside reproducibly raises blood NAD+ in human trials; no trial has shown a functional benefit in healthy older adults, and the results on mitochondria and stem cell decline come from mice."
access: "Nicotinamide riboside is sold freely as a dietary supplement in the United States; NMN's supplement status there is disputed and major retailers delisted it, and clinics sell intravenous NAD+ infusions direct to consumers."
reversibility: "reversible"
issues: ["The NMN regulatory determination and the retailer delistings are described without a citation."]
---

```infobox
{
  "caption": "Class of dietary supplement",
  "rows": [
    { "label": "Main compounds", "value": "Nicotinamide riboside, NMN" },
    { "label": "Target molecule", "value": "NAD+" },
    { "label": "Proposed mechanism", "value": "Sirtuin and PARP substrate supply" },
    { "label": "Raises blood NAD+", "value": "Yes, reproducibly" },
    { "label": "Proven functional benefit", "value": "No" },
    { "label": "US regulatory status", "value": "Disputed for NMN" },
    { "label": "Readiness", "value": "TRL 4" }
  ]
}
```

**NAD+ precursors** are compounds, chiefly nicotinamide riboside (NR) and nicotinamide
mononucleotide (NMN), taken to raise cellular levels of nicotinamide adenine dinucleotide, a
coenzyme that carries electrons in metabolism and is consumed as a substrate by several classes of
enzyme. The rationale is that NAD+ declines with age in at least some tissues and that restoring it
should restore the activity of the enzymes that depend on it. The supplements do raise NAD+ in blood
reliably; whether that produces any benefit in humans is unresolved, and the field's commercial
intensity considerably exceeds its evidence.

```keyfacts
[
  { "value": "2004", "label": "NR identified as an NAD+ precursor", "note": "Brenner and colleagues, in yeast" },
  { "value": "Yes", "label": "Blood NAD+ rises on supplementation", "note": "established in multiple human pharmacokinetic studies" },
  { "value": "None", "label": "Confirmed clinical outcome benefit", "note": "as of 2026, in healthy older adults" }
]
```

## The biology

NAD+ has two distinct jobs. As a redox cofactor it cycles between NAD+ and NADH without being
consumed, shuttling electrons through glycolysis and the mitochondrial electron transport chain. As a
substrate it is destroyed: sirtuins cleave it to remove acyl groups from proteins, PARP enzymes
consume it while responding to DNA damage, and the ectoenzyme CD38 degrades it outright. The second
role is what links NAD+ to aging, because the consuming enzymes are all more active in aged tissue —
more DNA damage means more PARP activity, and CD38 rises with the immune activation characteristic
of [[inflammaging]] and with the accumulation of [[cellular-senescence|senescent cells]]. Falling
NAD+ availability is usually placed downstream of genomic instability and upstream of
[[mitochondrial-dysfunction|mitochondrial decline]] in the standard maps of the
[[hallmarks-of-aging]].

The salvage pathway recycles nicotinamide back to NAD+ through NMN. NR enters one step earlier,
converted to NMN by nicotinamide riboside kinases. Orally administered NMN is largely broken down in
the gut and liver before reaching tissues, and how much of an oral dose ever reaches a muscle cell
intact remains contested.

> [!caution] The decline is less certain than it sounds
> "NAD+ falls with age" is repeated as settled. NAD+ degrades within minutes of tissue collection,
> making measurement difficult, and while declines are well documented in aged mouse tissues, human
> data are mixed — several studies have failed to find an age-related fall in skeletal muscle. The
> premise of the whole intervention is less secure than its marketing implies.

## Sirtuins and the resveratrol legacy

Interest in NAD+ derives largely from sirtuins, a family of NAD-dependent enzymes proposed in the
early 2000s as conserved longevity regulators and as the mediators of
[[caloric-restriction|dietary restriction]]. That programme has aged poorly. The claim that
resveratrol directly activates SIRT1 was traced to an artefact of the fluorophore-labelled peptide
used in the assay, with activation disappearing when native substrates were used. Reported lifespan
extension from sirtuin overexpression in worms and flies failed to replicate when background
genotypes were controlled.[^burnett2011]

Sirtuins remain real and important enzymes, but the case that they mediate any substantial part of
mammalian aging is weaker than it was in 2006. The commercial and popular framing of NAD+
supplementation still rests heavily on the older story, promoted most visibly by
[[david-sinclair]], whose claims in this area have drawn sustained criticism from other
biogerontologists.

## Human trials

Human studies divide cleanly into pharmacokinetics, where results are consistent, and outcomes, where
they are not.

Oral NR raises blood NAD+ in a dose-dependent way, a result established in the first human
pharmacokinetic study and replicated repeatedly.[^trammell2016] Beyond that, trials have been small,
short, and heterogeneous. Six weeks of NR in healthy middle-aged and older adults raised NAD+ and, in
exploratory analysis, was associated with lower systolic blood pressure and arterial stiffness in
participants with elevated readings.[^martens2018] In aged men, NR raised muscle NAD+ metabolites and
lowered circulating inflammatory cytokines but produced no change in mitochondrial bioenergetics or
physical function.[^elhassan2019] Several trials in insulin-resistant and obese participants found no
improvement in insulin sensitivity.

For NMN, the most cited result is a small trial in prediabetic postmenopausal women reporting improved
muscle insulin sensitivity, with most other endpoints unchanged.[^yoshino2021] Trials reporting
improved walking distance or reduced fatigue have generally been small and short.

The mouse literature is far more encouraging. NAD+ repletion improves mitochondrial function,
counteracts [[stem-cell-exhaustion|the decline of muscle stem cells]], and improves several measures
of tissue maintenance in aged mice. The gap between that literature and the human results is the
central fact about this intervention. It is a familiar gap, shared with [[senolytics]] and with
[[parabiosis-and-young-blood|circulating-factor therapies]]. Unlike [[rapamycin]], NAD+ precursors
have not extended lifespan in the multi-site protocols that the field treats as the standard for
rodent longevity claims.

## Regulation and commerce

NR is sold as a [[dietary-supplements|dietary supplement]] in the United States, with regulatory
acknowledgment as a new dietary ingredient. NMN's status is disputed: US regulators concluded that
because NMN had been authorized for investigation as a new drug, it is excluded from the dietary
supplement definition, a determination that led major retailers to delist NMN products and prompted
industry petitions that remained unresolved as of 2026. The rule at issue is the one that prevents a
company from short-circuiting drug development by selling an investigational compound as a
supplement; whether it was correctly applied here is contested, and the situation may have changed
since.

The sector has also generated substantial litigation between suppliers over patents and advertising
claims. None of that litigation has produced clinical evidence, and the commercial framing has done
real damage to the field's credibility: consumers encounter NAD+ as an established anti-aging measure
long before they encounter the trials.

## Risks

NAD+ precursors have been well tolerated in trials at the doses studied, with mild gastrointestinal
effects most commonly reported. Three theoretical concerns recur. High-dose nicotinamide consumes
methyl groups during clearance, raising questions about methyl-donor availability. NAD+ is required
for proliferation, and preclinical work has raised the possibility that supplementation could support
the metabolism of existing tumours; this has not been tested in humans and remains unresolved.
Finally, intravenous NAD+ infusions, sold by clinics, deliver a molecule that does not readily cross
cell membranes and have no controlled evidence behind them.

## Outlook

Three questions determine whether this class amounts to anything. Whether tissue NAD+ actually falls
in humans, measured properly and tissue by tissue. Whether raising it changes anything a person would
notice, which requires trials with functional endpoints rather than blood levels or shifts in an
[[epigenetic-clock|epigenetic clock]] reading. And whether the more direct approach, inhibiting the
enzymes that consume NAD+ and particularly CD38, works better than supplying more substrate.

The obstacle is the one that affects the whole field. Without accepted
[[aging-biomarkers|biomarkers of aging]], a trial powered on [[healthspan|healthspan outcomes]] takes
years and costs more than any supplement company will spend, and the incentive to run it disappears
once a product can be sold without it. That is the inverse of the problem facing
[[metformin|generic drugs like metformin]], where the trial is worth running and nobody profits from
it.

Until such a trial exists, NAD+ precursors sit in the same category as several other popular
geroprotectors: a coherent mechanism, an enormous consumer market, and, measured against
[[exercise-and-aging|exercise]], no demonstrated effect on anything that matters. The
[[geroscience-hypothesis]] does not require every plausible mechanism to work, and this one has had a
long run without delivering.

## See also

- [[hallmarks-of-aging]]
- [[mitochondrial-dysfunction]]
- [[caloric-restriction]]
- [[senolytics]]
- [[david-sinclair]]
- [[aging-biomarkers]]
- [[exercise-and-aging]]
- [[healthspan]]

## References

[^burnett2011]: `paper` Burnett, C. et al. "Absence of effects of Sir2 overexpression on lifespan in C. elegans and Drosophila." *Nature*, 2011.
[^trammell2016]: `paper` Trammell, S.A.J. et al. "Nicotinamide riboside is uniquely and orally bioavailable in mice and humans." *Nature Communications*, 2016. {A pharmacokinetic study: it establishes that blood NAD+ rises after a dose, not that anything downstream of NAD+ changes.}
[^martens2018]: `paper` Martens, C.R. et al. "Chronic nicotinamide riboside supplementation is well-tolerated and elevates NAD+ in healthy middle-aged and older adults." *Nature Communications*, 2018. {The stated aims were tolerability and NAD+ elevation; the blood-pressure and arterial-stiffness findings are secondary and hypothesis-generating.}
[^elhassan2019]: `paper` Elhassan, Y.S. et al. "Nicotinamide riboside augments the aged human skeletal muscle NAD+ metabolome and induces transcriptomic and anti-inflammatory signatures." *Cell Reports*, 2019.
[^yoshino2021]: `paper` Yoshino, M. et al. "Nicotinamide mononucleotide increases muscle insulin sensitivity in prediabetic women." *Science*, 2021. {Muscle insulin sensitivity was measured by clamp rather than by a surrogate index, but the trial was small and most other endpoints did not move.}

==============================================================================
ARTICLE: nanoparticle-diagnostics
TITLE:   Nanoscale diagnostics
PORTAL:  Nanomedicine
URL:     https://futurehumanwiki.com/wiki/nanoparticle-diagnostics
SOURCE:  https://futurehumanwiki.com/raw/nanoparticle-diagnostics
==============================================================================

---
title: "Nanoscale diagnostics"
slug: "nanoparticle-diagnostics"
type: "technology"
status: "emerging"
horizon: "present"
trl: 8
categories: ["nanomedicine"]
tags: ["nanomedicine", "diagnostics", "liquid biopsy", "cancer screening", "nanopore", "biomarkers"]
summary: "Detection methods that read single molecules or use nanoscale labels, including liquid biopsy, multi-cancer blood tests, and nanopore sensing, together with their overdiagnosis costs."
updated: "2026-07-27"
humanEvidence: "Tumour genotyping from blood is routine care, and a randomized trial used circulating tumour DNA to guide chemotherapy in stage II colon cancer; no multi-cancer blood test has been shown to reduce mortality in people."
access: "Lateral flow tests and tumour genotyping are routine clinical care; a cell-free DNA colorectal screen was approved in the United States in 2024, and multi-cancer tests are sold outside organised screening programmes rather than reimbursed by them."
reversibility: "reversible"
issues: ["The England multi-cancer screening trial is named and sized without a citation.", "Activity-based synthetic biomarkers are said to be in early clinical development without a source."]
---

```infobox
{
  "caption": "Class of diagnostic technology",
  "rows": [
    { "label": "Common formats", "value": "Liquid biopsy, nanopore, nanoparticle label" },
    { "label": "Ubiquitous example", "value": "Gold-nanoparticle lateral flow test" },
    { "label": "Typical target", "value": "Circulating tumour DNA" },
    { "label": "Signal fraction in early disease", "value": "Often below 0.1%" },
    { "label": "cfDNA colorectal screen approved", "value": "United States, 2024" },
    { "label": "Principal tradeoff", "value": "Overdiagnosis and false positives" },
    { "label": "Readiness", "value": "TRL 8–9 by format" }
  ]
}
```

**Nanoscale diagnostics** covers detection methods that operate on individual molecules or use nanoscale particles as labels and sensors. The category spans the mundane and the frontier: the coloured line on a home antigen test is a stripe of gold nanoparticles, while single-molecule sequencing through an engineered pore reads a DNA strand base by base. The clinically consequential branch as of 2026 is the blood test for fragments of tumour DNA, which is closer to changing cancer screening than any other technology in this wiki's nanomedicine section.

## Nanoparticle labels

The oldest working application uses nanoparticles as visible or magnetic reporters. Colloidal gold conjugated to antibodies aggregates at a capture line on a nitrocellulose strip, producing the red band of a lateral flow test; the format is cheap, needs no instrument, and was manufactured in billions of units during the COVID-19 pandemic. Superparamagnetic iron oxide particles serve as magnetic resonance contrast agents and, in one cleared platform, as the readout for detecting bloodstream pathogens by their effect on water relaxation times, avoiding the delay of culture.

Quantum dots — semiconductor nanocrystals with narrow, size-tunable emission — became standard in research imaging and multiplexed assays but have made little clinical progress, partly because the brightest formulations contain cadmium.

These are labels rather than sensors. The engineering that matters is surface chemistry: keeping particles from aggregating, attaching antibodies at the right density, and preventing non-specific binding. It is the same body of knowledge that governs [[targeted-drug-delivery]], applied to a much easier problem, because a particle in a test strip does not have to survive blood, evade phagocytes, or find an organ.

## Liquid biopsy

Cell-free DNA circulates in plasma in everyone. Tumours shed their own, and that circulating tumour DNA carries the mutations, copy-number changes, and methylation patterns of the cancer that produced it. Sampling it means sampling a tumour without a needle, repeatedly, and capturing heterogeneity a single core biopsy would miss.

The technical difficulty is signal fraction. In advanced disease tumour DNA may be a few percent of the total; in early-stage disease it is often below one part in a thousand, so detection requires deep sequencing, molecular barcoding to distinguish real variants from sequencing errors, and statistical models that aggregate weak evidence across many sites. Methylation-based methods have proved more sensitive than mutation-based ones for early detection, because a tumour genome carries far more differentially methylated positions than driver mutations, and the pattern also indicates the tissue of origin. The same methylation arrays and sequencing panels underlie the [[epigenetic-clock]] literature, which reads the marks for a different purpose and faces a much weaker validation standard.

Three applications have separated out. Genotyping advanced cancers to select targeted therapy is routine practice. Minimal residual disease testing after surgery detects recurrence months before imaging and has been used in randomised trials to decide who needs adjuvant chemotherapy, in one colon cancer study reducing chemotherapy use without worsening recurrence-free survival.[^tie2022] Screening asymptomatic people is the most consequential and the least settled.

## Multi-cancer early detection

Multi-cancer early detection tests look for a shared signal across dozens of cancer types in a single blood draw, most using methylation patterns. In a prospective study of several thousand adults, just over a third of positive signals corresponded to a confirmed cancer, with specificity above 99 percent.[^schrag2023] A specificity of 99 percent sounds excellent and is not, when the prevalence of undiagnosed cancer in a screened population is well under one percent: most positives in such a setting are false, and each one triggers imaging, biopsy, and months of uncertainty.

A single-cancer version cleared the regulatory bar first. A blood test for colorectal cancer approved in the United States in 2024 detected about 83 percent of cancers at 90 percent specificity in a large screening cohort, but only about 13 percent of advanced precancerous lesions.[^chung2024] That asymmetry matters: colonoscopy prevents cancer by removing adenomas, whereas a blood test that misses nearly nine in ten of them mostly finds disease that has already arrived. Its plausible value is in reaching people who decline colonoscopy at all.

The largest randomised trial of multi-cancer screening, enrolling roughly 140,000 people in England, was designed to measure whether the approach reduces late-stage diagnoses. As of 2026 no national screening programme has adopted a multi-cancer blood test.

> [!caution] The screening trap
> Finding cancer earlier is not the same as helping patients. Lead-time bias makes survival from diagnosis look longer whenever diagnosis moves earlier, even if death occurs at the same moment. Length bias means screening preferentially catches slow-growing tumours, which are the ones least likely to kill. Only a reduction in mortality, measured against an unscreened control arm, distinguishes benefit from statistical artefact.

## Overdiagnosis

Overdiagnosis is the detection of disease that would never have caused symptoms. It is not a false positive: the cancer is really there under the microscope. It would simply never have progressed.

The clearest natural experiment came from South Korea, where widespread ultrasound screening of the thyroid increased diagnosed incidence many times over while mortality from thyroid cancer stayed flat — the signature of finding disease that did not need finding.[^ahn2014] Most of those patients had their thyroid removed and took hormone replacement for life.

Blood-based multi-cancer tests could reproduce this at scale, or could avoid it. The optimistic argument is that a tumour shedding enough DNA to be detected is metabolically active and therefore more likely to be consequential, which would make the modality biased toward aggressive disease rather than indolent disease. The argument is plausible and unproven; the evidence that would settle it is mortality data from randomised trials that are due to report over the coming years.

## Single-molecule sensing

Nanopore sensing threads a molecule through a protein or solid-state pore a few nanometres wide and reads the resulting change in ionic current. The principle was demonstrated for polynucleotides in 1996 and became a commercial sequencing platform in the 2010s, giving very long reads from a device that fits in a hand and that has been used for genomic surveillance during outbreaks in field conditions.[^kasianowicz1996] Work on reading peptides and post-translational modifications by the same method has advanced considerably, and if it matures it would provide something proteomics currently lacks: single-molecule sequencing of proteins.

A different route uses [[crispr-cas9]] machinery as a detector. Collateral cleavage by certain Cas enzymes, triggered when a guide RNA finds its target, cuts a labelled reporter and produces a fluorescent or lateral-flow signal, allowing nucleic acid detection at low cost without thermal cycling.

Activity-based synthetic biomarkers invert the usual approach. Rather than looking for a molecule the body produces, they inject a nanoparticle carrying peptide substrates for disease-associated proteases; cleavage releases reporter fragments that concentrate in urine, amplifying a local enzymatic activity into an easily measured signal.[^kwong2013] The approach remains in early clinical development. It is the closest working thing to a diagnostic nanodevice: an injected particle that performs a computation of sorts in tissue and reports the answer, though without any of the autonomy the designs in [[medical-nanorobots]] assume. Structural containers built by [[dna-nanotechnology]] and the steerable devices of [[microrobots-in-medicine]] have both been proposed as sensing platforms, and neither has entered human testing.

## Limitations and outlook

Analytical sensitivity has outrun clinical interpretation. Detecting a hundred tumour DNA molecules in a tube of blood is now tractable; knowing what to do about them frequently is not, and clonal haematopoiesis — mutated blood cell clones that accumulate with age and shed the same variants tumours do — is a persistent source of confusing signals that also connects these assays to the biology of [[stem-cell-exhaustion]] and, more broadly, to the somatic mutation entry in the [[hallmarks-of-aging]].

Cost and access shape who benefits. A test priced for self-pay markets, delivered outside organised screening programmes and increasingly alongside the panels covered under [[consumer-blood-testing]], tends to reach people already well served by medicine, which is the pattern described in [[access-and-inequality]]. Insurance and employment consequences of a positive result raise the questions covered in [[genetic-discrimination]].

The most interesting unresolved application is not cancer at all. If the same sequencing and proteomic depth were applied longitudinally to healthy people, it might supply the surrogate endpoints that the field of [[aging-biomarkers]] has been unable to validate, put a measurement behind claims about [[biological-age]], and give trials in [[geroscience-hypothesis]] something to measure other than death. Whether a molecular signal read from plasma can track something as diffuse as biological ageing, rather than a discrete tumour, is a much harder question than early cancer detection, and no assay has yet answered it.

## See also

- [[targeted-drug-delivery]]
- [[dna-nanotechnology]]
- [[aging-biomarkers]]
- [[epigenetic-clock]]
- [[medical-nanorobots]]
- [[microrobots-in-medicine]]
- [[genetic-discrimination]]
- [[biological-age]]

## References

[^tie2022]: `paper` Tie, J. et al. "Circulating Tumor DNA Analysis Guiding Adjuvant Therapy in Stage II Colon Cancer." *New England Journal of Medicine*, 2022.
[^schrag2023]: `paper` Schrag, D. et al. "Blood-based tests for multicancer early detection (PATHFINDER): a prospective cohort study." *The Lancet*, 2023. {A single-arm cohort with no unscreened comparison group, so it reports how the test performed and cannot show whether screening changed any outcome.}
[^chung2024]: `paper` Chung, D.C. et al. "A Cell-free DNA Blood-Based Test for Colorectal Cancer Screening." *New England Journal of Medicine*, 2024.
[^ahn2014]: `paper` Ahn, H.S., Kim, H.J., Welch, H.G. "Korea's Thyroid-Cancer 'Epidemic' — Screening and Overdiagnosis." *New England Journal of Medicine*, 2014. {An analysis of national incidence and mortality trends rather than a trial; the overdiagnosis inference rests entirely on the divergence between the two curves.}
[^kasianowicz1996]: `paper` Kasianowicz, J.J., Brandin, E., Branton, D., Deamer, D.W. "Characterization of individual polynucleotide molecules using a membrane channel." *PNAS*, 1996.
[^kwong2013]: `paper` Kwong, G.A. et al. "Mass-encoded synthetic biomarkers for multiplexed urinary monitoring of disease." *Nature Biotechnology*, 2013. {The demonstration was in mouse disease models, not in patients.}

==============================================================================
ARTICLE: natasha-vita-more
TITLE:   Natasha Vita-More
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/natasha-vita-more
SOURCE:  https://futurehumanwiki.com/raw/natasha-vita-more
==============================================================================

---
title: "Natasha Vita-More"
slug: "natasha-vita-more"
type: "person"
status: "established"
horizon: "present"
categories: ["people"]
tags: ["transhumanism", "design", "cryonics", "human enhancement", "art"]
summary: "American designer and theorist who wrote the Transhumanist Arts Statement, designed the Primo Posthuman body concept, and led Humanity+ for much of the 2010s."
updated: "2026-07-27"
issues: ["Birth year and birth name in the infobox carry no source.", "The Reception section characterises critical posthumanist criticism without citing any of it."]
---

```infobox
{
  "caption": "Designer and transhumanist theorist",
  "rows": [
    { "label": "Born", "value": "1950, United States" },
    { "label": "Born as", "value": "Nancie Clark" },
    { "label": "Education", "value": "PhD, University of Plymouth, 2012" },
    { "label": "Known for", "value": "Primo Posthuman; Transhumanist Arts Statement", "link": "/wiki/posthuman" },
    { "label": "Field", "value": "Design; futures studies" },
    { "label": "Affiliation", "value": "Humanity+; University of Advancing Technology", "link": "/wiki/humanity-plus" },
    { "label": "Spouse", "value": "Max More", "link": "/wiki/max-more" }
  ]
}
```

**Natasha Vita-More** is an American designer, artist and theorist who has worked since the early
1980s on the aesthetics and embodiment of [[transhumanism]] — the question of what an enhanced human
body would actually look like and how a person would relate to it. She authored the Transhumanist
Arts Statement, designed the speculative whole-body concept Primo Posthuman, and served as chair
and then executive director of [[humanity-plus]]. She is one of the few figures in the movement to
approach enhancement as a design problem rather than an engineering or philosophical one.

## Career

Vita-More entered futurism through art and media rather than science. In the early 1980s she
produced video work on life extension and hosted a cable television programme in Los Angeles
covering emerging technology, aging research and futurist thought, at a time when the topics had
almost no mainstream outlet. She circulated the Transhumanist Arts Statement in 1982, which framed
artistic practice as a means of imagining and prototyping future forms of embodiment rather than
representing existing ones.[^tas1982]

She took a doctorate at the University of Plymouth in 2012 through its planetary collegium
programme, and has taught at the University of Advancing Technology in Arizona. She co-edited
*The Transhumanist Reader* (2013) with [[max-more]], to whom she is married; the volume remains the
standard anthology of the movement's primary texts.[^reader2013]

## Primo Posthuman

Primo Posthuman, first presented in 1997 and revised repeatedly since, is a conceptual design for a
future human body specified as a product might be: a labelled diagram with subsystems for
metabolism, sensory input, memory, error correction and replaceable components. It proposes
distributed data storage in place of a single vulnerable brain, sensory channels beyond the
biological range, skin capable of colour and texture change, and modular organs replaceable on
failure.

Nothing in it is buildable, and the design does not claim otherwise. Its function is to make the
implicit content of enhancement arguments visible. Philosophical discussions of
[[posthuman|posthumanity]] describe capacities — longer healthspan, greater cognitive range,
broader emotional repertoire — without specifying any body that has them; Primo Posthuman asks what
the specification looks like when someone tries to draw it. The resulting design surfaces the
questions the abstract arguments defer: where memory lives if it is distributed, what
[[personal-identity-and-continuity|continuity of identity]] means across modular replacement, and
what a person would find beautiful about a body they had specified themselves. Related themes run
through [[morphological-freedom]] and [[sensory-augmentation]].

> [!note] Design fiction, not a prototype
> Primo Posthuman belongs to a tradition of speculative design in which the artifact is an argument.
> Treating it as an engineering proposal — as some coverage has — misreads both the object and the
> claim.

## The C. elegans memory experiment

Vita-More's one experimental contribution is a study, published with Daniel Barranco in 2015, asking
whether learned information survives vitrification in a nematode.[^vmb2015] The worms were trained
using a standard olfactory imprinting protocol, cryopreserved by vitrification, revived, and then
tested for retention of the imprint. The trained animals retained the association after revival.

The result is real and modest, and it is worth stating what it does and does not show. *Caenorhabditis
elegans* has 302 neurons, is small enough for cryoprotectant to reach every cell, and is routinely
cryopreserved in laboratories; the learned behaviour is a simple chemosensory imprint, not an
episodic memory. The experiment establishes that vitrification need not erase a stored association
in a tiny nervous system. It says nothing directly about whether a human brain's structural
information survives the far harsher process of whole-body [[cryonics]], where perfusion is uneven,
ischemic time is long and cryoprotectant toxicity is the central unsolved problem. It is nonetheless
one of the few pieces of published experimental evidence anyone in the cryonics community has
produced, and it is cited in discussions of [[brain-preservation]] for that reason.

## Advocacy and organizational work

Vita-More chaired the board and later served as executive director of Humanity+, the successor to
the World Transhumanist Association founded by [[nick-bostrom]] and David Pearce, during the period
when the organization shifted from an activist body to a smaller publishing and conference
operation. Her positions within the movement have been consistently on the side of aesthetics,
access and diversity of forms — arguing that enhancement described only as capability increase
misses that people will choose bodies for expressive reasons, and that a movement composed largely
of men arguing about compute has a narrow view of what people want from their bodies.

## Reception and legacy

Academic engagement with her work comes mainly from design, art theory and media studies rather
than from bioethics or biology, which have largely ignored it. Critical posthumanist scholars,
whose tradition is separate from and often hostile to transhumanism, have criticized Primo Posthuman
as consumerist — a body reimagined as a serviceable product, with the market assumptions of its
period built in. Within the movement, her durable contribution is having insisted for four decades
that enhancement is about embodiment, and that a philosophy of the future which cannot describe a
body is incomplete.

That insistence has become more relevant, not less, as actual devices arrive. Implants,
[[neuroprosthetics]] and [[exoskeleton|powered exoskeletons]] are now designed by engineering teams
with no framework for the questions she raised about how a modified body should feel to inhabit and
to look at. Whether design practice can supply that framework before the devices standardize around
clinical convention is unresolved.

Her position also marks a real fault line inside the movement. Its dominant strand treats the body
as a temporary substrate to be escaped, with [[mind-uploading]] as the destination,
[[substrate-independence]] as the premise, and biology as an engineering problem to be routed
around. Vita-More's work assumes the opposite: that being embodied is not incidental to being a
person, and that a future worth wanting has to specify what the body is for rather than only what
replaces it. That disagreement is rarely
argued explicitly, which is part of why it persists.

The practical test is whether anyone builds to her specification. Primo Posthuman has never been
prototyped, and none of its components — the smart skin, the error-correcting organs, the
distributed backup of memory — exists at a readiness level that would let a designer specify them.
It remains a brief written for engineers who cannot yet accept it, which is either a failure of the
brief or an argument for writing briefs early.

## See also

- [[transhumanism]]
- [[posthuman]]
- [[max-more]]
- [[humanity-plus]]
- [[morphological-freedom]]
- [[cryonics]]
- [[extropianism]]
- [[human-enhancement]]

## References

[^tas1982]: `statement` Vita-More, N. "Transhumanist Arts Statement." 1982; revised in later editions. {A manifesto written by the subject of this article and revised repeatedly since, so it records a programme rather than documenting one.}
[^reader2013]: `book` More, M. and Vita-More, N. (eds.) *The Transhumanist Reader: Classical and Contemporary Essays on the Science, Technology, and Philosophy of the Human Future*. Wiley-Blackwell, 2013.
[^vmb2015]: `paper` Vita-More, N. and Barranco, D. "Persistence of Long-Term Memory in Vitrified and Revived *Caenorhabditis elegans*." *Rejuvenation Research*, 2015. {Rejuvenation Research is a small journal edited from within the life-extension community, and the result covers one nematode species with 302 neurons.}

==============================================================================
ARTICLE: negligible-senescence
TITLE:   Negligible senescence
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/negligible-senescence
SOURCE:  https://futurehumanwiki.com/raw/negligible-senescence
==============================================================================

---
title: "Negligible senescence"
slug: "negligible-senescence"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity"]
tags: ["aging", "comparative biology", "mortality", "naked mole-rat", "hydra", "biogerontology"]
summary: "The condition of organisms whose mortality rate and physiological function show no measurable decline with age after reaching maturity."
updated: "2026-07-28"
issues: ["The bowhead whale, rockfish and elephant TP53 claims carry no citations."]
---

```infobox
{
  "caption": "Concept in comparative biogerontology",
  "rows": [
    { "label": "Term coined by", "value": "Caleb Finch, 1990" },
    { "label": "Criterion", "value": "No detectable rise in mortality with age" },
    { "label": "Clearest cases", "value": "Hydra, naked mole-rat" },
    { "label": "Not equivalent to", "value": "Immortality" },
    { "label": "Field", "value": "Comparative biology, demography" },
    { "label": "Human relevance", "value": "Indirect; no mechanism transferred wholesale" }
  ]
}
```

**Negligible senescence** describes organisms in which the risk of death does not measurably increase with age after maturity, and in which reproductive output and physiological function show no age-related decline. The term was introduced by the biogerontologist Caleb Finch in 1990.[^finch1990] It is a statistical claim about mortality curves, not a claim of immortality: animals with negligible senescence still die, from predation, disease, injury and starvation, and their populations still turn over.

## What the term means

Most animals follow something close to a Gompertz curve, in which the hazard of death rises roughly exponentially with age. Negligible senescence is the case where the slope of that curve is indistinguishable from zero. Mortality becomes age-independent: an individual's chance of dying next year is the same at ten as at a hundred, and survivorship declines exponentially rather than sigmoidally.

Finch's original criteria were three: no observed increase in mortality rate with age, no decline in reproductive capacity, and no measurable deterioration in physiological function. Later authors have added a fourth in practice — that the claim be supported by a cohort large enough and followed long enough to detect a slope if one existed.

> [!note] Terminology
> "Negligible" is a statement about measurement, not about biology. It means no aging effect has been detected at the available statistical power, which is not the same as no aging effect existing. Several species once described as negligibly senescent have shown age-related change under closer study.

## Documented cases

```compare
{
  "columns": ["Hydra", "Naked mole-rat", "Greenland shark", "Ocean quahog"],
  "rows": [
    { "label": "Longest documented age", "values": ["No natural death observed in lab cohorts", "More than 30 years", "Estimated at a few centuries", "Estimated at about five centuries"] },
    { "label": "Evidence type", "values": ["Cohort mortality over 8 years", "Thousands of captive lifespan records", "Eye-lens radiocarbon dating", "Shell growth-ring counting"] },
    { "label": "Mortality rise with age", "values": ["Not detected", "Not detected", "Not measured", "Not measured"] },
    { "label": "Proposed mechanism", "values": ["Continuous stem cell renewal", "Cancer resistance, protein quality control", "Extremely slow metabolism and growth", "Oxidation-resistant membranes"] },
    { "label": "Main caveat", "values": ["Asexual polyps only", "Captive colonies, protected environment", "Age estimates carry wide error", "Single specimen sets the record"] }
  ]
}
```

**Hydra** is the strongest case. A cohort study following more than two thousand individuals for eight years found mortality and fertility both constant with age, with no sign of decline in either.[^schaible2015] The animal continuously replaces all of its cells from three stem cell lineages, so no somatic tissue persists long enough to accumulate damage. Hydra of species that reproduce sexually do senesce, which suggests the trait is a consequence of the reproductive mode rather than an intrinsic property of the genus.

**[[naked-mole-rat|Naked mole-rats]]** live roughly ten times longer than similarly sized rodents and show no increase in hazard with age across thousands of captive records extending past thirty years.[^ruby2018] They are exceptionally resistant to spontaneous tumors, though not entirely immune — cancers have been documented in captive animals. Their tissues produce an unusually high-molecular-mass form of hyaluronan that triggers early contact inhibition in cultured cells and has been implicated in that resistance.[^tian2013] Mice engineered to carry the naked mole-rat version of the responsible synthase gene showed reduced tumor incidence and a modest lifespan extension, one of the few instances in which a comparative finding has been transferred to a laboratory model.

**Very long-lived vertebrates and molluscs** — the Greenland shark, aged by radiocarbon in the eye lens to a few centuries;[^nielsen2016] the bowhead whale, whose age has been inferred from harpoon fragments and lens protein racemization; rougheye rockfish; the ocean quahog specimen named Ming — establish extreme longevity but not negligible senescence, because nobody has assembled an age-structured mortality curve for them. Extreme lifespan and a flat hazard are different claims and the literature often conflates them.

**Turtles and tortoises** occupy the middle ground. Comparative analyses published in 2022 covering many ectothermic tetrapod species found several testudines with aging rates near zero, alongside others that senesce normally, and used the contrast to test evolutionary theories of aging.[^dasilva2022]

## Why it is hard to establish

Proving a null is the field's chronic problem, and four difficulties recur.

Cohorts are small. Detecting a shallow Gompertz slope requires either very many individuals or very long observation, and for a 200-year animal nobody has either.

Captivity confounds. Naked mole-rat data come from protected colonies with veterinary care, which removes exactly the extrinsic mortality that would otherwise mask or mimic a trend. Wild data for the same species barely exist.

Ages are estimated, not recorded. Radiocarbon dating of lens tissue, growth-ring counts and racemization all carry substantial error, and the error grows with the age being estimated.

Selective reporting inflates records. The oldest individual of any species is by definition an outlier, and outliers attract publication. The same one-directional error that contaminates human [[maximum-human-lifespan|longevity records]] operates here.

## What comparative biology has taught

The most durable lesson is evolutionary. Aging rates track the risk of dying from something else: species with protected niches — subterranean, flying, armored, very large, or chemically defended — consistently age more slowly, as the classical theory of Medawar, Williams and Hamilton predicts. Selection cannot maintain function past the age at which most individuals are already dead, so lowering extrinsic mortality is what permits long life to evolve.

The second lesson is that the mechanisms are not shared. Elephants carry extra copies of the tumor-suppressor gene *TP53*; naked mole-rats use hyaluronan chemistry; bowhead whale genomes show duplications in DNA-repair genes; long-lived rockfish species differ from short-lived congeners in DNA repair and insulin-signaling pathways. Each lineage solved the same problem differently, which means there is no single mechanism to copy.

The third is negative. Long-lived species do not lack the [[hallmarks-of-aging]]. They have [[telomeres-and-telomerase|telomeres]], [[mitochondrial-dysfunction|mitochondria]], senescent cells and proteostatic burdens like everything else, and their longevity comes from better maintenance of each — higher [[autophagy|autophagic flux]], more accurate translation, more aggressive DNA repair — rather than from absence of the underlying processes. This is why comparative work supports the general claim that aging is modifiable while offering little in the way of a specific intervention.

## Relevance to human aging

The concept lends its name to [[aubrey-de-grey]]'s programme, Strategies for Engineered Negligible Senescence, pursued by the [[sens-research-foundation]]: the goal is to produce in humans by repeated intervention what these species achieve by construction. Nothing about the comparative data establishes that this is possible, but the existence of animals with flat mortality curves does undercut the argument that a rising hazard is a physical necessity.

Concrete transfers have been sparse. The naked mole-rat hyaluronan result is the clearest, and it produced a small effect in mice rather than a large one; moving such a gene into humans would be a [[gene-therapy-for-aging]] problem with all the delivery difficulties that entails. The broader claim that a single upstream process can be shifted to delay many diseases at once — the [[geroscience-hypothesis]] — draws indirect support from these species, since their advantage is general rather than disease-specific. Work on [[cellular-senescence]] in long-lived species, on [[proteostasis]] capacity in naked mole-rat tissue, on [[stem-cell-exhaustion|stem cell maintenance]] in hydra and on the regenerative capacity that connects to [[limb-regeneration]] in amphibians has generated candidate mechanisms and few candidate drugs. Companies including [[calico]] have run comparative programmes on naked mole-rats and killifish for over a decade without a clinical output.

## Outlook

The most useful thing negligible senescence provides is an existence proof and a set of measurement standards. Any claim that a human intervention has slowed aging has to be evaluated against the same criterion applied to hydra — a change in the slope of the mortality curve, not a change in an [[epigenetic-clock|epigenetic marker]] or a mean. Negligibly senescing species also make the [[healthspan]] point cleanly: their function does not decline before death, so the interval of frailty that dominates late human life is absent rather than shortened.

The unresolved question is whether flat mortality is a state that can be entered from a Gompertzian starting point, or only a state an organism can be built into. Hydra does not stop aging; it never starts, because it retains no long-lived somatic tissue. A human body is the opposite kind of system — long-lived post-mitotic neurons and cardiomyocytes that cannot be replaced without losing what they encode. Whether maintenance can substitute for replacement in tissue of that kind is the point on which the analogy either holds or fails.

## See also

- [[maximum-human-lifespan]]
- [[hallmarks-of-aging]]
- [[longevity-escape-velocity]]
- [[sens-research-foundation]]
- [[cellular-senescence]]
- [[limb-regeneration]]
- [[calico]]
- [[cynthia-kenyon]]

## References

[^finch1990]: `book` Finch, C.E. *Longevity, Senescence, and the Genome*. University of Chicago Press, 1990.
[^schaible2015]: `paper` Schaible, R. et al. "Constant mortality and fertility over age in *Hydra*." *PNAS*, 2015. {Eight years of laboratory observation under constant conditions; the flat hazard is measured within that window rather than over the animal's whole possible life.}
[^ruby2018]: `paper` Ruby, J.G., Smith, M. and Buffenstein, R. "Naked mole-rat mortality rates defy Gompertzian laws by not increasing with age." *eLife*, 2018.
[^tian2013]: `paper` Tian, X. et al. "High-molecular-mass hyaluronan mediates the cancer resistance of the naked mole rat." *Nature*, 2013.
[^nielsen2016]: `paper` Nielsen, J. et al. "Eye lens radiocarbon reveals centuries of longevity in the Greenland shark (*Somniosus microcephalus*)." *Science*, 2016. {Ages come from radiocarbon in eye lens nuclei and carry wide confidence intervals at the upper end; the study estimates lifespan, not a mortality curve.}
[^dasilva2022]: `paper` da Silva, R. et al. "Slow and negligible senescence among testudines challenges evolutionary theories of senescence." *Science*, 2022.

==============================================================================
ARTICLE: neural-correlates-of-consciousness
TITLE:   Neural correlates of consciousness
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/neural-correlates-of-consciousness
SOURCE:  https://futurehumanwiki.com/raw/neural-correlates-of-consciousness
==============================================================================

---
title: "Neural correlates of consciousness"
slug: "neural-correlates-of-consciousness"
type: "concept"
status: "emerging"
horizon: "present"
categories: ["minds"]
tags: ["consciousness", "neuroscience", "integrated information", "global workspace", "anaesthesia", "experiment"]
summary: "The minimal neural activity jointly sufficient for a specific conscious experience, and the experimental programme that tries to identify it."
updated: "2026-07-27"
issues: ["The 2023 open letter on whether IIT is testable is mentioned without a citation.", "Recurrent processing, higher-order and attention schema theories are stated without sources."]
---

```infobox
{
  "caption": "Research programme in cognitive neuroscience",
  "rows": [
    { "label": "Proposed as a programme", "value": "Crick and Koch, 1990" },
    { "label": "Standard definition", "value": "Chalmers, 2000" },
    { "label": "Leading theories", "value": "GNWT, IIT, RPT, higher-order" },
    { "label": "Key method", "value": "Contrastive and no-report paradigms" },
    { "label": "Adversarial test", "value": "Cogitate consortium, published 2025" },
    { "label": "Consensus", "value": "None on mechanism" }
  ]
}
```

**Neural correlates of consciousness** (NCC) are the neural events minimally sufficient for a
particular conscious experience. The concept was proposed as a deliberately modest research
programme: rather than solving the question of why physical processes give rise to experience at
all, find out which processes reliably accompany which experiences, and let the mechanism constrain
the theory. Francis Crick and Christof Koch put the programme on the agenda in 1990, proposing
synchronized gamma-band activity as a first candidate.[^crick1990] Three and a half decades later
the field has robust experimental paradigms, several well-specified theories, and no consensus on
which is right.

## Defining an NCC

David Chalmers's standard formulation distinguishes the *content* NCC — the minimal neural system
whose state determines what a person is conscious of — from the *background* or state NCC, the
conditions of arousal that make consciousness possible at all.[^chalmers2000] Brainstem structures
that regulate wakefulness are part of the second but plainly not the seat of visual experience.

Two qualifiers do real work. "Minimal" excludes the whole brain, which is trivially sufficient.
"Sufficient" is weaker than "identical to": a correlate can be a reliable accompaniment without
being the thing itself, which is why NCC results do not by themselves settle whether a machine or an
[[organoids|organoid]] is conscious. A further complication is that many candidate correlates are
consequences of consciousness rather than constituents — attention, working memory, and the
preparation of a verbal report all follow conscious perception and are recorded alongside it.

## Methods

The workhorse is the *contrastive* paradigm: hold the physical stimulus constant and vary whether it
is consciously perceived, then look for the difference. Binocular rivalry presents each eye with a
different image and lets perception alternate spontaneously. Masking renders a stimulus invisible by
flanking it in time. The attentional blink makes a target invisible when it follows another too
closely. Bistable figures, continuous flash suppression, and threshold detection tasks all serve the
same function.

The main confound is report. Asking a participant what they saw recruits decision-making, motor
preparation, and memory, and the resulting frontal activity may reflect reporting rather than
experiencing. No-report paradigms address this by inferring perceptual state from optokinetic eye
movements or pupil size, and they systematically reduce the frontal signature seen in report-based
studies.[^tsuchiya2015] How much of the classic prefrontal NCC survives is one of the field's live
disputes.

Measurement runs from scalp electroencephalography and functional MRI up to intracranial recording
in patients implanted for epilepsy monitoring, which supplies the field's highest-resolution human
data and uses the same hardware as [[ecog-interfaces]]. Analysis increasingly borrows the
multivariate classifiers developed for [[neural-decoding]], asking not whether a region is more
active but whether the identity of what a person saw can be read out from it. In animals, causal
tests using [[optogenetics]] can silence or drive candidate circuits, though animals cannot report
experience, which limits what those manipulations establish.

## The leading theories

**Global neuronal workspace theory**, developed from Bernard Baars's cognitive model by Stanislas
Dehaene and Jean-Pierre Changeux, holds that a stimulus becomes conscious when it is amplified and
broadcast to a distributed network of long-range pyramidal neurons, particularly in prefrontal and
parietal cortex. The signature is a nonlinear "ignition" event, with late widespread activity
following a threshold crossing.[^dehaene2011] On this account consciousness is about making
information globally available for report, reasoning, and action.

**Integrated information theory**, developed by Giulio Tononi, starts from the properties of
experience and asks what a physical system must be like to have them. It identifies consciousness
with a system's intrinsic, maximally irreducible cause-effect structure, quantified as Φ, and
locates the substrate in a posterior "hot zone" of parietal, temporal, and occipital cortex rather
than in prefrontal areas.[^koch2016] It makes the strong claim that a system's architecture, not its
input-output behaviour, determines whether it is conscious — the reason it denies
[[substrate-independence]].

**Recurrent processing theory** holds that local recurrent activity in sensory cortex suffices for
experience even without global broadcast. **Higher-order theories** hold that a first-order
representation becomes conscious only when represented by a further state, typically in prefrontal
cortex. **Attention schema theory** treats consciousness as the brain's simplified model of its own
attention. These are not merely verbal variants: they disagree about where to look and about which
patients and which systems are conscious.

## The adversarial collaboration

To break the pattern of each laboratory confirming its own theory, a large preregistered project
was organized in which proponents of global workspace theory and integrated information theory
agreed in advance on predictions that would count against their positions. The Cogitate consortium
ran the resulting experiments across multiple sites using functional MRI, magnetoencephalography,
and intracranial recordings, and published the results in 2025.[^cogitate2025]

Neither theory came through cleanly. Content-specific information about what participants saw was
decodable from posterior cortex and sustained for the duration of the stimulus, which fits the
posterior emphasis; but the sustained long-range synchronization that integrated information theory
predicted was not observed. Conversely, prefrontal cortex carried some content information, but the
ignition-like response global workspace theory predicted at the end of a stimulus was not found.
Both camps have argued that the tested predictions were not core commitments of their theories — a
response the collaboration's design was intended to forestall, and which illustrates how hard it is
to falsify a theory of consciousness.

> [!debate] A bet, settled
> In 1998 Christof Koch wagered David Chalmers a case of wine that a clear neural correlate of
> consciousness would be identified within 25 years. Koch conceded in 2023. The concession was
> collegial and the science had advanced considerably; the point is that a leading experimentalist
> judged the specific target unmet.

## Measures without report

The clinically important products of the field do not depend on settling the theory. The
perturbational complexity index uses transcranial magnetic stimulation — one of the techniques
covered under [[non-invasive-neuromodulation]] — to perturb cortex and measures the spatiotemporal
complexity of the electroencephalographic response. It separates wakefulness and dreaming from
anaesthesia, deep sleep, and vegetative states in ways that behavioural examination
misses.[^casali2013] Measures of signal diversity in the same family rise rather than fall under
psilocybin and LSD, which is why the drugs studied for [[psychedelic-therapy]] also serve the field
as a way of perturbing conscious state rather than only of suppressing it. Task-based imaging can
detect covert awareness: some behaviourally unresponsive
patients modulate their brain activity on command, for example by imagining playing tennis. A large
multi-centre study published in 2024 found this cognitive-motor dissociation in roughly one in four
patients who showed no observable response at the bedside.[^bodien2024] For those patients a
[[brain-computer-interface]] is the only available communication channel, and central thalamic
[[deep-brain-stimulation]] has been investigated as a way to raise arousal.

These methods are correlational and calibrated against people who can confirm their experience,
which limits how far they extend to systems very unlike humans — the problem that dominates
[[machine-consciousness]].

## Implications for artificial systems

Whether a machine could be conscious depends on which theory is true, and the theories give opposite
answers. Global workspace theory is functionalist: build the right architecture, with a bottleneck
that selects and broadcasts information, and the system qualifies. Integrated information theory
holds that a digital computer simulating a conscious system would have negligible Φ, so behaviour is
irrelevant. Higher-order and attention-schema theories fall in between.

This is why current work on assessing artificial systems takes an indicator-property approach —
extracting the computational features that each theory says matter and checking which are present —
rather than applying a behavioural test. It is also why progress on the NCC bears directly on
[[mind-uploading]] and [[whole-brain-emulation]]: an emulation's moral and personal status depends
on a question the NCC programme was explicitly designed to bracket, as does the status of anything
built on the path toward [[artificial-general-intelligence]]. The status of products marketed as
[[digital-immortality]] does not turn on it, since a text model of a dead person makes no claim to
experience in the first place.

## Open problems

The programme's founding bracket is now its main limitation. Correlates do not distinguish
constituents from consequences, and no experiment currently on offer separates a mechanism that
produces experience from one that merely accompanies it. Adjacent difficulties compound this:
theories are formulated at different levels of description, so a result can be read as supporting
either; the field's stimuli are almost entirely visual, so the correlates found may be correlates of
seeing rather than of consciousness; and the tools remain coarse, since even
[[connectomics]] and dense recording sample far less than the relevant circuitry. Tension over
whether integrated information theory is testable at all — expressed in a 2023 open letter from a
large group of researchers — reflects a deeper unresolved question about what would count as
evidence.

## See also

- [[machine-consciousness]]
- [[substrate-independence]]
- [[mind-uploading]]
- [[whole-brain-emulation]]
- [[organoids]]
- [[connectomics]]
- [[neural-decoding]]
- [[personal-identity-and-continuity]]

## References

[^chalmers2000]: `book` Chalmers, D. J. "What is a Neural Correlate of Consciousness?" In T. Metzinger (ed.), *Neural Correlates of Consciousness: Empirical and Conceptual Questions*, MIT Press, 2000.

[^crick1990]: `paper` Crick, F. and Koch, C. "Towards a neurobiological theory of consciousness." *Seminars in the Neurosciences*, 1990. {A programmatic essay rather than an experiment; the specific proposal it advanced, gamma-band synchrony as the correlate, did not survive later work.}

[^tsuchiya2015]: `paper` Tsuchiya, N., Wilke, M., Frässle, S. and Lamme, V. A. F. "No-Report Paradigms: Extracting the True Neural Correlates of Consciousness." *Trends in Cognitive Sciences*, 2015.

[^dehaene2011]: `paper` Dehaene, S. and Changeux, J.-P. "Experimental and theoretical approaches to conscious processing." *Neuron*, 2011.

[^koch2016]: `paper` Koch, C., Massimini, M., Boly, M. and Tononi, G. "Neural correlates of consciousness: progress and problems." *Nature Reviews Neuroscience*, 2016.

[^cogitate2025]: `paper` Cogitate Consortium. "Adversarial testing of global neuronal workspace and integrated information theories of consciousness." *Nature*, 2025. {Proponents of both theories fixed the predictions before data collection, which is why the later claims that those predictions were not core commitments arrived after the result.}

[^casali2013]: `paper` Casali, A. G. et al. "A theoretically based index of consciousness independent of sensory processing and behavior." *Science Translational Medicine*, 2013.

[^bodien2024]: `paper` Bodien, Y. G. et al. "Cognitive Motor Dissociation in Disorders of Consciousness." *New England Journal of Medicine*, 2024. {Detection requires the patient to sustain a mental task on command, so a negative result establishes nothing about whether that patient is aware.}

==============================================================================
ARTICLE: neural-decoding
TITLE:   Neural decoding
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/neural-decoding
SOURCE:  https://futurehumanwiki.com/raw/neural-decoding
==============================================================================

---
title: "Neural decoding"
slug: "neural-decoding"
type: "concept"
status: "established"
horizon: "present"
categories: ["cybernetics", "minds"]
tags: ["bci", "neural decoding", "machine learning", "fmri", "neural manifolds", "mental privacy"]
summary: "The inference of a stimulus, intention, or mental state from measured neural activity, and the statistical machinery that makes brain–computer interfaces work."
updated: "2026-07-27"
issues: ["Image reconstruction from fMRI with diffusion models is discussed without a citation."]
---

```infobox
{
  "caption": "Method in computational neuroscience",
  "rows": [
    { "label": "Field", "value": "Computational neuroscience" },
    { "label": "Inverse of", "value": "Encoding models" },
    { "label": "Classic result", "value": "Population vector, 1986" },
    { "label": "Signal sources", "value": "Spikes, LFP, ECoG, EEG, fMRI" },
    { "label": "Typical methods", "value": "Kalman filters, RNNs, transformers" },
    { "label": "Main obstacle", "value": "Non-stationarity across days" }
  ]
}
```

**Neural decoding** is the inference of something outside the brain — a stimulus, a movement, a
word, a state — from measurements of activity inside it. It is the statistical half of every
[[brain-computer-interface]], and the half where most recent progress has come from. Decoding is
formally the inverse of encoding, which asks how a neuron or population responds to a given input;
decoding asks what input or intention best explains an observed response.

The word "decoding" invites a misleading picture of a fixed cipher waiting to be broken. In
practice a decoder is a model fitted to one person's brain, on one day, for one task, and its
performance degrades as any of those conditions changes. What decoding recovers is what varied
systematically in the data it was trained on, and nothing else.

## Origins

Motor decoding began with the observation that neurons in primary motor cortex are broadly tuned to
the direction of arm movement, each firing most for a preferred direction and less for others. No
single cell specifies direction, but a weighted vector sum across a population does — the population
vector, described in the mid-1980s.[^georgopoulos1986] That result established both the possibility
of extracting a continuous variable from spike counts and the population-level framing the field
still uses.

Sensory decoding developed in parallel in human neuroimaging. Pattern-classification methods applied
to functional MRI showed that the category of an object a person was viewing could be read from
distributed activity in ventral temporal cortex,[^haxby2001] and later that the orientation of an
attended grating and even the content of binocular rivalry could be classified from voxel
patterns.[^kamitani2005] These demonstrations are the ancestors of every subsequent "mind-reading"
headline.

## From spikes to intent

Early [[utah-array]] decoders used linear filters mapping binned firing rates to cursor velocity.
The Kalman filter improved on this by treating the intended movement as a hidden state evolving
over time, with the neural data as noisy observations. A further gain came from recognising that
the training data are themselves flawed: during calibration the user is trying to correct the
cursor, so the true intention points toward the target rather than along the observed trajectory.
Retraining with that assumption produced a marked improvement in closed-loop
control.[^gilja2012]

Modern decoders are recurrent or transformer-based networks trained on many hours of data, often
pooled across sessions. For [[speech-neuroprosthesis]] systems they emit phoneme sequences that a
language model then resolves into text. For surface recordings in [[ecog-interfaces]] the input
features differ but the architecture does not. At the opposite extreme, a sixteen-electrode
endovascular device such as the [[stentrode]] supports only a binary classifier, and the decoding
problem shrinks accordingly — the design of a decoder is inseparable from the sensor feeding it.

Decoding is not confined to movement and speech. Closed-loop [[deep-brain-stimulation]] systems
decode a physiological marker of symptom state and adjust stimulation in response; seizure-warning
systems decode the approach of an ictal event; sleep and workload classifiers decode broad brain
states from a handful of channels. Hippocampal decoding underlies the [[memory-prosthesis]]
literature, and the sender's side of a [[brain-to-brain-interface]] is nothing more than a decoder
whose output happens to be routed to another person's stimulator.

### Population dynamics and manifolds

A shift in framing during the 2010s changed what decoders assume. Rather than treating each neuron
as encoding a variable, the dynamical-systems view treats motor cortex as a machine whose
population state evolves according to internal rules, with rotational structure that appears during
reaching regardless of the specific movement.[^churchland2012] Population activity turns out to
occupy a low-dimensional subspace — a neural manifold — of perhaps ten to a hundred dimensions,
whatever the number of recorded neurons.[^gallego2017]

This has two practical consequences. First, it explains the diminishing returns from higher channel
counts: additional electrodes sample the same manifold more densely rather than revealing new
independent signals, which is why the arms race in electrode numbers pursued by
[[neuralink]] and others has not translated proportionally into control performance. Second, it
offers a fix for drift.

## Calibration drift

A decoder trained on Monday works worse on Tuesday. Electrodes shift by micrometres, the set of
recorded neurons turns over, impedances change, and the user's own strategy changes as they learn.
The traditional response is to recalibrate at the start of every session, which costs minutes and
requires supervision — a serious obstacle to home use.

The manifold view suggests an alternative: the latent dynamics underlying a behaviour are far more
stable than the individual neurons expressing them, so a new day's recordings can be aligned to a
stored latent space rather than relabelled from scratch. Stabilizers built on this principle have
maintained performance across long gaps without new supervised calibration.[^degenhart2020] Related
work has shown that latent dynamics for a learned behaviour remain consistent over
years.[^gallego2020]

> [!key] Why drift is the practical bottleneck
> Peak decoding accuracy in a laboratory session is not what limits deployment. A device that needs
> a technician every morning cannot be a medical product. Most of the engineering distance between
> current research systems and an approved implant lies in making decoders that survive weeks
> unattended, not in raising a benchmark number.

## Non-invasive and semantic decoding

Decoding from outside the skull is far more limited, and the limits are informative. A widely
discussed 2023 study reconstructed the gist of continuous language from functional MRI while
subjects listened to stories, watched silent films, or imagined telling a story.[^tang2023] The
output was a paraphrase, not a transcript: it captured meaning while frequently getting exact words
wrong.

The conditions attached to that result are the important part. Each decoder required many hours of
training data from the specific individual, transferred poorly to other people, and could be
defeated by an uncooperative subject performing a distracting mental task. fMRI also requires
lying still inside a large magnet. The study's authors framed these findings as evidence that
mental privacy is currently protected by practical barriers rather than by principle — a framing
central to the policy discussion under [[mental-privacy]] and [[neurorights]].

Image reconstruction from fMRI using diffusion models produces striking pictures, but similarly
depends on tens of thousands of image-response pairs per subject, and the generative model supplies
a large share of the detail in the output. Distinguishing what the brain data contributed from what
the prior supplied is an active methodological problem.

## Open problems

Decoders recover trained categories. A speech decoder cannot output a language it never saw; a
movement decoder cannot produce a gesture absent from calibration. Whether richer or more abstract
content is recoverable at all from any practical measurement is unknown, and it is a different
question from whether that content is present in the signal.

Generalization across people is weak. Nearly every high-performing decoder is fitted to one
individual, and efforts to build cross-subject foundation models for neural data are early. Related
to this, the field lacks agreed benchmarks: results are typically reported on bespoke datasets from
single participants, which makes comparison across laboratories difficult.

Finally, decoding is silent about mechanism. That a variable can be read out of a population does
not establish that the brain uses it, a point that recurs in debates over the
[[neural-correlates-of-consciousness]] and in arguments about what a wiring diagram from
[[connectomics]] would and would not explain. Decoding accuracy is a lower bound on the information
present, not a description of how it is used.

## Outlook

The near-term trajectory is toward decoders that self-calibrate, transfer across sessions, and run
on the implant rather than on a laboratory computer. Those changes matter more for clinical
deployment than any further accuracy gain. Beyond that, the question that recurs in discussions of
[[whole-brain-emulation]] applies here in miniature: how much of a system's function can be
recovered from the measurements a practical device can make? For motor intention the answer has
turned out to be a surprising amount from surprisingly few neurons. Whether that generalizes beyond
motor variables is not established, and there is no strong theoretical reason to expect it to.

## See also

- [[brain-computer-interface]]
- [[speech-neuroprosthesis]]
- [[utah-array]]
- [[ecog-interfaces]]
- [[mental-privacy]]
- [[connectomics]]
- [[memory-prosthesis]]
- [[brain-to-brain-interface]]

## References

[^georgopoulos1986]: `paper` Georgopoulos, A. P., Schwartz, A. B. and Kettner, R. E. "Neuronal population coding of movement direction." *Science*, 1986. {Recorded in monkey motor cortex during reaching; the population vector is a way of reading direction out, not evidence that the brain computes one.}

[^haxby2001]: `paper` Haxby, J. V. et al. "Distributed and overlapping representations of faces and objects in ventral temporal cortex." *Science*, 2001.

[^kamitani2005]: `paper` Kamitani, Y. and Tong, F. "Decoding the visual and subjective contents of the human brain." *Nature Neuroscience*, 2005.

[^gilja2012]: `paper` Gilja, V. et al. "A high-performance neural prosthesis enabled by control algorithm design." *Nature Neuroscience*, 2012.

[^churchland2012]: `paper` Churchland, M. M. et al. "Neural population dynamics during reaching." *Nature*, 2012.

[^gallego2017]: `paper` Gallego, J. A., Perich, M. G., Miller, L. E. and Solla, S. A. "Neural manifolds for the control of movement." *Neuron*, 2017.

[^gallego2020]: `paper` Gallego, J. A. et al. "Long-term stability of cortical population dynamics underlying consistent behavior." *Nature Neuroscience*, 2020.

[^degenhart2020]: `paper` Degenhart, A. D. et al. "Stabilization of a brain–computer interface via the alignment of low-dimensional spaces of neural activity." *Nature Biomedical Engineering*, 2020. {Demonstrated in monkeys with chronically implanted arrays performing a trained task, not in a person using a device at home.}

[^tang2023]: `paper` Tang, J., LeBel, A., Jain, S. and Huth, A. G. "Semantic reconstruction of continuous language from non-invasive brain recordings." *Nature Neuroscience*, 2023. {Output was scored by similarity of meaning rather than word-level accuracy, and each decoder was fitted to one cooperating individual over many scanner hours.}

==============================================================================
ARTICLE: neural-dust
TITLE:   Neural dust and ultrasonic implants
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/neural-dust
SOURCE:  https://futurehumanwiki.com/raw/neural-dust
==============================================================================

---
title: "Neural dust and ultrasonic implants"
slug: "neural-dust"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 4
categories: ["cybernetics", "nanomedicine"]
tags: ["implants", "ultrasound", "wireless", "miniaturization", "bci", "neural recording"]
summary: "A proposed class of millimetre-scale wireless implants powered and read out by ultrasound, intended to replace tethered electrode arrays with distributed untethered sensors."
updated: "2026-07-27"
humanEvidence: "No human study of ultrasonic neural dust has been published; the in vivo work is in anaesthetized rats and covers peripheral nerve and muscle, not cortex, while related magnetoelectric stimulators have reached first-in-human study."
access: "Nothing to obtain: these are laboratory devices with no approved product, no commercial supplier, and no trial of an ultrasonic mote in a person."
reversibility: "difficult"
issues: ["The magnetoelectric first-in-human studies are described without a citation."]
---

```infobox
{
  "caption": "Class of miniature neural implant",
  "rows": [
    { "label": "Proposed", "value": "2013, UC Berkeley" },
    { "label": "Key figures", "value": "Seo, Maharbiz, Carmena, Rabaey, Alon" },
    { "label": "Power and readout", "value": "Ultrasonic backscatter" },
    { "label": "Demonstrated size", "value": "Millimetre-scale motes" },
    { "label": "First in vivo demonstration", "value": "2016, rat peripheral nerve" },
    { "label": "Cortical demonstration", "value": "None published as of 2026" },
    { "label": "Main obstacle", "value": "Skull attenuation; power budget" },
    { "label": "Readiness", "value": "TRL 4" }
  ]
}
```

**Neural dust and ultrasonic implants** are miniature, battery-free devices that draw their power
from an external ultrasound beam and report their measurements by modulating the sound they reflect
back. The concept was proposed in 2013 as a way past the central failure mode of implanted
recording: not the electrode itself but the wires, connectors, and skull penetrations that tether it
to the outside world. Despite the name, the demonstrated devices are millimetre-scale, not nanoscale,
and the published in vivo work is confined to peripheral nerve and muscle in rodents.

## The problem it addresses

Chronic neural recording fails at its interfaces. Penetrating arrays such as the
[[utah-array]] lose channels over months to years as tissue encapsulates the shanks, but the more
mundane failures are mechanical: a rigid array tethered to a skull-mounted connector transmits every
micromotion of the brain into the tissue, insulation on the wire bundle delaminates, and a
percutaneous pedestal is a permanent infection route. Scaling up channel count makes all of this
worse, because more channels mean more wires through the same skull opening.

The neural dust proposal inverts the architecture. Instead of one large array with many wires,
distribute many independent, sealed, wireless motes, each recording locally and each small enough
that the tissue response around it is minimal. There is nothing to tether and nothing to route. The idea belongs to a broader
push in [[neuroprosthetics]] toward fully implanted wireless systems, and it is the most aggressive
version of it.

## How ultrasonic backscatter works

Each mote contains a piezoelectric crystal, a pair of recording electrodes, and — in the simplest
designs — a single transistor. An external transducer sends ultrasound pulses; the crystal converts
the incident acoustic energy into a voltage, which powers the device. The local extracellular
potential at the electrodes modulates the transistor's conductance, which changes the electrical
load on the crystal, which changes how much of the incident ultrasound is reflected. The external
transceiver listens for the echo and reads the neural signal off its amplitude. Time-of-flight
distinguishes echoes from motes at different depths, so many devices can be interrogated with one
beam.

The scheme's elegance is that the mote needs no oscillator, no radio, no battery, and almost no
digital logic. Its cost is bandwidth: backscatter is a low-rate uplink, and the signal must compete
with reflections from tissue interfaces.

## Why not radio

The choice of ultrasound over electromagnetic power transfer follows from physics rather than
preference. At the frequencies where an antenna a millimetre across would radiate efficiently, the
wavelength in tissue is short and absorption is high, so the power that reaches an implanted device
is small and most of the transmitted energy heats the intervening tissue. Ultrasound at a few
megahertz has a wavelength on the order of a millimetre in soft tissue and attenuates far less per
centimetre, so a millimetre-scale piezoelectric receiver is well matched to the wavelength and the
safe acoustic intensity limits permit useful power delivery at depth.

Bone breaks this argument. The skull attenuates, reflects, and refracts ultrasound severely and
heats under it, which is why proposals for cortical neural dust place a sub-cranial transceiver — a
thin device implanted beneath the skull — that relays between the motes and an external unit. That
relay is itself a conventional implant requiring surgery, which removes much of the claimed
simplicity for brain applications and explains why every published in vivo result to date sits
outside the skull.

## Development history

```timeline
[
  { "year": "2013", "title": "Concept published", "text": "A Berkeley group proposes ultrasonically powered sub-millimetre motes as a route to chronic brain-machine interfaces, arguing that electromagnetic power transfer does not scale to that size in tissue." },
  { "year": "2016", "title": "First in vivo demonstration", "text": "Roughly millimetre-scale motes implanted on the sciatic nerve and gastrocnemius muscle of anaesthetized rats record nerve and muscle activity and report it by ultrasonic backscatter." },
  { "year": "2018–2020", "title": "Stimulation added", "text": "A millimetre-scale ultrasonically powered stimulator with bidirectional communication delivers current to peripheral nerve, showing the link can carry commands as well as data." },
  { "year": "2021", "title": "Distributed microimplant networks", "text": "A separate approach demonstrates dozens of sub-millimetre radio-frequency-powered chiplets recording from rodent cortex as a coordinated network, without ultrasound." },
  { "year": "2020s", "title": "Magnetoelectric and clinical spin-offs", "text": "Magnetoelectric films are used to power millimetre-scale stimulators, and companies begin first-in-human studies of pea-sized wireless stimulators for psychiatric indications." }
]
```

## Current state

Two published results anchor the field. The 2016 demonstration implanted millimetre-scale motes on
rat peripheral nerve and muscle and recovered electroneurogram and electromyogram signals
wirelessly.[^seo2016] A later device shrank an ultrasonically powered
*stimulator* to a few cubic millimetres with a bidirectional link, showing that the same channel can
deliver instructions as well as carry data.[^piech2020]

Neither has recorded single-neuron action potentials in cortex, which is the application the concept
was proposed for. The gap is substantial: extracellular spikes are tens of microvolts against a
noise floor that a passive backscatter link struggles to resolve, and they require kilohertz sampling
rather than the slow envelope signals that peripheral nerve recording can tolerate.

Parallel work reaches for the same goal by other routes. Networks of tens of sub-millimetre chiplets
powered by radio-frequency fields have recorded from rodent cortex as a coordinated
ensemble.[^lee2021] Magnetoelectric films, which convert an external magnetic field into a local
electric field, power millimetre-scale stimulators without the skull's acoustic penalty, and this
approach has moved fastest toward humans: pea-sized wireless cortical stimulators for psychiatric
indications have entered first-in-human studies in the mid-2020s, aiming at indications currently
served by [[deep-brain-stimulation]] or by [[non-invasive-neuromodulation]] with far less hardware
than the former and more precision than the latter.

## Limits

Four constraints govern how far the approach can scale.

**Power.** A mote's harvested power scales with its cross-sectional area. Shrinking a device by a
factor of ten in linear dimension cuts available power by a hundred, while the amplifier and
digitizer needed for spike-band recording have a floor set by thermal noise. This is the reason
motes have not shrunk to the tens of microns the original name implies.

**Uplink bandwidth.** Backscatter carries little information per interrogation pulse. Recording many
channels at spike-band rates requires either many interrogations per second, which raises the
acoustic dose, or on-mote compression, which requires power the mote does not have.

**Addressing and registration.** With hundreds of motes, the interrogator must know where each one
is, and motes move as tissue moves. Localization by time-of-flight works for a handful and becomes
an inverse problem for many.

**Surgery does not disappear.** Distributing hundreds of motes through cortex requires inserting
them, which means either an injection needle or an insertion tool per mote, plus a sub-cranial relay.
The tissue trauma of many small insertions is not obviously less than that of one array, and it is
not well characterized.

A fifth consideration is not technical. A recording device with no wires, no external marker, and no
battery to replace is also a device that is difficult to audit, which sharpens the questions raised
under [[mental-privacy]] about who holds neural data and under what terms it is read.

> [!caution] Naming versus scale
> "Neural dust" invites comparison with the devices imagined under [[medical-nanorobots]], which are
> proposed to be a thousand times smaller and to move under their own power. The demonstrated motes
> are passive millimetre-scale electronics, closer in kind to a very small pacemaker than to anything
> nanoscale, and the mismatch between the name and the hardware has repeatedly caused the work to be
> reported as more advanced than it is.

## Outlook

The most defensible near-term application is peripheral: chronic, untethered monitoring or
stimulation of nerves in the abdomen, chest, or limbs, where there is no skull, the signals are
larger and slower, and the clinical need — closed-loop control of bladder function, inflammation, or
cardiac rhythm — is real. This is bioelectronic medicine rather than a
[[brain-computer-interface]], and it is where ultrasonic implants are most likely to reach patients.
It also overlaps with the untethered devices covered under
[[microrobots-in-medicine]], which are steered through the body rather than fixed in place, and with
the delivery-focused work in [[targeted-drug-delivery]].

The cortical version competes with approaches that are further along. Penetrating arrays behind a
fully implanted wireless can, as pursued by [[neuralink]] and others; surface arrays as used in
[[ecog-interfaces]]; and the endovascular route of the [[stentrode]] all address the tethering
problem without requiring a new power-transfer physics. Whether distributed motes offer enough
additional coverage to justify their difficulty depends on a question that
[[neural-decoding]] research has partly answered in the negative: information from motor cortex
saturates well before channel count does, so a thousand scattered sensors may not decode much better
than a hundred well-placed ones. The case for neural dust is stronger where broad, sparse sampling
across many regions matters more than dense sampling of one — the regime relevant to
[[whole-brain-emulation]] arguments and, more immediately, to any attempt at a
[[memory-prosthesis]] spanning distributed circuits.

## See also

- [[brain-computer-interface]]
- [[utah-array]]
- [[neuroprosthetics]]
- [[medical-nanorobots]]
- [[microrobots-in-medicine]]
- [[deep-brain-stimulation]]
- [[stentrode]]
- [[neurorights]]

## References

[^seo2016]: `paper` Seo, D. et al. "Wireless recording in the peripheral nervous system with ultrasonic neural dust." *Neuron*, 2016. {The motes sat on sciatic nerve and calf muscle in anaesthetized rats and recorded compound signals; nothing was placed in brain and no single neuron was resolved.}

[^piech2020]: `paper` Piech, D. K. et al. "A wireless millimetre-scale implantable neural stimulator with ultrasonically powered bidirectional communication." *Nature Biomedical Engineering*, 2020.

[^lee2021]: `paper` Lee, J. et al. "Neural recording and stimulation using wireless networks of microimplants." *Nature Electronics*, 2021. {These chiplets are powered by radio frequency rather than ultrasound, so the result shows distributed wireless recording works without validating the acoustic route.}

==============================================================================
ARTICLE: neuralink
TITLE:   Neuralink
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/neuralink
SOURCE:  https://futurehumanwiki.com/raw/neuralink
==============================================================================

---
title: "Neuralink"
slug: "neuralink"
type: "organization"
status: "emerging"
horizon: "late 2020s"
categories: ["organizations", "cybernetics"]
tags: ["bci", "neurotechnology", "implants", "elon musk", "clinical trials", "paralysis"]
summary: "American neurotechnology company founded in 2016 developing a wireless implanted brain–computer interface with flexible electrode threads inserted by a surgical robot."
updated: "2026-07-27"
issues: ["Series E size and valuation are attributed to the company with no citation.", "Blindsight and CONVOY paragraphs rest on company statements not cited as footnotes."]
---

```infobox
{
  "caption": "Neurotechnology company",
  "rows": [
    { "label": "Founded", "value": "2016" },
    { "label": "Headquarters", "value": "Fremont, California" },
    { "label": "Co-founder", "value": "Elon Musk" },
    { "label": "Products", "value": "N1 implant, R1 surgical robot" },
    { "label": "First human implant", "value": "January 2024" },
    { "label": "Regulatory status", "value": "Investigational (FDA IDE)" },
    { "label": "Focus", "value": "Motor decoding, visual prosthesis" }
  ]
}
```

**Neuralink** is a privately held American company developing an implanted
[[brain-computer-interface]] consisting of a coin-sized wireless processor and a set of thin
polymer threads carrying electrodes, inserted into cortex by a purpose-built surgical robot. It was
founded in 2016 by a group including Elon Musk, who has been its public face and principal funder.
The company began implanting the device in humans in January 2024 under an FDA investigational
device exemption.

Neuralink's technical contribution is largely in engineering and surgery rather than neuroscience:
flexible electrode threads that move with the brain, automated insertion, high channel counts in a
sealed wireless package, and a hermetic implant with no percutaneous connector. Its scientific
results to date have been communicated mainly through livestreams and company blog posts rather
than peer-reviewed publication, which is the most common criticism levelled at it by academic BCI
researchers.

## Technology

The N1 implant houses a battery, wireless radio, and signal-processing electronics in a package
that sits in a recess milled into the skull, flush with the surrounding bone. From it extend 64
polymer threads bearing 16 electrodes each — 1,024 channels in total — each thread narrower than a
human hair and far more compliant than the rigid silicon shanks of a [[utah-array]]. The design
intent is to reduce the micromotion-driven tissue damage that drives glial encapsulation and signal
loss in stiff arrays.

Because the threads are too floppy to insert by hand, the R1 robot uses a needle and optical
guidance to place each one individually, steering around surface vasculature. The company described
the architecture in a 2019 white paper based on rodent work.[^musk2019] Recording, spike detection,
and compression happen on the implant; decoded output is sent over Bluetooth to an external device.
Users control a cursor by attempting or imagining movement, with the mapping learned by a decoder
that is periodically recalibrated — the same basic paradigm of [[neural-decoding]] established in
academic laboratories two decades earlier.[^hochberg2006][^gilja2012]

The device is read-only in its current form. It records but does not stimulate, which places it in
a different category from the implanted neurotechnologies already in routine clinical use, such as
the [[cochlear-implant]] and [[deep-brain-stimulation]], which write signals in rather than
reading them out. Adding a stimulation channel is a stated goal and is required for the visual
programme, but the safety profile of chronic cortical stimulation is less well characterized than
that of chronic recording.

```timeline
[
  { "year": "2016", "title": "Company founded", "text": "Registered in California by Elon Musk and a group of neuroscientists and engineers; publicly reported in early 2017." },
  { "year": "2019", "title": "Threads and robot unveiled", "text": "A launch event and accompanying white paper describe polymer electrode threads, robotic insertion, and rodent recordings." },
  { "year": "2020", "title": "Pig demonstration", "text": "A livestream shows live neural recordings from an implanted pig, demonstrating a sealed wireless device rather than any new decoding capability." },
  { "year": "2021", "title": "Macaque cursor control", "text": "A monkey plays a cursor-based video game using the implant, replicating a capability shown in primate laboratories years earlier." },
  { "year": "2023", "title": "Human trial cleared", "text": "The FDA grants an investigational device exemption; the PRIME study opens recruitment for people with quadriplegia." },
  { "year": "2024", "title": "First human implants", "text": "The first participant receives the device in January; the company later discloses that many threads retracted from the cortex, and modifies its technique for subsequent surgeries." },
  { "year": "2024–2026", "title": "International expansion", "text": "Regulators in Canada, the United Kingdom, and the United Arab Emirates permit studies; the company reports additional participants and a robotic-arm extension study." }
]
```

## Human trials

The PRIME study enrols adults with quadriplegia from cervical spinal cord injury or amyotrophic
lateral sclerosis. The first participant, Noland Arbaugh, received an implant in January 2024 and
within weeks was using it to move a cursor, play chess and video games, and operate a computer for
long stretches daily.

In the weeks after that surgery a substantial fraction of the threads retracted from the cortex,
reducing the number of channels recording usable signal. Neuralink disclosed this in a blog post
several months later and reported recovering performance through changes to signal processing and
decoding rather than reimplantation.[^nlink2024] For the second participant the company said it
altered its technique, sculpting the skull surface to reduce the gap under the implant and
inserting threads deeper. Thread retraction is a specific instance of the general problem that
afflicts every penetrating array: the brain moves, pulses, and reacts to foreign material.

A companion study, CONVOY, extends control from a cursor to an assistive robotic arm, moving the
device from computer access toward the broader ambitions of [[neuroprosthetics]]. The company has
also said it intends to pursue speech decoding, a capability demonstrated by academic groups using
surface and penetrating arrays and covered under [[speech-neuroprosthesis]]. A separate
programme, Blindsight, aims to restore a crude visual percept by stimulating visual cortex directly;
it received an FDA breakthrough device designation in 2024 but had no published human results as of
early 2026. Cortical visual prostheses have a long and largely disappointing history, and the
resolution limits that constrained earlier efforts — described under [[retinal-implant]] — apply
with equal force here.

> [!caution] Company reporting, not trial reporting
> Participant counts, performance figures, and adverse-event descriptions for Neuralink's studies
> have come predominantly from company statements and social media. As of 2026 there is no
> peer-reviewed clinical publication from the PRIME study, which makes independent assessment of
> durability, complication rates, and decoder performance impossible.

## Funding and structure

Neuralink is funded by venture capital and by Musk personally. It raised a Series E in 2025 that
the company said totalled several hundred million dollars at a valuation in the billions, a figure
far above what any BCI company has justified on revenue, since none has an approved product. The
capital advantage is real: implant development, robot engineering, animal work, and multi-country
regulatory submissions are expensive in a way academic consortia such as BrainGate cannot match.

The founding team has largely dispersed. Several co-founders left within the first few years, and
one went on to co-found Precision Neuroscience, which pursues a thin-film surface approach closer
to [[ecog-interfaces]] than to penetrating threads. Competitors now span the invasiveness spectrum,
including [[synchron]] with its endovascular [[stentrode]], Paradromics with a high-channel
penetrating array, and Blackrock Neurotech.

## Reception

Assessments divide along a consistent line. Engineers credit the company with solving practical
problems the field had left unsolved: a fully implanted wireless device with no skull-mounted
pedestal, an insertion robot that makes hundreds of precise placements, and a manufacturing process
for polymer electrode arrays. Those are the bottlenecks between a laboratory demonstration and a
product.

Neuroscientists are more sceptical of the framing. The cursor control demonstrated in 2024 was
qualitatively similar to results published from academic implants more than a decade earlier, and
the product name "Telepathy" describes typing with a decoder, not thought transmission — an
overstatement of the kind examined under [[brain-to-brain-interface]]. Claims that the device will
one day enable symbiosis with artificial intelligence, discussed as [[human-ai-merger]], are
untested and rest on a bandwidth argument that the low-dimensional structure of cortical activity
does not obviously support.

The company has also faced sustained criticism over animal research. Reuters reported in 2022 that
federal investigators were examining animal-welfare practices following internal complaints, and a
separate inquiry concerned the transport of hardware removed from primate brains.[^reuters2022] An
advocacy organization filed complaints over primate experiments conducted at a university partner.
Neuralink has denied wrongdoing and published its own account of its animal-care programme.

A third strand of criticism concerns data. A device that streams cortical activity to a phone
creates a record of neural signals held by a private company, raising the questions taken up under
[[mental-privacy]] — not because a cursor decoder reads thoughts, but because the underlying
recordings support inferences the user never intended to make available.

## Outlook

The questions that will determine whether Neuralink matters are unglamorous. Do threads stay put
over five years? What fraction of participants develop infection or require explant? Can a decoder
run for weeks without a technician? Does the device work in a home, unsupervised, on a bad day? The
company's willingness to run multi-country studies suggests a push toward a pivotal trial, but no
implanted BCI from any developer has yet been approved for routine clinical use, and the regulatory
path for a permanent cortical implant in a non-fatal condition has no precedent to follow.

A separate question is what happens if the company stops. Implanted devices create obligations that
outlast product lines, an issue that the emerging [[neurorights]] frameworks have begun to name but
not yet resolved.

## See also

- [[brain-computer-interface]]
- [[stentrode]]
- [[utah-array]]
- [[speech-neuroprosthesis]]
- [[neural-decoding]]
- [[neuroprosthetics]]
- [[mental-privacy]]
- [[synchron]]

## References

[^musk2019]: `paper` Musk, E. and Neuralink. "An Integrated Brain-Machine Interface Platform With Thousands of Channels." *Journal of Medical Internet Research*, 2019. {Written by the company about its own platform; the recordings it reports are from rats, and no human implant existed when it was published.}

[^hochberg2006]: `paper` Hochberg, L. R. et al. "Neuronal ensemble control of prosthetic devices by a human with tetraplegia." *Nature*, 2006.

[^gilja2012]: `paper` Gilja, V. et al. "A high-performance neural prosthesis enabled by control algorithm design." *Nature Neuroscience*, 2012.

[^nlink2024]: `statement` Neuralink. "PRIME Study Progress Update." Company blog post, 2024. {The sole source for the thread-retraction episode and the recovery from it; none of it has been peer-reviewed or published as a clinical report.}

[^reuters2022]: `news` Levy, R. "Musk's Neuralink faces federal probe, employee backlash over animal tests." *Reuters*, 2022. {Based on internal documents and staff accounts; it reports that investigations were opened, which is not a finding that violations occurred.}

==============================================================================
ARTICLE: neuroprosthetics
TITLE:   Neuroprosthetics
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/neuroprosthetics
SOURCE:  https://futurehumanwiki.com/raw/neuroprosthetics
==============================================================================

---
title: "Neuroprosthetics"
slug: "neuroprosthetics"
type: "technology"
status: "emerging"
horizon: "present"
trl: 7
categories: ["cybernetics", "bodies"]
tags: ["neurotechnology", "implants", "sensory restoration", "paralysis", "neuromodulation", "assistive technology"]
summary: "The engineering field that builds devices interfacing directly with the nervous system to restore lost sensory, motor, cognitive, or autonomic function."
updated: "2026-07-27"
humanEvidence: "Cochlear implants, deep brain stimulators and spinal cord stimulators are in routine clinical use in large patient populations; every bidirectional cortical result rests on single-digit numbers of participants."
access: "Hearing, movement-disorder, pain and bladder devices are approved and reimbursed in high-income countries; cortical, spinal and sensory-feedback systems are available only inside research studies."
reversibility: "difficult"
issues: ["The late-2024 transcutaneous spinal stimulation clearance is unnamed and uncited.", "The withdrawal of retinal implant support is cited to no source."]
---

```infobox
{
  "caption": "Field of biomedical engineering",
  "rows": [
    { "label": "Scope", "value": "Sensory, motor, cognitive, autonomic" },
    { "label": "Most deployed device", "value": "Cochlear implant", "link": "/wiki/cochlear-implant" },
    { "label": "Interface sites", "value": "Cortex, deep nuclei, spinal cord, peripheral nerve" },
    { "label": "Field established", "value": "1960s–1970s" },
    { "label": "Main bottleneck", "value": "Writing information in, not reading out" },
    { "label": "Chronic failure mode", "value": "Foreign-body encapsulation" },
    { "label": "Readiness", "value": "TRL 9 (hearing) to TRL 3 (memory)" }
  ]
}
```

**Neuroprosthetics** is the engineering discipline concerned with devices that exchange signals
directly with the nervous system in order to replace a function the nervous system has lost. It
spans devices in routine clinical use — the [[cochlear-implant]], the spinal cord stimulator,
[[deep-brain-stimulation]] — and devices that exist only in single-patient research studies, such as
cortical arrays that restore a sense of touch. The unifying problem is not electronics but
interface: how to place a durable transducer against excitable tissue that reacts to its presence,
and how to encode information in a form neurons will accept.

## Scope and taxonomy

Four families are usually distinguished, though real devices increasingly cross the boundaries.

**Sensory prostheses** convert an external stimulus into neural activity. The cochlear implant and
the [[retinal-implant]] are the archetypes; vestibular implants for bilateral vestibular loss have
reached small human trials. These devices must solve an encoding problem, because the natural code
of the sense organ is not reproducible with electrodes.

**Motor prostheses** read intent and act on it. Cortical systems for people with paralysis fall
under [[brain-computer-interface]]; peripheral systems that read muscle activity fall under
[[myoelectric-prosthetics]]; functional electrical stimulation drives the user's own paralysed
muscles or spinal circuitry rather than an external actuator.

**Autonomic and organ-directed devices** are numerically the largest category and receive the least
attention. Sacral neuromodulation for bladder dysfunction, vagus nerve stimulation for epilepsy and
for rehabilitation after stroke, phrenic pacing for ventilator dependence, and spinal cord
stimulation for chronic pain together account for far more implants than every cortical device
combined.

**Cognitive prostheses** attempt to restore a computation rather than a signal path. The only
serious example is the hippocampal [[memory-prosthesis]], and it remains at the proof-of-concept
stage.

## How a neuroprosthesis works

Every device is a loop with four parts, and each part fails differently.

The *transducer* couples electrical or optical energy to tissue. Choices range from penetrating
microelectrode arrays such as the [[utah-array]], through surface arrays used in
[[ecog-interfaces]], to nerve cuffs, intrafascicular electrodes, and the endovascular approach of
the [[stentrode]]. Closer coupling means more information per channel and more tissue damage.

The *front end* amplifies microvolt signals in the presence of stimulation artefacts thousands of
times larger, and does so within a power budget that will not heat the surrounding tissue by more
than a fraction of a degree.

The *algorithm* converts signals into commands or converts sensor data into stimulation patterns.
On the read side this is [[neural-decoding]], which has improved faster than the hardware. On the
write side it is an encoding problem with no equivalent body of theory.

The *actuator* is a robotic limb, a synthesized voice, a stimulator on the user's own spinal cord,
or a pattern of current in a sensory nucleus. Systems that both read and write are called
bidirectional, and they are where the field's hardest problems live.

## Development history

```timeline
[
  { "year": "1957–1961", "title": "Direct stimulation of a sense", "text": "Djourno and Eyriès stimulate the auditory nerve of a deaf patient; William House begins implanting single-channel cochlear devices in Los Angeles." },
  { "year": "1968", "title": "Cortical visual prosthesis", "text": "Brindley and Lewin evoke patterned phosphenes from an electrode array over the occipital cortex of a blind volunteer." },
  { "year": "1987–1997", "title": "Deep brain stimulation becomes clinical", "text": "Chronic high-frequency thalamic stimulation for tremor moves from a Grenoble case series to regulatory approval." },
  { "year": "2004–2006", "title": "Human cortical motor decoding", "text": "A participant with tetraplegia in the BrainGate pilot controls a cursor and a prosthetic hand with a penetrating array." },
  { "year": "2014–2016", "title": "Touch written back in", "text": "Peripheral nerve cuffs in amputees and intracortical microstimulation of somatosensory cortex in a participant with spinal cord injury evoke localized tactile percepts." },
  { "year": "2018–2023", "title": "Spinal interfaces restore walking", "text": "Targeted epidural stimulation restores stepping after spinal cord injury; a 2023 system links a cortical implant to a spinal stimulator to give volitional control of walking." },
  { "year": "2020s", "title": "Commercial implants enter trials", "text": "Multiple companies run early feasibility studies of implanted interfaces, and adaptive sensing-and-stimulating devices reach the market for movement disorders." }
]
```

## Current state

The clinically settled devices restore hearing, suppress movement-disorder symptoms, manage pain,
and control bladder and bowel function. They are unglamorous, well reimbursed in wealthy countries,
and improving incrementally.

The research frontier is bidirectional restoration after paralysis and amputation. Three results
define it. Intracortical microstimulation of human somatosensory cortex produces tactile sensations
that a participant localizes to specific fingers of a paralysed hand, and adding that feedback
roughly halved the time needed to complete an object-transfer task with a robotic
arm.[^flesher2016][^flesher2021] Implanted electrodes in residual peripheral nerves evoke graded
touch in amputees and, combined with skeletal attachment through
[[osseointegration]], have supported prostheses used at home for years.[^ortizcatalan2020] And
epidural spinal stimulation, targeted to the dorsal roots that recruit specific leg muscle groups,
restores stepping in people with severe spinal cord injury.[^wagner2018] A cortical implant wired to
such a stimulator gave one participant volitional, thought-driven walking, and produced neurological
recovery that persisted with the system switched off.

Each of these involves single-digit numbers of participants. A device cleared in the United States
in late 2024 delivers non-invasive transcutaneous spinal stimulation to improve hand and arm
strength after cervical injury, and represents the first of this generation to reach the market.

## The feedback problem

Reading intent out of the nervous system is comparatively tractable, because a decoder can be
trained on labelled attempts and improved offline without touching the tissue. Writing information
in is harder for reasons that are structural rather than temporary.

An electrode excites every axon within its field indiscriminately, including axons of passage
belonging to unrelated circuits. Natural sensory input arrives as spatiotemporally patterned
activity across specific cell types; electrical stimulation produces synchronous activation of a
sphere of tissue, which the brain has never encountered and has no code for. Percepts evoked this
way are describable but unnatural: pressure, tingling, or a vague spot of light rather than texture,
temperature, or shape. Repeated stimulation drives adaptation, so amplitude must be modulated to
maintain a stable percept. And every stimulation pulse saturates the recording amplifiers, which is
why bidirectional devices need artefact-rejection schemes that discard part of the signal they were
built to collect.

Two approaches attack the specificity limit. Biomimetic stimulation shapes pulse trains to imitate
the dynamics of natural afferent firing rather than delivering constant-amplitude trains, which
improves the naturalness of evoked touch. [[optogenetics]] achieves genuine cell-type specificity in
animals but requires gene delivery and implanted light sources; no human neuroprosthesis uses it.

> [!key] Why this asymmetry matters
> A motor prosthesis that reads well and writes badly is still useful, because vision substitutes for
> missing touch. A sensory prosthesis has nothing to substitute for a bad write channel, which is
> why hearing and vision restoration have plateaued at low information rates while motor decoding
> has kept improving.

## Limitations and failure modes

Chronic implants fail in predictable ways. Penetrating electrodes provoke a foreign-body response:
microglia activate, astrocytes encapsulate the shank, and neurons retreat from the recording tip, so
signal yield declines over months to years. Insulation absorbs water and delaminates; thin metal
traces corrode; hermetic feedthroughs leak. Percutaneous connectors, still standard in academic
systems, are an infection route and tether the user to a laboratory.

Beyond the hardware, three limits recur. Decoders drift as the recorded population changes,
requiring recalibration. Almost all headline performance is measured with a technician present in a
controlled setting, and degrades outside it. And the number of independent control or perceptual
dimensions a person can use saturates well below the electrode count, for the same reason in every
modality: current spread makes nearby channels non-independent.

## Ethics and access

Implanted neurotechnology raises questions that ordinary medical devices do not. Recorded neural
activity supports inferences the user did not intend to disclose, the concern formalized under
[[mental-privacy]] and addressed legislatively under [[neurorights]]. Devices that modulate mood or
motivation raise questions about authorship of action. And a device implanted for therapy in a
person who cannot easily have it removed creates an asymmetric dependence on the manufacturer's
continued existence — an exposure demonstrated when retinal implant support was withdrawn from
several hundred patients.

Access follows wealth rather than need. Cochlear implantation rates differ by more than an order of
magnitude between high- and low-income countries, and every newer device is more expensive. The
distributional pattern examined in [[access-and-inequality]] is not hypothetical here; it is the
current state of the most successful neuroprosthesis ever built.

Where restoration shades into [[human-enhancement]] is a live boundary rather than a settled line.
Devices such as those catalogued in [[sensory-augmentation]] use the same interfaces to add
capacities rather than restore them, and speculative uses such as
[[brain-to-brain-interface]] communication borrow the vocabulary of clinical neuroprosthetics while
resting on far weaker demonstrations.

## Outlook

Two trends look likely to shape the next decade. Materials and packaging are improving faster than
algorithms — thin-film flexible arrays, carbon-fibre electrodes, and fully implanted wireless
systems address the specific failure modes that have limited chronic recording, and commercial
programmes at [[neuralink]], [[synchron]], and others are funding that engineering at a scale
academic laboratories never could. Miniaturized wireless nodes of the kind described under
[[neural-dust]] would change the placement calculus if they can be made to work at cortical depth.

The competing trajectory is biological. Regenerative approaches under
[[limb-regeneration]] and cell replacement therapies would remove the need for an interface
entirely in some indications, and external assistive devices such as the
[[exoskeleton]] avoid implantation altogether. Which path wins depends on an unresolved empirical
question: whether the write-channel problem yields to better encoding and better electrodes, or
whether artificial stimulation is permanently limited to coarse, unnatural percepts that the brain
can learn to use but never to interpret as its own.

## See also

- [[brain-computer-interface]]
- [[cochlear-implant]]
- [[retinal-implant]]
- [[deep-brain-stimulation]]
- [[myoelectric-prosthetics]]
- [[speech-neuroprosthesis]]
- [[memory-prosthesis]]
- [[non-invasive-neuromodulation]]

## References

[^flesher2016]: `paper` Flesher, S. N. et al. "Intracortical microstimulation of human somatosensory cortex." *Science Translational Medicine*, 2016. {One participant with a spinal cord injury; the percepts were localized to individual fingers, and the outcome is what he reported feeling rather than a measured functional gain.}

[^flesher2021]: `paper` Flesher, S. N. et al. "A brain-computer interface that evokes tactile sensations improves robotic arm control." *Science*, 2021.

[^ortizcatalan2020]: `paper` Ortiz-Catalan, M. et al. "Self-contained neuromusculoskeletal arm prostheses." *New England Journal of Medicine*, 2020. {A brief report on a small series of arm amputees using the system in daily life, with no control group and no randomization.}

[^wagner2018]: `paper` Wagner, F. B. et al. "Targeted neurotechnology restores walking in humans with spinal cord injury." *Nature*, 2018.

==============================================================================
ARTICLE: neurorights
TITLE:   Neurorights
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/neurorights
SOURCE:  https://futurehumanwiki.com/raw/neurorights
==============================================================================

---
title: "Neurorights"
slug: "neurorights"
type: "concept"
status: "emerging"
horizon: "present"
categories: ["society", "cybernetics"]
tags: ["law", "human rights", "privacy", "neurotechnology", "governance", "regulation"]
summary: "A proposed set of human rights covering mental privacy, cognitive liberty, and mental integrity, drafted in response to brain-recording and brain-stimulation technology."
updated: "2026-07-27"
issues: ["No statute, judgment or recommendation is cited directly; all three sources are journal articles.", "Chile's 2023 Supreme Court ruling and the 2025 UNESCO recommendation need sources."]
---

```infobox
{
  "caption": "Proposed human-rights category",
  "rows": [
    { "label": "Proposed", "value": "2017" },
    { "label": "Principal advocates", "value": "Ienca, Andorno, Yuste" },
    { "label": "First constitutional protection", "value": "Chile, 2021" },
    { "label": "US state statutes", "value": "From 2024" },
    { "label": "International instrument", "value": "UNESCO recommendation, 2025" },
    { "label": "Binding force", "value": "National law only" }
  ]
}
```

**Neurorights** are a proposed category of human rights protecting the brain and mental life against interference by neurotechnology, covering at minimum the privacy of neural data, freedom from unconsented modification of mental states, and control over one's own cognition. The concept was formulated in 2017 and has moved into law faster than almost any other proposal in technology governance — a constitutional amendment within four years, state statutes within seven, and an international recommendation within eight — largely before the technology it anticipates exists in deployable form.

## The proposed rights

Two overlapping formulations circulate.

Marcello Ienca and Roberto Andorno proposed four: **cognitive liberty**, the right to use or refuse neurotechnology; **mental privacy**, protection of neural information; **mental integrity**, freedom from unauthorised alteration of neural computation; and **psychological continuity**, protection of the sense of self against unconsented change.[^ienca2017]

The NeuroRights Foundation, founded by the Columbia neuroscientist Rafael Yuste out of the group that published a 2017 set of ethical priorities for neurotechnology in *Nature*, advocates five: mental privacy, personal identity, free will, fair access to mental augmentation, and protection from algorithmic bias.[^yuste2017] The fourth restates the [[human-enhancement]] question as an entitlement rather than a liberty and connects the framework to [[access-and-inequality]]; the fifth borrows from AI governance and fits awkwardly with the rest.

The two lists disagree on whether neurorights are new rights or applications of existing ones. Freedom of thought is already an absolute, non-derogable right under the International Covenant on Civil and Political Rights, and has almost never been litigated because until recently no technology could reach the thought itself.

> [!note] What the rights are protecting against
> Not one mechanism but three: reading neural signals without consent, writing to the brain without consent, and inferring mental content from data that is not obviously neural at all. The first is technically hard, the second is routine in medicine, and the third is the one already happening at scale.

## Why the proposal arrived when it did

Three developments converged. Implanted recording devices moved from laboratories to multi-year clinical use, described in [[brain-computer-interface]] and [[ecog-interfaces]]. Consumer electroencephalography headsets became cheap enough to sell as meditation and focus trainers, generating neural data outside any medical-privacy regime. And decoding methods improved to the point where research groups could reconstruct approximate meaning from brain activity, a body of work covered in [[neural-decoding]] and [[speech-neuroprosthesis]].

Stimulation supplies the second half of the concern. [[deep-brain-stimulation]] is an established therapy in which a device alters neural activity continuously, and a minority of patients report changes in mood, impulsivity, or the sense that their actions are their own. [[non-invasive-neuromodulation]] devices are sold direct to consumers with no clinical oversight. Mental integrity is not a hypothetical right when a manufacturer can push a firmware update to an implanted stimulator.

## Law enacted so far

**Chile** amended its constitution in 2021, adding to the article on physical and mental integrity a clause directing that scientific and technological development respect life and physical and mental integrity, and that the law regulate the requirements for activity involving neural data and brain activity. It is the first constitutional protection of its kind. In 2023 the Chilean Supreme Court applied it, ruling against a consumer EEG manufacturer in a case brought by a former senator and ordering the deletion of his neural data — the only judicial decision anywhere resolving a neurorights claim on the merits.

**US states** have legislated through data-protection statutes rather than rights instruments. Colorado amended its privacy act in 2024 to bring biological and neural data within the definition of sensitive data requiring consent; California added neural data to the sensitive-information category of its consumer privacy act in the same year; further states followed in 2025. The approach is narrower than Chile's — it regulates processing, not interference — and it applies only to companies meeting the statutes' size thresholds.

**International instruments** are non-binding. The OECD issued a recommendation on responsible innovation in neurotechnology in 2019, the first intergovernmental instrument on the subject. UNESCO adopted a Recommendation on the Ethics of Neurotechnology in late 2025, following the model of its earlier AI ethics recommendation: member states report on implementation, and nothing obliges them to implement.

## Criticism

The main objection is that neurorights are redundant. Jan Christoph Bublitz and others argue that existing rights to privacy, bodily and mental integrity, and freedom of thought already cover the described harms, that creating new rights dilutes the human-rights framework, and that the proposals are drafted around imagined capabilities rather than actual ones.[^bublitz2022] On this view the correct response to a company selling EEG headsets is enforcement of ordinary data-protection law, not a new constitutional category.

A second objection concerns definitions. "Neural data" has no settled meaning. If it covers only signals recorded directly from neural tissue, it excludes the inference channels that actually reveal mental states — eye movement, heart rate variability, keystroke dynamics, typing latency — and a company can escape the statute by using better sensors of the wrong kind. If it covers anything from which mental states may be inferred, it covers nearly all behavioural data and becomes unadministrable. Existing statutes have chosen the narrow definition.

A third is that the framing overstates what decoding can do. As [[mental-privacy]] sets out, the best non-invasive semantic decoding requires many hours of per-subject training data, a cooperative participant lying still in a scanner, and fails when the subject deliberately thinks about something else. Nothing resembling involuntary mind-reading exists. Advocates reply that rights are properly established before the capability arrives rather than after, which is a reasonable position but concedes the empirical point.

> [!debate] New rights or old ones enforced
> Advocates argue that neural data is categorically different because it is generated involuntarily and can reveal states the subject has not chosen to express. Critics reply that the same is true of a consumer [[wearable-health-sensors|heart-rate monitor]], and that treating the brain as special repeats the error diagnosed in genetic exceptionalism — see [[genetic-discrimination]].

## Enforcement difficulty

Even where law exists, three features make it hard to apply.

Consent is the mechanism most statutes rely on, and it is doing work it cannot bear. A user agreeing to a headset's terms of service cannot evaluate what will be inferable from their data in five years, because the inference depends on models not yet trained.

Jurisdiction is the second problem. Neural data collected by a device sold in one country is processed by a service operated in another and may be sold to a broker in a third. Chile's constitutional protection binds conduct in Chile; the company in its Supreme Court case was foreign.

Medical and consumer devices are governed by different regimes with different assumptions. An implanted stimulator is a regulated medical device subject to clinical oversight; a wearable making similar claims about attention or mood is often a consumer product. The gap is where most current activity sits, including workplace fatigue-monitoring headsets used in transport and mining.

## Open problems

The distributive right — fair access to augmentation — has attracted the least analysis and would be the hardest to make justiciable. No court has any basis for ordering a state to provide cognitive enhancement, and it is not obvious what a violation would look like.

The relationship to therapy is unsettled. A right against unconsented alteration of mental states, read literally, is in tension with involuntary psychiatric treatment, which every legal system permits under conditions. Neurorights instruments have not addressed how the two fit. Nor have they addressed devices that write to memory rather than read from it: the hippocampal systems described in [[memory-prosthesis]] would, if they worked well, raise a question about the authenticity of recall that no privacy framework is shaped to answer.

And the timing question remains open. Constitutional protections written for capabilities that do not exist risk either freezing the wrong definition into law or accumulating as symbolic provisions that courts decline to apply. Chile's court did apply its provision, to a consumer headset whose recordings contain little of what the amendment was meant to protect. Whether the framework will be available and correctly shaped when devices of the kind described in [[neuralink]] and [[synchron]] research reach scale is not yet possible to judge.

## See also

- [[mental-privacy]]
- [[brain-computer-interface]]
- [[neural-decoding]]
- [[deep-brain-stimulation]]
- [[genetic-discrimination]]
- [[non-invasive-neuromodulation]]
- [[morphological-freedom]]
- [[access-and-inequality]]

## References

[^ienca2017]: `paper` Ienca, M. and Andorno, R. "Towards new human rights in the age of neuroscience and neurotechnology." *Life Sciences, Society and Policy*, 2017.
[^yuste2017]: `paper` Yuste, R. et al. "Four ethical priorities for neurotechnologies and AI." *Nature*, 2017. {A comment piece signed by a working group of neuroscientists and ethicists, arguing for priorities; it reports no data and was not peer-reviewed as research.}
[^bublitz2022]: `paper` Bublitz, J.C. "Novel Neurorights: From Nonsense to Substance." *Neuroethics*, 2022.

==============================================================================
ARTICLE: newlimit
TITLE:   NewLimit
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/newlimit
SOURCE:  https://futurehumanwiki.com/raw/newlimit
==============================================================================

---
title: "NewLimit"
slug: "newlimit"
type: "organization"
status: "emerging"
horizon: "2030s"
categories: ["organizations", "longevity"]
tags: ["reprogramming", "transcription factors", "epigenetics", "biotech", "aging", "liver"]
summary: "A biotechnology company founded in 2021 to screen transcription-factor combinations that restore youthful function to aged cells without changing their identity."
updated: "2026-07-27"
issues: ["The 2024 public data release and the 2025 funding round are described without a source."]
---

```infobox
{
  "caption": "Private biotechnology company",
  "rows": [
    { "label": "Founded", "value": "2021" },
    { "label": "Founders", "value": "Brian Armstrong, Blake Byers" },
    { "label": "Chief scientist", "value": "Jacob Kimmel" },
    { "label": "Initial commitment", "value": "$105 million" },
    { "label": "Headquarters", "value": "South San Francisco, California" },
    { "label": "Approach", "value": "Transcription-factor screening" },
    { "label": "Lead cell type", "value": "Hepatocytes" }
  ]
}
```

**NewLimit** is a biotechnology company founded in 2021 by Brian Armstrong and Blake Byers to develop [[partial-reprogramming|epigenetic reprogramming]] into a therapy, using large-scale screens of transcription-factor combinations to find sets that restore youthful function in aged cells without converting them to a different cell type. It is smaller than [[altos-labs]] and unusual among longevity companies in publishing negative and intermediate results as it goes.

## Overview

The company's technical bet is that the canonical [[yamanaka-factors]] are a poor tool for rejuvenation. The four-factor set was selected by Takahashi and Yamanaka for its ability to drive cells all the way to pluripotency,[^takahashi2006] which is precisely the property that makes it dangerous in a living animal — cells that lose identity form teratomas. NewLimit's premise is that the space of transcription factors contains combinations that produce the rejuvenation without the dedifferentiation, and that finding them is a search problem rather than a mechanism problem.

That framing determines the company's structure. Most of its investment is in the screening platform: delivering combinatorial sets of factors into primary human cells, measuring the result with single-cell transcriptomics and functional assays, and using the data to select the next round.

## History

Armstrong, the chief executive of the cryptocurrency exchange Coinbase, and Byers, a venture investor with a bioengineering doctorate, announced NewLimit in 2021 with a personal commitment of $105 million. Jacob Kimmel, a computational biologist who had worked at [[calico]] on aging and machine learning, joined as a co-founder and heads research.

The company spent its first years building the platform rather than announcing a target, and published a public data release describing its screening results in liver cells. In 2025 it raised additional venture funding, reported at roughly $130 million, to extend the platform to further cell types and move toward a development candidate.

```timeline
[
  { "year": "2021", "title": "Founded", "text": "Brian Armstrong and Blake Byers commit $105 million to a company built around transcription-factor screening, with Jacob Kimmel leading research." },
  { "year": "2022–2024", "title": "Platform build", "text": "The company focuses on combinatorial delivery, single-cell readouts and functional assays in primary human hepatocytes and T cells rather than on a named indication." },
  { "year": "2024", "title": "Public data release", "text": "NewLimit publishes screening results and methodology openly, an unusual posture for a private longevity company." },
  { "year": "2025", "title": "Further funding", "text": "A venture round reported at roughly $130 million extends the platform and moves the company toward candidate selection." }
]
```

## Research programme

### The screening approach

Reprogramming in its original form uses four factors delivered together and continuously. NewLimit instead treats the identity of the factors, their combination, their dose and their duration as variables to be optimized against a functional readout. The combinatorial space is enormous — the human genome encodes well over a thousand transcription factors — so the practical question is how to sample it efficiently, which is where machine-learning models trained on the company's own screening data are applied. It is the search logic of [[ai-drug-discovery]] pointed at factor combinations rather than at small molecules.

The readouts matter as much as the search. A shift in an [[epigenetic-clock]] reading is cheap to measure and weakly informative; the company has emphasized functional assays instead, on the reasoning that a cell that looks younger by methylation but does not perform better is not a therapeutic result. This is the same measurement problem that constrains all claims about [[biological-age]].

### Hepatocytes

The company's lead cell type is the hepatocyte. The choice is pragmatic rather than sentimental: liver function declines measurably with age, primary human hepatocytes can be obtained and cultured, the liver is the tissue most readily reached by [[lipid-nanoparticles]] and by [[aav-vectors]], and there are age-associated liver conditions with recognized clinical endpoints. A rejuvenation therapy that only ever worked in the liver would still be a drug.

### Immune cells

A second programme targets T cells, on the reasoning that immune decline with age — thymic involution, accumulation of exhausted and senescent lymphocytes, and the [[inflammaging|chronic inflammation]] that accompanies them — drives a large share of age-related mortality through infection and cancer. Chronic inflammation was added as a distinct hallmark of aging in the 2023 revision of that framework, which reflects how central immune dysfunction has become to the field's account of aging.[^lopez2023] Restoring T-cell function is also one of the three functional domains that the [[xprize-healthspan]] competition scores.

> [!key] Why identity is the constraint
> The therapeutic target is a cell that is old but still the right kind of cell. Push reprogramming too far and a hepatocyte stops being a hepatocyte; push it too little and nothing changes. There is no theory that predicts where that boundary lies for a given factor set, which is why the field has resorted to empirical screening at scale.

## Funding and posture

NewLimit is funded by its founders and by venture investors rather than by philanthropy, and its capital is roughly a tenth of what Altos Labs raised. The company has been explicit that this is a deliberate constraint: a smaller budget forces earlier decisions about what to measure.

Its more distinctive choice is disclosure. NewLimit has published data releases and technical write-ups, including results that did not work, at a stage when most private biotechnology companies publish nothing. The stated rationale is that the field's central questions are pre-competitive and that public data attracts collaborators and scientific staff. The commercial cost of the policy is real, and whether it survives contact with a clinical programme is untested.

## Reception

Scientific reception has been cautiously positive, largely for the disclosure policy and for the focus on functional rather than clock-based endpoints. Researchers who are skeptical of reprogramming as a rejuvenation strategy have nonetheless noted that the transcription-factor search is a well-posed question that will produce a usable answer either way.

The standard criticisms of the field apply. Nothing NewLimit works on has been tested in a human. The rejuvenation effects that motivate the whole programme have been demonstrated in mice and in cultured cells, and the founding in vivo result came from a progeria model together with short-duration measures in normal animals rather than from lifespan extension in healthy wild-type mice.[^ocampo2016] Delivery of multiple transcription factors to a solid organ at a controlled dose remains unsolved, and is the same obstacle that has limited [[somatic-gene-therapy]] for decades.

A more specific criticism concerns founder attention. A company whose principal funder runs a large public company in an unrelated industry has a governance profile that some investors treat as a risk, though the same is true of [[retro-biosciences]] and of much of the sector.

## Outlook

The decisions that will reveal whether the approach works are near. Does NewLimit name a factor set and an indication? Does it show rejuvenation of a functional parameter — regenerative capacity, metabolic output, drug clearance — in aged primary human cells and then in a large animal? Does it publish, as it has said it will, the results that fail?

If the search succeeds, the outcome is likely to be modest and specific: a defined combination that improves one cell type's function, delivered locally, for one age-associated disease. That is a smaller claim than [[epigenetic-reprogramming]] rhetoric usually makes, and it is the version most likely to reach a patient. Whether such a therapy would generalize to the organism, which is what the [[geroscience-hypothesis]] requires, is a separate question that no cell-level screen can answer.

## See also

- [[partial-reprogramming]]
- [[altos-labs]]
- [[yamanaka-factors]]
- [[epigenetic-reprogramming]]
- [[calico]]
- [[retro-biosciences]]
- [[epigenetic-clock]]
- [[induced-pluripotent-stem-cells]]

## References

[^takahashi2006]: `paper` Takahashi, K. and Yamanaka, S. "Induction of Pluripotent Stem Cells from Mouse Embryonic and Adult Fibroblast Cultures by Defined Factors." *Cell*, 2006. {Mouse fibroblasts; the four factors were selected for their power to erase cell identity, which is the property NewLimit is trying to avoid.}
[^ocampo2016]: `paper` Ocampo, A. et al. "In Vivo Amelioration of Age-Associated Hallmarks by Partial Reprogramming." *Cell*, 2016. {Lifespan extension was in a progeria mouse model; in normally aging mice the paper reports short-term tissue measures, not longer life.}
[^lopez2023]: `paper` López-Otín, C., Blasco, M.A., Partridge, L., Serrano, M. and Kroemer, G. "Hallmarks of aging: An expanding universe." *Cell*, 2023.

==============================================================================
ARTICLE: nick-bostrom
TITLE:   Nick Bostrom
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/nick-bostrom
SOURCE:  https://futurehumanwiki.com/raw/nick-bostrom
==============================================================================

---
title: "Nick Bostrom"
slug: "nick-bostrom"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "society"]
tags: ["existential risk", "transhumanism", "artificial intelligence", "philosophy", "forecasting"]
summary: "Swedish philosopher who defined existential risk as a field, wrote Superintelligence, and formulated the simulation argument and the vulnerable world hypothesis."
updated: "2026-07-27"
issues: ["Every reference is by Bostrom; the reception and criticism section cites no critic directly.", "The 2024 closure of FHI and the 2023 mailing-list episode carry no source."]
---

```infobox
{
  "caption": "Philosopher",
  "rows": [
    { "label": "Born", "value": "10 March 1973, Helsingborg, Sweden" },
    { "label": "Nationality", "value": "Swedish" },
    { "label": "Education", "value": "PhD, London School of Economics, 2000" },
    { "label": "Known for", "value": "Existential risk; simulation argument", "link": "/wiki/existential-risk" },
    { "label": "Founded", "value": "Future of Humanity Institute (2005–2024)" },
    { "label": "Co-founded", "value": "World Transhumanist Association", "link": "/wiki/humanity-plus" },
    { "label": "Key book", "value": "Superintelligence (2014)" }
  ]
}
```

**Nick Bostrom** is a Swedish philosopher whose work established existential risk as a research
programme and whose 2014 book *Superintelligence* moved the risk from advanced artificial
intelligence into mainstream policy discussion. He founded and directed the Future of Humanity
Institute at Oxford from 2005 until its closure in 2024, and he is one of the small number of
academic philosophers who has taken transhumanist claims seriously enough to argue for them in
peer-reviewed venues.

## Career

Bostrom studied physics, computational neuroscience and philosophy across four institutions before
taking a doctorate at the London School of Economics in 2000 on anthropic reasoning — the problem
of what an observer may infer from the fact of their own existence. That thesis became
*Anthropic Bias* (2002), and the observation-selection machinery in it recurs throughout his later
work.

In 1998 he co-founded the World Transhumanist Association with the philosopher David Pearce; the
organization later became [[humanity-plus]]. In 2005 he founded the Future of Humanity Institute
within Oxford's Faculty of Philosophy, assembling an unusual group of philosophers, mathematicians
and computer scientists, among them [[anders-sandberg]], Toby Ord and, later,
[[eric-drexler]]. FHI produced work on [[whole-brain-emulation]], global catastrophic risk,
[[human-enhancement]] and AI alignment for nearly two decades. It closed in April 2024 after
prolonged administrative conflict with the faculty, including restrictions on hiring and
fundraising. Bostrom left Oxford and now runs the Macrostrategy Research Initiative.

## Key ideas

### Existential risk

Bostrom's 2002 paper defined an existential risk as one that would either annihilate Earth-originating
intelligent life or permanently and drastically curtail its potential, and classified such risks as
bangs, crunches, shrieks and whimpers.[^bostrom2002] The definition does real work: it separates
catastrophes that a civilization recovers from within centuries from those that foreclose the
future, and it makes the size of the loss depend on how large that foreclosed future could have
been — an argument he developed in "Astronomical Waste".[^waste2003] Engineered pandemics,
discussed here under [[dual-use-research]], are usually treated as the leading biological case. The
framework is now standard vocabulary; see [[existential-risk]].

### The simulation argument

The simulation argument is a trilemma rather than a claim that reality is simulated. Bostrom argues
that at least one of the following holds: almost no civilization reaches a stage capable of running
ancestor simulations; almost no such civilization chooses to run them; or almost all observers with
experiences like ours are simulated.[^sim2003] The argument depends on
[[substrate-independence]] — that a sufficiently detailed computational process could instantiate
conscious experience — which he flags explicitly as a premise rather than a conclusion.

### The vulnerable world hypothesis

Bostrom asks what follows if some technology, not yet invented, is easy to build and reliably
destructive: an "easy nukes" scenario in which civilization draws a black ball from the urn of
possible inventions.[^vwh2019] His conclusion is uncomfortable and he says so — that surviving such
a draw would require either extremely effective preventive policing or strong global governance —
and the paper functions partly as an argument for [[differential-technological-development]], the
sequencing of technologies so that defensive capability arrives before offensive.

### Enhancement and transhumanism

Bostrom has defended human enhancement in analytic terms. "In Defense of Posthuman Dignity" (2005)
answers Leon Kass, Francis Fukuyama and Jürgen Habermas by arguing that dignity does not attach to
a particular biological configuration.[^dignity2005] With Sandberg he formulated the "wisdom of
nature" heuristic, which asks why evolution has not already made a proposed improvement and treats
a good answer as a precondition for attempting it — an unusual concession to the
[[precautionary-principle]] from a pro-enhancement writer. "The Fable of the Dragon-Tyrant" (2005)
recasts aging as a monster society has rationalized into acceptance, and remains the most widely
circulated argument for treating [[longevity-escape-velocity|defeating aging]] as urgent.

> [!note] What the fable does and does not argue
> The Dragon-Tyrant is an argument against the "wisdom of acceptance" position on mortality, not a
> forecast. Bostrom does not claim aging is close to solved; the fable's point is that the moral
> case for trying does not depend on how close it is.

### Superintelligence

*Superintelligence: Paths, Dangers, Strategies* (2014) argues that a machine intelligence exceeding
human capability across the board would be difficult to control by default, because
capability and goal content vary independently, and because a wide range of final goals generate
the same instrumental subgoals of self-preservation, resource acquisition and resistance to
modification.[^super2014] The book gave AI safety a shared vocabulary — orthogonality, instrumental
convergence, the treacherous turn, decisive strategic advantage — and moved the subject from
mailing lists into governments. Its central scenario, a single system undergoing fast recursive
self-improvement in isolation, has aged less well than its vocabulary: the actual trajectory of
[[artificial-general-intelligence]] since 2022 has involved many large models improving gradually
and in public.

## Reception

Bostrom is cited across philosophy, AI policy and popular writing, and *Superintelligence*
influenced public statements by several technology executives and national AI strategies. Criticism
comes from three directions. Philosophers dispute the probability assignments underlying the
simulation argument and the anthropic reasoning it uses. AI researchers, including
prominent figures who consider present systems far from general intelligence, have argued that his
scenarios abstract away from how machine learning actually works. Critics of longtermism argue that
weighting vast hypothetical future populations distorts present priorities.

In 2023 a message Bostrom sent to an [[extropianism|extropian]] mailing list in 1996, containing a
racial slur, resurfaced. He apologized; the apology drew further criticism, and Oxford said it was
looking into the matter. The episode preceded FHI's closure, though the institute's difficulties
with the faculty were longstanding and predated it.

## Legacy

Two of Bostrom's constructions have outlived the debates that produced them: the definition of
existential risk, which now organizes an entire research and funding ecosystem, and the vocabulary
of AI alignment. *Deep Utopia* (2024) turns to the harder question left over — what people would do
with themselves in a world where technology has solved the instrumental problems, and whether a
"solved world" leaves any content to a human life. That question, not the risk arguments, is the
one the field has least equipment to answer.

## See also

- [[existential-risk]]
- [[technological-singularity]]
- [[whole-brain-emulation]]
- [[differential-technological-development]]
- [[transhumanism]]
- [[anders-sandberg]]
- [[posthuman]]
- [[leon-kass]]

## References

[^bostrom2002]: `paper` Bostrom, N. "Existential Risks: Analyzing Human Extinction Scenarios and Related Hazards." *Journal of Evolution and Technology*, 2002. {Published in a small transhumanist-affiliated journal rather than a mainstream philosophy venue; its standing comes from later citation, not from where it appeared.}
[^waste2003]: `paper` Bostrom, N. "Astronomical Waste: The Opportunity Cost of Delayed Technological Development." *Utilitas*, 2003.
[^sim2003]: `paper` Bostrom, N. "Are You Living in a Computer Simulation?" *Philosophical Quarterly*, 2003.
[^vwh2019]: `paper` Bostrom, N. "The Vulnerable World Hypothesis." *Global Policy*, 2019.
[^dignity2005]: `paper` Bostrom, N. "In Defense of Posthuman Dignity." *Bioethics*, 2005.
[^super2014]: `book` Bostrom, N. *Superintelligence: Paths, Dangers, Strategies*. Oxford University Press, 2014.

==============================================================================
ARTICLE: nir-barzilai
TITLE:   Nir Barzilai
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/nir-barzilai
SOURCE:  https://futurehumanwiki.com/raw/nir-barzilai
==============================================================================

---
title: "Nir Barzilai"
slug: "nir-barzilai"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "longevity"]
tags: ["aging", "centenarians", "genetics", "metformin", "clinical trials", "geroscience", "regulation"]
summary: "Israeli-American physician-scientist whose centenarian genetics work and proposed TAME trial aim to make aging itself an acceptable regulatory indication."
updated: "2026-07-27"
---

```infobox
{
  "caption": "Physician-scientist and geroscience advocate",
  "rows": [
    { "label": "Born", "value": "1955, Haifa, Israel" },
    { "label": "Nationality", "value": "Israeli-American" },
    { "label": "Education", "value": "MD, Technion, 1985" },
    { "label": "Field", "value": "Endocrinology; geroscience" },
    { "label": "Affiliation", "value": "Albert Einstein College of Medicine" },
    { "label": "Directs", "value": "Institute for Aging Research" },
    { "label": "Known for", "value": "Centenarian genetics; TAME", "link": "/wiki/metformin" },
    { "label": "Key book", "value": "Age Later (2020)" }
  ]
}
```

**Nir Barzilai** is an Israeli-American physician-scientist at the Albert Einstein College of
Medicine in New York, where he directs the Institute for Aging Research and leads a long-running
study of exceptionally long-lived Ashkenazi Jewish families. He is best known as the principal
designer of TAME, a proposed trial of [[metformin]] whose endpoint was
constructed to force a regulator to treat aging as something a drug can be approved to address. As
of 2026 TAME had not been funded at the scale it requires, had not enrolled, and had reported no
results.

## Overview

Barzilai's two bodies of work rest on different kinds of evidence. One is observational human
genetics: a cohort of centenarians and their children,
assembled to ask which inherited variants separate people who reach extreme age from people who do
not. The other is regulatory strategy: an argument that the [[geroscience-hypothesis]] cannot be
tested until an endpoint exists that a health authority will accept for a drug aimed at aging
rather than at a named disease. The genetics is descriptive and largely uncontested. The trial
built to settle the second question has never been funded at the scale it requires.

## Career

Barzilai took his MD at the Technion in Haifa and trained in internal medicine and endocrinology
before joining the faculty of the Albert Einstein College of Medicine in the Bronx. His early
laboratory work concerned insulin action, visceral fat, and the metabolic changes that accompany
aging in rodents, which is the route by which he arrived at the question of why some people escape
them. He is the founding director of Einstein's Institute for Aging Research and directs its Nathan
Shock Center of Excellence in the Basic Biology of Aging, one of a small number designated by the US
National Institute on Aging. He co-founded CohBar, a company formed to
develop [[mitochondrial-dysfunction|mitochondria-derived peptides]] as drugs, and holds advisory
roles across the sector. His 2020 trade book *Age Later* sets out the case for
treating aging as the shared upstream cause of the diseases of old age.[^agelater2020]

```timeline
[
  { "year": "Late 1990s", "title": "Longevity Genes Project begins", "text": "Barzilai's group at Albert Einstein College of Medicine starts recruiting Ashkenazi Jewish people aged 95 and over, their adult children, and age-matched controls." },
  { "year": "2003", "title": "Lipoprotein genotype reported", "text": "A CETP variant is found over-represented among the long-lived, whose carriers have larger lipoprotein particles. The association later fails to yield a drug." },
  { "year": "2011", "title": "Lifestyle comparison published", "text": "People in the cohort who passed 95 report drinking, exercise, diet and weight histories no better than those of a national survey sample from the same birth cohort, a result widely over-read in popular coverage." },
  { "year": "2015", "title": "Composite endpoint discussed with US regulators", "text": "Barzilai and colleagues present the TAME design; regulators indicate a composite of age-related events could be acceptable in principle." },
  { "year": "2017", "title": "Growth hormone receptor result", "text": "A deletion in exon 3 of the growth hormone receptor is associated with longer life in men across several cohorts, and with taller stature." },
  { "year": "2018", "title": "TAME biomarkers framework", "text": "A convened workgroup publishes which blood-based measures a geroscience trial should collect, building infrastructure a future application would need." },
  { "year": "2020", "title": "Age Later published", "text": "A trade book arguing that aging is the shared upstream cause of the diseases of old age." },
  { "year": "2026", "title": "TAME still unreported", "text": "More than a decade after it was designed, the trial has not been funded at scale and has reported no results." }
]
```

## The Longevity Genes Project

From the late 1990s Barzilai's group recruited several hundred Ashkenazi Jewish people aged 95 and
over, along with their adult children and a comparison group of similar age whose parents had
ordinary lifespans. The children are the analytically useful part of the design: young enough to be
measured directly, sharing half the genome of a demonstrated survivor, and comparable on disease
rates with their contemporaries, which is how the project connects to
[[compression-of-morbidity|compressed morbidity]] rather than to lifespan alone.

The most-quoted result is negative. On self-reported alcohol use, physical activity, diet and body
weight, the people in the cohort who had passed 95 looked no more virtuous than a general-population
sample from their own birth cohort.[^rajpathak2011] The finding says that in this population extreme
longevity is not explained by behaviour; it does not say that behaviour is irrelevant to ordinary lifespan, and
Barzilai has been careful in print to keep the two apart. The distinction is routinely lost in
popular coverage, where the result is read as licence to ignore the evidence behind
[[exercise-and-aging|exercise]], and it sits awkwardly beside the lifestyle-first framing of
[[blue-zones|long-lived population studies]].

The genetic findings are associations, and are described here as such. A variant of the cholesteryl ester transfer protein gene, CETP, was over-represented among the
long-lived, and carriers had larger high- and low-density lipoprotein particles.[^barzilai2003]
Functionally significant mutations in the insulin-like growth factor 1 receptor were later reported
as enriched among centenarian women, connecting the human cohort to the insulin and IGF-1 signalling
axis that governs lifespan in nematodes and mice, the pathway opened by
[[cynthia-kenyon|Kenyon's daf-2 work]].[^suh2008] A deletion in the third exon of the growth
hormone receptor was associated with longer life in men across several cohorts, and with taller
stature in carriers, a male-specific result that complicates the simple story that less growth
signalling means more years.[^benavraham2017]

> [!note] What a founder population buys and costs
> A genetically homogeneous population reduces the background variation that swamps association
> studies, so a modest cohort can detect variants a general-population study would miss. The price
> is generalizability: a variant enriched in Ashkenazi centenarians may be rare, differently
> penetrant, or absent elsewhere, and several of these associations have replicated unevenly.

### From variant to drug

The CETP story shows how far an association sits from a therapy. Pharmaceutical companies pursued CETP inhibition hard on the strength of its lipid biology. The
first such drug to reach a large outcome trial, torcetrapib, raised HDL cholesterol and also raised
mortality, and the trial was stopped.[^barter2007] Later compounds in the class were safer and did
little; none became a longevity drug. The inference that a variant carried by long-lived people
names a druggable target has not held.

> [!caution] Association is not mechanism
> None of the variants identified in centenarian cohorts has been shown to cause exceptional
> longevity, and no drug derived from one has extended human healthy life. The cohorts also carry a
> survivorship problem: people who reach 95 are, by construction, unrepresentative in ways no
> covariate captures.

## TAME and the regulatory argument

TAME, Targeting Aging with Metformin, was designed by Barzilai and colleagues and organized through
the American Federation for Aging Research, where he has held a scientific leadership role. As
proposed it would randomize roughly 3,000 adults aged 65 to 79 to metformin or placebo across
multiple US sites, follow them for about six years, and count time to the first of a composite of
unrelated age-related events.[^barzilai2016] The design and the evidence behind the drug are set out
under [[metformin]].

The composite is the substance of the proposal. Regulators approve drugs for diseases, and there is
no approval pathway for aging, so a drug that modestly delayed several conditions at once would fail
every existing trial design while doing exactly what geroscience predicts. US regulators met the
investigators in the mid-2010s and indicated that such an endpoint could be acceptable in principle.
A related workgroup Barzilai convened set out which blood-based [[aging-biomarkers|biomarkers]] a
geroscience trial should collect, an attempt to build the measurement infrastructure such an
application would need.[^justice2018] Neither an [[epigenetic-clock]] reading nor any other
[[biological-age|composite age measure]] is accepted as a surrogate endpoint, which is why TAME was
built around hard clinical events.

The trial has never been fully funded. Metformin is generic, so no sponsor can recover a budget in
the tens of millions of dollars from sales, and public money has arrived in amounts far short of
what a multi-centre outcome trial costs. That the field's most discussed trial has gone a decade
without enrolling, through years of fundraising, is itself a substantive fact: the interventions
cheap enough to give to everyone are the ones nobody owns.

## Reception

Barzilai is taken seriously as a clinician-scientist and is among the field's most effective
advocates with funders and regulators. The criticism is aimed at the vehicle rather than the design.
Metformin's geroprotective evidence in people without diabetes is weak, resting largely on
observational comparisons vulnerable to confounding by indication; the US National Institute on
Aging's Interventions Testing Program has not reported a lifespan extension from metformin in
genetically heterogeneous mice, in contrast to its result for [[rapamycin]]. A trial published in
2019 found that metformin blunted the mitochondrial adaptation of older adults to aerobic exercise
training,[^konopka2019] which would be an unfortunate property in a drug given to healthy people for
decades.

> [!debate] Right argument, wrong drug
> Critics generally accept that a multi-disease endpoint is what geroscience needs and doubt that
> metformin should carry it. Their argument is that a null result would be read as evidence about
> aging as an indication rather than about one weak biguanide, so a cheap drug with thin preclinical
> support puts the precedent at risk. Barzilai's answer has been that metformin was chosen for a
> safety record accumulated in millions of older people and for a price that would make a positive
> result usable everywhere.

His public framing is more restrained than that of several contemporaries. He argues for delaying
the onset of disease and extending [[healthspan]] rather than for indefinite lifespan, a position
well short of [[longevity-escape-velocity|escape-velocity]] claims of the kind associated with
[[aubrey-de-grey]], and unaccompanied by the supplement advocacy that has drawn criticism onto
[[david-sinclair]]. Colleagues who reject the drug rarely reject the framing, which is part of why
the proposal keeps a hearing.

## Legacy

The design has outlasted the effort to pay for it. Composite age-related endpoints, offspring-of-centenarian
cohorts, and the language of the [[longevity-dividend]] — the economic argument that delaying aging
returns more than curing any single disease — now appear routinely in geroscience proposals.
Aging-biology centres such as the [[buck-institute]] frame their work as one upstream target with
many diseases downstream, and the [[xprize-healthspan]] competition defines winning as function
restored across several domains rather than one disease improved.
His insistence that regulatory acceptance, not molecular novelty, is the binding constraint is now
close to consensus among people who want geroprotectors deployed rather than discussed, and it
shapes how [[senolytics]] and other candidate classes plan their trials.

The counterfactual is instructive. While TAME waited, the [[glp-1-receptor-agonists]] accumulated
large randomized outcome data across weight, cardiovascular, and kidney endpoints in adults with
obesity or diabetes, because a patented class had sponsors who could pay. Whether those results say
anything about aging as such is unsettled, but they show how multi-organ evidence actually gets
produced, and it is not the route TAME was designed to take.
The open question Barzilai's career poses is not whether the [[hallmarks-of-aging]] describe a
shared biology. It is whether a public good with no owner can be tested at all, and
[[access-and-inequality|who is left out]] if the answer arrives only through drugs someone can sell.

## See also

- [[metformin]]
- [[geroscience-hypothesis]]
- [[longevity-dividend]]
- [[aging-biomarkers]]
- [[compression-of-morbidity]]
- [[cynthia-kenyon]]
- [[healthspan]]
- [[access-and-inequality]]

## References

[^barzilai2003]: `paper` Barzilai, N. et al. "Unique lipoprotein phenotype and genotype associated with exceptional longevity." *JAMA*, 2003. {A case-control comparison within one founder population; the lipoprotein particle-size phenotype was the more robust finding.}
[^suh2008]: `paper` Suh, Y. et al. "Functionally significant insulin-like growth factor I receptor mutations in centenarians." *PNAS*, 2008. {Sequencing in a few hundred Ashkenazi Jewish people averaging about 98 years old, three-quarters of them women; the enrichment and the reduced receptor activity were reported in the women.}
[^benavraham2017]: `paper` Ben-Avraham, D. et al. "The GH receptor exon 3 deletion is a marker of male-specific exceptional longevity associated with increased GH sensitivity and taller stature." *Science Advances*, 2017.
[^rajpathak2011]: `paper` Rajpathak, S. N. et al. "Lifestyle factors of people with exceptional longevity." *Journal of the American Geriatrics Society*, 2011. {The comparison group came from a national health survey rather than from the study's own controls, and lifestyle was self-reported across decades.}
[^barter2007]: `paper` Barter, P. J. et al. "Effects of torcetrapib in patients at high risk for coronary events." *New England Journal of Medicine*, 2007. {Torcetrapib had off-target effects on blood pressure and aldosterone, so the trial does not by itself condemn CETP inhibition as a class.}
[^barzilai2016]: `paper` Barzilai, N., Crandall, J. P., Kritchevsky, S. B. and Espeland, M. A. "Metformin as a Tool to Target Aging." *Cell Metabolism*, 2016.
[^justice2018]: `paper` Justice, J. N. et al. "A framework for selection of blood-based biomarkers for geroscience studies: report of the TAME Biomarkers Workgroup." *GeroScience*, 2018.
[^agelater2020]: `book` Barzilai, N. *Age Later: Health Span, Life Span, and the New Science of Longevity*. St. Martin's Press, 2020.
[^konopka2019]: `paper` Konopka, A. R. et al. "Metformin inhibits mitochondrial adaptations to aerobic exercise training in older adults." *Aging Cell*, 2019. {A small trial with a mechanistic readout, not a clinical outcome; a companion 2019 paper reported blunted muscle hypertrophy after resistance training.}

==============================================================================
ARTICLE: non-invasive-neuromodulation
TITLE:   Non-invasive neuromodulation
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/non-invasive-neuromodulation
SOURCE:  https://futurehumanwiki.com/raw/non-invasive-neuromodulation
==============================================================================

---
title: "Non-invasive neuromodulation"
slug: "non-invasive-neuromodulation"
type: "technology"
status: "emerging"
horizon: "present"
trl: 8
categories: ["cybernetics", "enhancement"]
tags: ["tms", "tdcs", "ultrasound", "vagus nerve", "cognitive enhancement", "replication"]
summary: "Techniques that alter neural excitability through the intact skull or peripheral nerves — magnetic, electrical, and ultrasonic — without surgery or implanted hardware."
updated: "2026-07-27"
humanEvidence: "Repetitive TMS for depression is supported by sham-controlled trials in patients and is cleared and reimbursed; tDCS enhancement studies in healthy adults are small and showed no reliable effect when pooled."
access: "TMS is delivered in clinics and covered by many health systems; tDCS has no approved US indication yet is sold as a consumer wellness device and built at home; ultrasound is research-only."
reversibility: "reversible"
issues: ["The mid-2020s sham-controlled vagus nerve stimulation trial for depression is unnamed and uncited."]
---

```infobox
{
  "caption": "Family of neuromodulation techniques",
  "rows": [
    { "label": "Main modalities", "value": "TMS, tDCS/tACS, focused ultrasound, VNS" },
    { "label": "TMS introduced", "value": "1985, Sheffield" },
    { "label": "First depression clearance", "value": "2008 (rTMS, United States)" },
    { "label": "tDCS regulatory status", "value": "No US approval for any indication" },
    { "label": "Typical tDCS current", "value": "1–2 mA" },
    { "label": "Depth reachable", "value": "Cortex (TMS/tDCS); deep nuclei (ultrasound)" },
    { "label": "Readiness", "value": "TRL 9 (rTMS) to TRL 4 (ultrasonic)" }
  ]
}
```

**Non-invasive neuromodulation** covers the techniques that change nervous-system activity from
outside the body: magnetic pulses that induce currents in cortex, weak direct or alternating
currents applied to the scalp, focused ultrasound aimed through the skull, and electrical
stimulation of peripheral nerves reachable at the skin. They avoid the surgery, cost, and
hardware dependence of [[deep-brain-stimulation]] and of the implanted devices grouped under
[[neuroprosthetics]] — whose stimulation is adjustable and can be switched off, but whose electrodes
come out only with another operation — and they pay for that with far less spatial precision,
shallower reach, and — for several of the most popular methods — an evidence base that has not
survived rigorous replication.

## Modalities

**Transcranial magnetic stimulation** (TMS) discharges a capacitor through a coil held against the
scalp, producing a rapidly changing magnetic field that induces an electric field in the underlying
cortex strong enough to depolarize axons. A single pulse over motor cortex produces a measurable
twitch, which gives TMS something no other non-invasive method has: a direct, objective readout that
the stimulation reached its target — a form of ground truth that recording-based systems such as a
[[brain-computer-interface]] have to establish statistically. Repetitive TMS delivered in trains over days is the therapeutic
form, and patterned protocols such as intermittent theta burst compress a session from forty minutes
to a few.

**Transcranial electrical stimulation** passes 1–2 mA between scalp electrodes. Direct current
(tDCS) shifts resting membrane potentials, making neurons under the anode marginally more likely to
fire and those under the cathode marginally less; it does not itself trigger action potentials.
Alternating current (tACS) aims to entrain endogenous oscillations at a chosen frequency. Both are
cheap, portable, and easy to build, which explains both their popularity in research and their
regulatory problems.

**Transcranial focused ultrasound** (tFUS) at low intensity modulates neural activity, most likely
through mechanical effects on ion channels and membranes rather than heating. Its distinguishing
property is depth with focus: an ultrasound beam can be steered to a target a few millimetres across
in the thalamus or amygdala, structures unreachable by magnetic or electrical scalp stimulation. The
first human sensory-cortex demonstrations date from the 2010s.[^legon2014] High-intensity focused
ultrasound is a different technology, used to ablate the thalamic tremor target under magnetic
resonance guidance, and is approved for essential tremor.

**Peripheral approaches** stimulate nerves that project into the brain. Vagus nerve stimulation is
long established in implanted form for epilepsy and treatment-resistant depression; transcutaneous
variants stimulate the auricular branch at the ear or the cervical vagus at the neck, and are
cleared for headache indications.

## How they act on tissue

The physics sets the limits. The skull is a poor conductor and a strong acoustic barrier, and the
scalp is a low-resistance shunt. Direct measurements in human cadaver preparations found that the
great majority of current applied at conventional tDCS intensities is shunted through the scalp,
leaving field strengths in cortex of roughly a quarter of a volt per metre — well below the
threshold for driving spikes, and at the low end of what modulates firing probability in animal
recordings.[^voroslakos2018] That finding did not show tDCS does nothing; it showed that whatever it
does must be a subtle modulation of ongoing activity, and it rationalized both the small effect
sizes and the high inter-individual variability the literature reports.

Spatial focality is worse than most figures suggest. A conventional two-electrode tDCS montage
spreads current across large regions of cortex; TMS achieves perhaps a centimetre of focality at the
gyral crown and cannot reach the depths where the targets of implanted therapy sit. Ultrasound is
the exception, and its focality is why interest has moved that way.

Individual variability is the other structural problem. Skull thickness, cortical folding, baseline
neurotransmitter state, and prior activity all change the response, and studies of standard
plasticity protocols find that a substantial fraction of healthy participants show no effect or the
opposite of the expected one. Group averages are therefore a poor guide to any individual's
response.

## Clinical evidence

Repetitive TMS for major depression is the best-supported application. It was cleared in the United
States in 2008 after a sham-controlled trial, is reimbursed in many health systems, and produces
response rates in medication-resistant patients that are clinically meaningful without approaching
the remission rates of electroconvulsive therapy. An accelerated protocol delivering many theta-burst
sessions over five days with individualized functional-connectivity targeting reported high
remission rates in a small randomised sham-controlled trial and was cleared in 2022, though
replication at scale is still in progress.[^cole2022] TMS is also cleared for obsessive-compulsive
disorder and as an aid to smoking cessation.

Transcranial electrical stimulation has no approved US indication. European expert guidelines
assign tDCS probable efficacy for depression and for some pain conditions and note insufficient
evidence for most other proposed uses.[^fregni2021] A tDCS headset is regulated as a medical device
for depression in some European jurisdictions. The clinical case is real but modest, and it is
weaker than the volume of published literature implies.

Vagus nerve stimulation for treatment-resistant depression has been studied for two decades with
persistently ambiguous results, and a large sham-controlled trial reported mixed outcomes in the
mid-2020s. Non-invasive ear stimulation has a much thinner evidence base than its commercial
presence suggests.

## Cognitive enhancement and the replication problem

The largest gap between claim and evidence concerns tDCS as a cognitive enhancer in healthy people.
Hundreds of studies report improvements in working memory, attention, learning rate, and motor skill
acquisition, typically with fewer than twenty participants per group, one session, and one outcome
measure among several collected. A meta-analysis pooling single-session tDCS studies in healthy
adults found no reliable effect on any cognitive measure once the whole literature was
considered.[^horvath2015] The analysis was itself contested on methodological grounds, and the
argument has not fully resolved, but the field's own subsequent work has moved toward larger samples,
preregistration, and individualized dosing precisely because the earlier literature did not hold up.

The structural causes are familiar: small samples, flexible analysis, publication bias toward
positive results, and sham conditions that participants can often detect from the scalp sensation.
A detectable sham is not peculiar to stimulation — it is the central objection to trials of
[[psychedelic-therapy]], where an active dose is unmistakable within the hour — but a tingle under an
electrode is a far weaker cue than that, so the blind here is degraded rather than absent.
A specific complication is that tDCS effects are frequently state-dependent, helping poor performers
and impairing good ones, so an average across a group can be near zero while real effects exist in
subgroups. That is a reason for careful design, not a defence of the existing literature.

The same caution applies to claims made for [[nootropics]] and to any proposal that
[[intelligence-amplification]] can be achieved by stimulating cortex. Nothing in the current
evidence supports durable, general cognitive gain in healthy adults from any non-invasive
stimulation method, and the implanted approach to the same goal — the hippocampal
[[memory-prosthesis]] — has only within-session results in patients. Claims that stimulation can
reduce sleep need, discussed under [[sleep-reduction]], rest on still weaker foundations.

> [!caution] What a positive result usually means
> A typical enhancement finding is a within-session improvement of a few percent on one task, in
> around twenty people, relative to a sham the participants may have identified. It is not evidence
> that a device makes anyone smarter, and it does not transfer to untrained tasks.

## The consumer market and its regulation

Direct-to-consumer stimulation devices occupy a regulatory gap. Marketed for relaxation, sleep,
focus, or athletic training rather than for treating disease, many are sold under general-wellness
provisions that do not require efficacy evidence. A sports-oriented tDCS headset attracted
significant attention in the 2010s before its maker ceased operations; cranial electrotherapy
devices continue to be sold for anxiety and insomnia; and an active do-it-yourself community builds
its own units, an activity that sits alongside the practices described under
[[biohacking]].

Consumer electroencephalography headsets are increasingly sold alongside these devices, and scalp
recordings support more inference than their crudeness suggests — the concern taken up under
[[mental-privacy]] and legislated under [[neurorights]].

Harms are mostly minor — skin irritation and burns under electrodes, headache, transient mood
change — with the notable exception of TMS, where seizure is a rare but documented risk that keeps
it in clinical settings. The more consequential problem is opportunity cost and misinformation:
devices sold on the strength of a literature that has not replicated. Military interest in
stimulation for sustained vigilance has produced similar claims with similar evidentiary weakness,
and the question of whether such devices belong in competitive settings is unresolved, as discussed
under [[enhancement-in-sport]] and [[bioethics-of-enhancement]].

## Outlook

Focused ultrasound is the modality most likely to change what non-invasive stimulation can do,
because it is the only one that combines depth with a focal spot. If low-intensity ultrasound can
reliably modulate specific deep nuclei in humans, it would offer a reversible, repeatable
alternative to implanted electrodes for some psychiatric and pain indications, and would let
investigators run causal experiments in deep human structures that currently require surgery. As of
2026 the human evidence consists of small studies with heterogeneous protocols, and the dose-response
relationship is not well characterized.

The second trend is closing the loop. Stimulation timed to a measured brain state — an oscillation
phase, a sleep spindle, a decoded attentional lapse — should outperform stimulation delivered on a
fixed schedule, and it borrows directly from the methods of [[neural-decoding]] and from the
adaptive control now used in implanted devices. Whether that gain materializes in practice will
determine if non-invasive methods remain a weaker substitute for implanted devices or become a
distinct clinical tool with their own indications. The techniques that achieve genuine cell-type
specificity, [[optogenetics]] above all, work in animal brains and have reached people only in the
eye; a brain target would need gene delivery to a defined population and a light source implanted
next to it, which leaves ultrasound as the only non-invasive candidate combining depth with focus.

## See also

- [[deep-brain-stimulation]]
- [[brain-computer-interface]]
- [[neuroprosthetics]]
- [[nootropics]]
- [[intelligence-amplification]]
- [[human-enhancement]]
- [[biohacking]]
- [[mental-privacy]]

## References

[^legon2014]: `paper` Legon, W. et al. "Transcranial focused ultrasound modulates the activity of primary somatosensory cortex in humans." *Nature Neuroscience*, 2014.

[^voroslakos2018]: `paper` Vöröslakos, M. et al. "Direct effects of transcranial electric stimulation on brain circuits in rats and humans." *Nature Communications*, 2018. {The human part measures how much of a scalp-applied current actually reaches the cortex; it quantifies dose, not clinical or cognitive effect.}

[^cole2022]: `paper` Cole, E. J. et al. "Stanford Neuromodulation Therapy (SNT): a double-blind randomized controlled trial." *American Journal of Psychiatry*, 2022.

[^fregni2021]: `paper` Fregni, F. et al. "Evidence-based guidelines and secondary meta-analysis for the use of transcranial direct current stimulation in neurological and psychiatric disorders." *International Journal of Neuropsychopharmacology*, 2021.

[^horvath2015]: `paper` Horvath, J. C., Forte, J. D., Carter, O. "Quantitative review finds no evidence of cognitive effects in healthy populations from single-session transcranial direct current stimulation (tDCS)." *Brain Stimulation*, 2015. {Covers single-session studies in healthy adults only; its exclusion criteria were contested, and it does not bear on multi-session clinical protocols in patients.}

==============================================================================
ARTICLE: nootropics
TITLE:   Nootropics
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/nootropics
SOURCE:  https://futurehumanwiki.com/raw/nootropics
==============================================================================

---
title: "Nootropics"
slug: "nootropics"
type: "intervention"
status: "contested"
horizon: "present"
trl: 5
categories: ["enhancement"]
tags: ["cognition", "enhancement", "pharmacology", "supplements", "evidence", "regulation"]
summary: "Substances taken to improve cognition in healthy people, a category whose best-evidenced members produce small effects and whose best-selling members produce none."
updated: "2026-07-27"
humanEvidence: "Meta-analyses in healthy adults find small effects of stimulants and modafinil on laboratory tasks; no trial has shown any nootropic improves a real-world outcome, and the racetams have almost no modern human data."
access: "Caffeine and supplement blends are sold over the counter with no efficacy review; stimulants and modafinil are prescription-only and used off-label; racetams are bought online outside any approval."
reversibility: "reversible"
---

```infobox
{
  "caption": "Class of purported cognitive enhancers",
  "rows": [
    { "label": "Term coined", "value": "1972, Corneliu Giurgea" },
    { "label": "Prototype compound", "value": "Piracetam (1964)" },
    { "label": "Best-evidenced agents", "value": "Caffeine, modafinil, stimulants" },
    { "label": "Typical effect size", "value": "Small; d ≈ 0.2–0.3" },
    { "label": "Regulatory status", "value": "Prescription, supplement, or unapproved" },
    { "label": "Main hazard", "value": "Adulterated supplements, dependence" }
  ]
}
```

**Nootropics** are substances taken with the intention of improving memory, attention, motivation or executive function in people who are not cognitively impaired. The category is defined by user intent rather than by pharmacology, and it spans prescription stimulants, an over-the-counter supplement industry, and compounds sold as research chemicals. The evidence divides sharply: a few agents produce small, replicable effects on laboratory tasks, and most of the market produces none.

## Origins and definition

The Romanian pharmacologist Corneliu Giurgea synthesised piracetam at the Belgian firm UCB in 1964 and coined "nootropic" — from the Greek for mind-turning — around 1972 to name a class of drug he believed acted on higher integrative brain function without the sedation, stimulation or toxicity of existing psychoactive agents.[^giurgea1973] He proposed five criteria: enhancement of learning and memory; protection of learned behaviour against disruption; protection of the brain against physical or chemical injury; improvement of tonic cortical control mechanisms; and an absence of the usual psychotropic side effects.

The criteria have not aged well as a research programme. No compound has been shown to meet all five in humans, and piracetam itself, the drug the definition was built around, has never been approved in the United States and has weak evidence for the indications it is licensed for elsewhere. In practice "nootropic" now denotes a market segment rather than a pharmacological class, and much of what is sold under the label would have failed Giurgea's fifth criterion immediately.

```timeline
[
  { "year": "1964", "title": "Piracetam synthesised", "text": "Corneliu Giurgea's team at UCB produces the first racetam while looking for a compound active on the cortex." },
  { "year": "c. 1972", "title": "The term coined", "text": "Giurgea proposes 'nootropic' and a five-part definition intended to mark out a genuinely new drug class." },
  { "year": "1998", "title": "Modafinil approved", "text": "Modafinil is approved in the United States for narcolepsy, and becomes the most-studied off-label wakefulness agent." },
  { "year": "2008", "title": "The Nature poll", "text": "An informal reader survey finds about one in five respondents reporting non-medical use of drugs for focus, memory or concentration, and puts the issue on the scientific agenda." },
  { "year": "2015", "title": "Systematic reviews", "text": "Reviews of modafinil and of prescription stimulants in healthy people converge on small, task-dependent effects rather than general cognitive gain." },
  { "year": "2019–2020", "title": "Adulteration documented", "text": "Analyses of commercial cognitive-enhancement supplements find unapproved drugs present, sometimes at doses exceeding pharmaceutical ranges, and regulators issue warning letters." }
]
```

## What the evidence supports

**Caffeine** is the most widely used psychoactive substance in the world and the best-established agent in the category. It reliably improves alertness, vigilance and reaction time, especially under fatigue. A long-standing dispute concerns how much of the benefit in habitual users is restoration from overnight withdrawal rather than enhancement above an unmedicated baseline; the withdrawal-reversal effect is real, though evidence in non-consumers suggests it does not explain the whole effect.

**Modafinil** is licensed for narcolepsy and shift-work sleep disorder and used off-label for wakefulness. In sleep-deprived people its effects are clear. In rested healthy volunteers, a systematic review by Battleday and Brem found consistent improvement on complex tasks involving attention, executive function and learning, and little or no effect on simple tasks or on mood.[^battleday2015] An earlier review by Repantis and colleagues had reached a more cautious conclusion, finding limited evidence for enhancement outside sleep deprivation.[^repantis2010] Both note that longer and more demanding tasks show effects that brief tests miss.

**Prescription stimulants** — methylphenidate and amphetamine salts — have been studied extensively in healthy adults. A meta-analysis by Ilieva, Hook and Farah found small effects on inhibitory control, working memory and delayed episodic memory, on the order of a fifth to a third of a standard deviation.[^ilieva2015] Subjective reports of improvement consistently exceed measured performance, which is the finding most relevant to how these drugs are actually used. Effects are larger in low performers than in high performers, an inversion of the usual assumption about who benefits from enhancement.

> [!stat] The size of the effect
> The largest replicated cognitive gains from a nootropic in rested healthy adults are comparable to the effect of a night of adequate sleep or of moderate aerobic exercise, both of which are free and have no dependence liability. Any evaluation of the class has to start from that comparison.

## What the evidence does not support

**The racetams.** Piracetam, aniracetam, oxiracetam and phenylpiracetam are the compounds most closely associated with the word "nootropic". A Cochrane review of piracetam for dementia and cognitive impairment concluded the evidence was insufficient to support its use.[^flicker2001] Controlled data in healthy young adults are sparse, small and mostly old. The newer racetams have essentially no published human trials at all.

**Herbal and supplement agents.** Ginkgo biloba was tested in a large, long randomised prevention trial and did not reduce the incidence of dementia.[^dekosky2008] Bacopa monnieri, Rhodiola, lion's mane and citicoline have small trials with mixed results and heterogeneous preparations. Effects reported in this literature are rarely distinguishable from expectancy, and the placebo response for a subjectively assessed cognitive outcome is large.

**Proprietary blends.** Products combining a dozen ingredients at undisclosed doses cannot be evaluated. Where they have been analysed chemically, results have been poor: Cohen and colleagues found several unapproved drugs, including omberacetam, phenibut, vinpocetine, picamilon and aniracetam, present in commercially sold cognitive-enhancement supplements, sometimes at doses above those used pharmaceutically.[^cohen2020] Phenibut in particular carries dependence and withdrawal risk that consumers buying a "focus" supplement have no reason to expect.

> [!caution] Regulatory gap
> In the United States, [[dietary-supplements]] are not reviewed for efficacy before sale, and enforcement is post-market. A compound withdrawn from a supplement after a warning letter can reappear under a different name. Consumers of this market are not protected by the evidence standards that apply to the prescription drugs in the same category.

## Mechanisms and why they are hard to establish

The agents with real effects act on arousal and catecholamine signalling. Caffeine antagonises adenosine receptors; stimulants increase synaptic dopamine and noradrenaline; modafinil's action is less well characterised but involves dopamine transporter inhibition among other effects. None of these mechanisms is specific to cognition. They shift arousal and motivation, and cognitive performance changes as a consequence, which is why effects follow an inverted-U with dose and baseline state.

Wakefulness agents add a further complication: suppressing sleepiness is not the same as reducing the need for sleep, and the debt continues to accumulate while the drug masks it, an asymmetry examined in [[sleep-reduction]]. That structure has three consequences. Gains at one point on the curve are losses at another, so a drug that helps a tired person can impair a rested one. Improvement on one function often costs another: stimulants can narrow attention in a way that helps rote tasks and hurts creative or divergent ones. And a drug that raises confidence more than performance will be judged effective by the person taking it regardless of what it does. Careful studies of putative cognitive enhancers therefore separate objective measures from self-report, and the two frequently diverge.

## Use, ethics and policy

Non-medical stimulant use is concentrated in academic and competitive settings, where the pressure is positional. The ethical structure is the one described in [[enhancement-arms-race]]: where enough peers use a drug, declining ceases to be a neutral choice. Bioethicists writing in support of open access — the position developed in [[bioethics-of-enhancement]] and defended by [[julian-savulescu]] among others, and grounded in the autonomy claim of [[morphological-freedom]] — argue that a safe and effective cognitive drug should be treated like caffeine. Critics reply that the conditional has not been satisfied, since none of the drugs in question have been assessed for long-term safety in healthy users, and that prescription diversion is not a considered policy. The unregulated end of the market overlaps with the self-experimentation described in [[biohacking]], where dosing decisions are made from forum consensus.

Sport treats the question differently. Stimulants are prohibited in competition and modafinil is on the World Anti-Doping Agency list, with the therapeutic-use exemption process drawing an operational line between treatment and enhancement; see [[enhancement-in-sport]] and, for the anticipated genetic version, [[gene-doping]]. Military organisations have used modafinil and, historically, amphetamine for sustained operations, which is the clearest institutional endorsement of pharmacological cognitive enhancement anywhere and the setting in which that collective pressure is least deniable.

The wider context is set by [[human-enhancement]] and by the alternatives: [[non-invasive-neuromodulation]] devices marketed for cognition have their own replication problems, the psychedelic microdosing taken up under [[psychedelic-therapy]] reports benefits that self-blinded designs do not confirm, [[intelligence-amplification]] argues that external tools have delivered far more than any drug, and the genetic route described in [[genetic-enhancement-of-intelligence]] is constrained by polygenicity. Interest has also moved toward compounds that target [[hallmarks-of-aging|ageing biology]] on the hypothesis that preserving cognition is easier than augmenting it, a framing shared with [[healthspan]] research and pursued in [[exercise-and-aging]], which remains the best-evidenced intervention for cognitive function across the lifespan.

## Open questions

No nootropic has been shown to improve a real-world outcome — a grade, a diagnosis rate, a piece of work — in a controlled trial in healthy adults. Laboratory tasks are the entire evidence base, and their relation to consequential performance is weak. Long-term safety in non-patients is similarly unstudied: the trials that exist run for weeks, and the use pattern that matters runs for years.

Whether any pharmacological agent can raise cognitive capacity, as opposed to shifting arousal along a curve the brain already occupies, is the question Giurgea posed and it remains open. Sixty years after piracetam, no compound has clearly answered it in the affirmative.

## See also

- [[human-enhancement]]
- [[intelligence-amplification]]
- [[non-invasive-neuromodulation]]
- [[enhancement-in-sport]]
- [[genetic-enhancement-of-intelligence]]
- [[bioethics-of-enhancement]]
- [[biohacking]]
- [[sleep-reduction]]

## References

[^giurgea1973]: `paper` Giurgea, C. "The 'nootropic' approach to the pharmacology of the integrative activity of the brain." *Conditional Reflex*, 1973.
[^battleday2015]: `paper` Battleday, R. M. and Brem, A.-K. "Modafinil for cognitive neuroenhancement in healthy non-sleep-deprived subjects: A systematic review." *European Neuropsychopharmacology*, 2015.
[^repantis2010]: `paper` Repantis, D., Schlattmann, P., Laisney, O. and Heuser, I. "Modafinil and methylphenidate for neuroenhancement in healthy individuals: A systematic review." *Pharmacological Research*, 2010.
[^ilieva2015]: `paper` Ilieva, I. P., Hook, C. J. and Farah, M. J. "Prescription Stimulants' Effects on Healthy Inhibitory Control, Working Memory, and Episodic Memory: A Meta-analysis." *Journal of Cognitive Neuroscience*, 2015.
[^flicker2001]: `paper` Flicker, L. and Grimley Evans, J. "Piracetam for dementia or cognitive impairment." *Cochrane Database of Systematic Reviews*, 2001. {The trials pooled were in patients with dementia or cognitive impairment; the review says nothing about piracetam in healthy adults, who are most of the market.}
[^dekosky2008]: `paper` DeKosky, S. T. et al. "Ginkgo biloba for prevention of dementia: a randomized controlled trial." *JAMA*, 2008. {A long prevention trial in older adults with dementia incidence as the endpoint; it did not test short-term cognitive performance in healthy younger users.}
[^cohen2020]: `paper` Cohen, P. A. et al. "Five unapproved drugs found in cognitive enhancement supplements." *Neurology: Clinical Practice*, 2020. {A chemical analysis of products bought off the shelf, not a study in people; it establishes what was in the bottles tested, not what the ingredients do.}

==============================================================================
ARTICLE: crispr-off-target-effects
TITLE:   Off-target effects in genome editing
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/crispr-off-target-effects
SOURCE:  https://futurehumanwiki.com/raw/crispr-off-target-effects
==============================================================================

---
title: "Off-target effects in genome editing"
slug: "crispr-off-target-effects"
type: "concept"
status: "established"
horizon: "present"
categories: ["genetics", "society"]
tags: ["crispr", "genome editing", "safety", "gene therapy", "regulation", "mutagenesis"]
summary: "Unintended DNA changes caused by genome editors, either at sites resembling the target or at the target itself, and the assays used to find them."
updated: "2026-07-27"
issues: ["The 2024 FDA final guidance on genome editing products is described without a citation."]
---

```infobox
{
  "caption": "Safety problem in genome editing",
  "rows": [
    { "label": "Applies to", "value": "All programmable editors" },
    { "label": "Two classes", "value": "Off-target and on-target" },
    { "label": "First reported", "value": "2013, human cell lines" },
    { "label": "Main assays", "value": "GUIDE-seq, CIRCLE-seq, DISCOVER-Seq" },
    { "label": "Regulatory guidance", "value": "FDA final guidance, 2024" },
    { "label": "Hardest setting", "value": "Human embryos" }
  ]
}
```

**Off-target effects in genome editing** are DNA changes that an editor makes where it was not intended to act. In practice the term has widened to cover a second and arguably more consequential category: unintended outcomes at the correct site, where the editor cut exactly where instructed and the cell's repair produced something other than the desired edit. Off-target cutting was reported within a year of [[crispr-cas9]] being adapted to human cells.[^fu2013] The on-target failure modes took most of the following decade to characterise, and both classes have absorbed a substantial share of the field's effort since.

## What counts as an off-target effect

Guide RNAs tolerate mismatches, particularly at positions distant from the PAM, so a twenty-nucleotide spacer typically has dozens to hundreds of partially matching sites in a mammalian genome. Whether any of them is actually cut depends on chromatin accessibility, the concentration and persistence of the editor, and the specific mismatch pattern. Cutting at such a site produces the same indels as an on-target cut, in a gene nobody chose.

Editors that do not cut have their own version of the problem. Base editors carry a deaminase that can act on transiently exposed single-stranded DNA anywhere in the genome, entirely independently of where the guide RNA directs the protein — a mechanism that leaves scattered single-base changes with no cut site to search near. Prime editors show lower guide-dependent off-target activity than nucleases, since three separate base-pairing events are required, but can insert fragments of their own guide RNA at the target. Editors that only bind DNA, as in [[epigenome-editing]], perturb transcription at sites they would never cut, because binding tolerates more mismatches than cleavage does.

## Detection

No single assay is adequate, and regulators now expect at least two orthogonal ones. The families differ in what they trade away.

```compare
{
  "columns": ["In silico prediction", "Biochemical (cell-free)", "Cell-based or in vivo"],
  "rows": [
    { "label": "What it examines", "values": ["Reference genome sequence similarity", "Purified genomic DNA cut by the editor", "Breaks in living cells"] },
    { "label": "Sensitivity", "values": ["Low; misses non-obvious sites", "Very high", "Moderate"] },
    { "label": "False positives", "values": ["Many", "Many; naked DNA lacks chromatin", "Few"] },
    { "label": "Reflects the therapeutic cell type", "values": ["No", "Partly, if patient DNA is used", "Yes"] },
    { "label": "Typical role", "values": ["Nominating candidate sites", "Nominating candidate sites", "Confirming which are real"] }
  ]
}
```

The standard workflow nominates candidate sites broadly, then sequences those sites deeply in the actual edited cell product. Cell-based tagging methods capture double-strand breaks as they occur; one approach instead pulls down a DNA-repair protein that gathers at breaks, which allows the measurement to be made in a living animal rather than a dish.[^wienert2019] Whole-genome sequencing is conceptually cleaner but too insensitive for events present in a small fraction of cells.

> [!key] The reference genome is not the patient
> Off-target site nomination usually starts from a reference sequence. A patient carrying a common variant inside a candidate site may have a near-perfect match that the reference does not show. This became a specific question for sickle cell editing therapies, where the affected population is genetically diverse and under-represented in reference datasets.

## On-target damage

The larger surprise of the past decade was that cutting the right place can still go wrong. Repair of a Cas9-induced break produces, at measurable frequency, deletions of many kilobases, inversions and complex rearrangements that standard short-amplicon sequencing does not detect because the primer sites themselves are lost.[^kosicki2018] More severe outcomes have been documented: whole-chromosome-arm loss, loss of heterozygosity extending far from the cut, and chromothripsis, in which a chromosome shatters and is stitched back together in scrambled order.[^leibowitz2021]

Breaks also activate a p53-mediated damage response, which reduces editing efficiency in cells with intact p53 and therefore selects, weakly but systematically, for cells in which that pathway is impaired.[^ihry2018] The practical worry is not that editing causes cancer in any demonstrated case, but that the standard editing protocol applies a selection pressure whose direction is unfavourable.

Multiplexed editing compounds the problem. Making several cuts at once — routine in engineered cell therapies and in the heavily edited donor animals used for [[xenotransplantation]] — creates opportunities for translocation between the cut sites, and the frequency rises with the number of simultaneous breaks.

## Editing in embryos

The embryo is the worst case for every failure mode at once. Editing after the first cell division produces mosaicism, so the resulting individual carries a mixture of genotypes and a biopsy of a few cells does not represent the rest. Studies of human embryos have reported large on-target segmental losses and loss of heterozygosity around the cut site at appreciable frequency.[^zuccaro2020] An influential 2017 report interpreted apparent correction of a paternal allele as repair templated by the maternal chromosome; critics argued the more likely explanation was allele dropout, in which the paternal allele was deleted and simply failed to amplify.

This is why editing in embryos is not a scaled-down version of editing in a dish. The verification problem is structural: certifying that an embryo carries no unintended change requires sequencing cells that will not become the person, and the mosaicism that makes editing risky is the same property that makes the biopsy uninformative. Where the goal is avoiding a known monogenic disease, [[embryo-selection]] achieves it without editing in almost every case, which is the strongest practical argument against [[germline-editing]] independent of any ethical position.

## Mitigation

Several strategies reduce guide-dependent off-target activity, none to zero.

- **Engineered high-fidelity nucleases** with weakened non-specific DNA contacts, which lose activity faster on mismatched sites than on matched ones.[^kleinstiver2016]
- **Short exposure**, by delivering preassembled protein–RNA complexes that degrade within hours rather than plasmids or viral vectors that express for weeks. This is one reason ex vivo editing is safer than in vivo editing with [[aav-vectors]], and why messenger RNA in [[lipid-nanoparticles]] is preferred to persistent expression.
- **Paired nickases**, requiring two adjacent binding events to produce a break.
- **Guide selection** that avoids sites with close genomic relatives and avoids known polymorphic positions.
- **Switching editor class**, using [[base-editing]] or [[prime-editing]] where a double-strand break is not needed — which trades one error spectrum for another rather than eliminating error.

## Regulatory practice

The United States Food and Drug Administration issued final guidance on genome editing products in 2024 setting out expectations: orthogonal off-target nomination assays, confirmation in the clinical cell type, justification of any residual risk, and long-term follow-up of treated patients. The approval of [[casgevy]] in late 2023 was the first regulatory test of that framework, and the advisory discussion around it turned less on whether off-target edits had been found than on whether the assays used could have found them. That distinction — absence of evidence versus evidence of absence — remains the crux of every editing safety package.

For [[somatic-gene-therapy]] the residual uncertainty is managed the way oncology risk is managed: quantified where possible, monitored over years, and weighed against the disease. The frameworks surveyed in [[governance-of-genome-editing]] apply a different logic to heritable edits, and the [[he-jiankui-affair]] is the standing example of what happens when the safety analysis is performed by the person with an interest in the answer.

## Open problems

The field lacks a validated way to say how much off-target editing is too much, because no assay's sensitivity floor is well characterised in the tissue that matters. There is no established method for detecting rare structural rearrangements across a whole treated organ, no consensus on how to handle patient-specific variants at candidate sites, and little long-term human data — the treated cohorts are small and young. Editors deployed outside medicine, notably in [[gene-drive|gene drives]] released into wild populations, face the same detection limits with none of the follow-up infrastructure, which is one reason the [[precautionary-principle]] is invoked more forcefully there than in the clinic.

## See also

- [[crispr-cas9]]
- [[base-editing]]
- [[prime-editing]]
- [[germline-editing]]
- [[casgevy]]
- [[governance-of-genome-editing]]
- [[embryo-selection]]
- [[he-jiankui-affair]]

## References

[^fu2013]: `paper` Fu, Y. et al. "High-frequency off-target mutagenesis induced by CRISPR-Cas nucleases in human cells." *Nature Biotechnology*, 2013. {Human cell lines with sustained editor expression, conditions that exaggerate off-target rates relative to a transient ribonucleoprotein dose.}
[^kosicki2018]: `paper` Kosicki, M., Tomberg, K., Bradley, A. "Repair of double-strand breaks induced by CRISPR–Cas9 leads to large deletions and complex rearrangements." *Nature Biotechnology*, 2018.
[^ihry2018]: `paper` Ihry, R. J. et al. "p53 inhibits CRISPR–Cas9 engineering in human pluripotent stem cells." *Nature Medicine*, 2018.
[^wienert2019]: `paper` Wienert, B. et al. "Unbiased detection of CRISPR off-targets in vivo using DISCOVER-Seq." *Science*, 2019.
[^zuccaro2020]: `paper` Zuccaro, M. V. et al. "Allele-specific chromosome removal after Cas9 cleavage in human embryos." *Cell*, 2020. {Human embryos edited for research and never transferred; the losses were found only with assays designed to catch allele dropout.}
[^leibowitz2021]: `paper` Leibowitz, M. L. et al. "Chromothripsis as an on-target consequence of CRISPR–Cas9 genome editing." *Nature Genetics*, 2021.
[^kleinstiver2016]: `paper` Kleinstiver, B. P. et al. "High-fidelity CRISPR–Cas9 nucleases with no detectable genome-wide off-target effects." *Nature*, 2016. {The title claims absence of detection, which is bounded by the sensitivity of the 2016 assays rather than by the editor's fidelity.}

==============================================================================
ARTICLE: optogenetics
TITLE:   Optogenetics
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/optogenetics
SOURCE:  https://futurehumanwiki.com/raw/optogenetics
==============================================================================

---
title: "Optogenetics"
slug: "optogenetics"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 5
categories: ["cybernetics", "genetics"]
tags: ["neuroscience", "opsins", "gene delivery", "circuit mapping", "vision restoration", "research tools"]
summary: "A technique that makes selected neurons light-sensitive by expressing microbial opsin genes in them, allowing their activity to be switched on or off on a millisecond timescale."
updated: "2026-07-27"
humanEvidence: "One published case report describes partial visual recovery in a patient with retinitis pigmentosa after opsin delivery to retinal ganglion cells; every use outside the eye, from engram control to cochlear and cardiac work, is in animals."
access: "Nothing to obtain clinically: no optogenetic therapy is approved anywhere and the only human use is inside ophthalmology trials, while opsin constructs are standard laboratory reagents."
reversibility: "irreversible"
issues: ["Opsin trials targeting bipolar cells are mentioned without a source."]
---

```infobox
{
  "caption": "Neuroscience technique",
  "rows": [
    { "label": "Core components", "value": "Microbial opsin plus light source" },
    { "label": "First neural demonstration", "value": "2005" },
    { "label": "Key developers", "value": "Boyden, Deisseroth, Hegemann, Nagel, Miesenböck" },
    { "label": "Temporal precision", "value": "Milliseconds" },
    { "label": "Main research use", "value": "Causal circuit mapping" },
    { "label": "First human use", "value": "2021, retinal ganglion cells" },
    { "label": "Main limit in humans", "value": "Gene and light delivery" },
    { "label": "Readiness", "value": "TRL 9 research tool, TRL 5 therapy" }
  ]
}
```

**Optogenetics** is the use of light-sensitive proteins, introduced into cells by gene transfer, to
control the electrical activity of those cells with light. In neuroscience it made a specific and
previously unavailable experiment possible: activating or silencing one genetically defined
population of neurons, in a behaving animal, on the timescale of a single action potential, while
leaving neighbouring cells untouched. That combination of cell-type specificity and millisecond
timing is what electrical methods such as [[deep-brain-stimulation]] cannot
provide.[^boyden2005]

## How it works

The active ingredient is a microbial opsin, a membrane protein from algae, archaea, or bacteria that
absorbs light and moves ions across the membrane. Channelrhodopsin-2, from the green alga
*Chlamydomonas reinhardtii*, opens a cation channel when illuminated with blue light, depolarizing
the cell. Halorhodopsin, a chloride pump from a haloarchaeon, and archaerhodopsin, a proton pump,
hyperpolarize the cell under yellow or green light and therefore silence it. All require the
cofactor all-*trans* retinal, which vertebrate brain tissue supplies in sufficient quantity — a
biological accident without which the technique would not work in mammals.

An experiment has three parts. The opsin gene is delivered to the target cells, usually by
[[aav-vectors]] carrying a cell-type-specific promoter, or by crossing a transgenic animal line
expressing Cre recombinase in a defined population with a Cre-dependent opsin construct. Light is
delivered by an optical fibre, an implanted micro-LED array, or, in transparent preparations, from
outside. Activity is then read out by behaviour, electrophysiology, or fluorescent calcium imaging.

The engineered variants matter as much as the original discovery. Trafficking sequences improve
membrane targeting; faster mutants follow spike trains above 100 Hz; step-function opsins stay open
for minutes after a brief pulse; red-shifted opsins absorb longer wavelengths that scatter less in
tissue; and soma-targeted opsins restrict expression to the cell body so that holographic
two-photon illumination can address individual neurons without exciting passing axons.

## Development history

```timeline
[
  { "year": "1971–1979", "title": "Precursors", "text": "Bacteriorhodopsin is characterized as a light-driven proton pump. Francis Crick later argues that neuroscience needs a method to control one cell type at a time, and speculates that light might provide it." },
  { "year": "2002–2003", "title": "Channelrhodopsins identified", "text": "Georg Nagel, Peter Hegemann and colleagues characterize channelrhodopsin-1 and channelrhodopsin-2 from green algae as directly light-gated ion channels." },
  { "year": "2002", "title": "First genetically targeted photostimulation", "text": "Gero Miesenböck's group controls neurons using a multi-component invertebrate phototransduction system, establishing the concept before a single-gene tool existed." },
  { "year": "2005", "title": "Optogenetics in neurons", "text": "Boyden, Zhang, Deisseroth and colleagues show that channelrhodopsin-2 alone confers millisecond-precision light control on mammalian neurons." },
  { "year": "2007–2010", "title": "Inhibition and colour range", "text": "Halorhodopsin and archaerhodopsin provide optical silencing; red-shifted and faster variants broaden the toolkit." },
  { "year": "2012–2013", "title": "Memory engrams manipulated", "text": "Optogenetic reactivation of hippocampal cells labelled during learning elicits the corresponding memory in mice, and a false memory is created by pairing artificial reactivation with a new experience." },
  { "year": "2013–2021", "title": "Recognition", "text": "The Brain Prize, the Shaw Prize and the Lasker Basic Medical Research Award are given for the development of optogenetics." },
  { "year": "2021", "title": "First human result", "text": "A patient with retinitis pigmentosa recovers partial visual function after retinal ganglion cells are made light-sensitive and driven by image-projecting goggles." }
]
```

## What it is used for

Optogenetics is, in practice, a research tool, and its influence on neuroscience comes from turning
correlational claims into causal ones. Before it, showing that a population of neurons was active
during a behaviour was straightforward and showing that the activity caused the behaviour was not.
Optogenetics supplies the missing manipulation, and it has been used to map circuits for fear,
reward, feeding, aggression, thirst, sleep-wake transitions, and locomotion, generally by activating
or silencing a projection defined by both its cell type and its anatomical target.

The most striking demonstrations concern memory. Cells active during fear learning can be tagged
with an opsin using activity-dependent promoters; reactivating those cells later elicits the
behavioural signature of the memory in the absence of the original cue.[^liu2012] Pairing artificial
reactivation of a context representation with an aversive stimulus produces behaviour consistent
with a memory of an event that never occurred. These experiments are the strongest existing evidence
that a discrete, manipulable physical substrate of a specific memory exists, which bears on the
premises of [[whole-brain-emulation]] and on the encoding assumptions behind a
[[memory-prosthesis]].

Optogenetics also underpins "all-optical" physiology: an opsin for writing, a genetically encoded
calcium or voltage indicator for reading, and holographic light shaping to address chosen neurons.
This is the closest thing neuroscience has to a bidirectional interface with single-cell resolution,
and it is available only in animals; no [[brain-computer-interface]] in a person approaches that
precision on either the read or the write side. Structural mapping methods such as
[[connectomics]] give the wiring; optogenetics tests what the wiring does.

## Human applications

Only one clinical route has produced a published human result. In an approach that treats the eye as
a uniquely favourable target — accessible, transparent, small, and relatively immune-privileged — a
red-shifted opsin was delivered by intravitreal AAV to the surviving retinal ganglion cells of a
patient blinded by retinitis pigmentosa, and paired with goggles that convert the visual scene into
amber light pulses. The patient regained the ability to locate, count, and touch objects, which he
could not do beforehand.[^sahel2021] This is a single case reported from an ongoing trial, and the
recovered function is far below normal vision, comparable in magnitude to what a
[[retinal-implant]] delivers. Other opsin-based programmes target bipolar cells rather than ganglion
cells and remain in trials; the eye is still the only tissue in which the technique has been tried in
people at all.

The advantage over an electronic implant is that no hardware goes inside the eye and the number of
addressable cells is limited by expression rather than by electrode count. The disadvantages are
that it requires a functioning ganglion cell layer, that the opsin's light sensitivity is far below
that of natural photoreceptors so external amplification is mandatory, and that expressing a
microbial protein in a human tissue for life carries an immunological question that no long-term
data yet answers.

Beyond the eye, proposals are earlier. An optical [[cochlear-implant]] using opsin-expressing spiral
ganglion neurons would exploit light's poor spread in tissue to deliver more independent channels
than electrical stimulation allows; the concept has been demonstrated in rodents. Cardiac
optogenetics for pacing and defibrillation works in animal hearts. Neither has a human study.

> [!caution] A research tool, not yet a therapy
> Optogenetics transformed animal neuroscience within a decade. Its clinical footprint after twenty
> years is one published case report in a rare form of blindness. The gap is not scientific
> uncertainty about the mechanism; it is that every therapeutic use requires solving gene delivery
> and light delivery simultaneously, in a specific tissue, with acceptable immunology.

## Delivery and physical limits

Gene delivery is the same problem faced by [[somatic-gene-therapy]] generally: viral tropism,
pre-existing immunity to common AAV serotypes, limited packaging capacity, and the difficulty of
restricting expression to one cell type without a Cre driver line that humans do not have. Editing
tools such as [[crispr-cas9]] do not solve this, because the problem is delivery and cell-type
targeting rather than the sequence change itself. Non-viral
carriers such as [[lipid-nanoparticles]] do not efficiently reach neurons after systemic
administration.

Light delivery is the constraint with no biological workaround. Blue light is scattered and absorbed
within roughly a millimetre of brain tissue, so deep targets require an implanted fibre or LED —
reintroducing exactly the invasiveness optogenetics was supposed to avoid, plus tissue heating and a
foreign-body response indistinguishable from that around a recording electrode. Red-shifted opsins
buy a modest depth increase. Upconversion nanoparticles that convert tissue-penetrating near-infrared
light into visible emission at the target have been demonstrated in mice, adding a second delivery
problem to the first.[^chen2018]

Related approaches attempt to remove the light source altogether. Chemogenetic receptors activated
by an inert drug provide cell-type specificity with minutes-to-hours resolution instead of
milliseconds. Sonogenetics uses mechanosensitive channels driven by ultrasound, which penetrates
tissue far better than light. Magnetogenetic schemes, which claimed magnetic control of engineered
channels, have been challenged on the grounds that the reported effects exceed what the physics of
magnetic energy transfer to a protein permits, and remain unresolved — a useful reminder that
plausible-sounding tools in this area do not always survive scrutiny.

## Outlook

The research trajectory is toward larger-scale all-optical control: reading and writing hundreds of
individually specified neurons in a behaving animal, which would let investigators test whether
specific activity patterns are sufficient for a percept or a decision. That work bears directly on
the questions raised under [[neural-correlates-of-consciousness]], since it is the only method that
can impose a precise pattern rather than merely observe one.

Clinically, the near-term expectation is narrow: retinal indications where the anatomy cooperates,
and possibly the inner ear. A general-purpose optogenetic alternative to implanted electrodes would
require gene delivery to defined deep populations and a light source that does not need surgery,
and no current approach supplies both. Whether the technique's extraordinary value in animal
research ever converts into a broad clinical technology is, twenty years in, still an open
question — and the more likely path may be that optogenetics remains the instrument that tells other
technologies, from [[neuroprosthetics]] to [[neural-decoding]], what to aim at.

## See also

- [[deep-brain-stimulation]]
- [[retinal-implant]]
- [[aav-vectors]]
- [[neuroprosthetics]]
- [[connectomics]]
- [[memory-prosthesis]]
- [[somatic-gene-therapy]]
- [[non-invasive-neuromodulation]]

## References

[^liu2012]: `paper` Liu, X. et al. "Optogenetic stimulation of a hippocampal engram activates fear memory recall." *Nature*, 2012. {Mice; the readout is freezing behaviour, which indexes a fear response and says nothing about the content of what was recalled.}

[^sahel2021]: `paper` Sahel, J.-A. et al. "Partial recovery of visual function in a blind patient after optogenetic therapy." *Nature Medicine*, 2021. {A single patient from an ongoing trial, tested on locating and counting objects while wearing the stimulating goggles; there is no control comparison.}

[^chen2018]: `paper` Chen, S. et al. "Near-infrared deep brain stimulation via upconversion nanoparticle-mediated optogenetics." *Science*, 2018.

[^boyden2005]: `paper` Boyden, E. S., Zhang, F., Bamberg, E., Nagel, G., Deisseroth, K. "Millisecond-timescale, genetically targeted optical control of neural activity." *Nature Neuroscience*, 2005.

==============================================================================
ARTICLE: organ-bioprinting
TITLE:   Organ bioprinting
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/organ-bioprinting
SOURCE:  https://futurehumanwiki.com/raw/organ-bioprinting
==============================================================================

---
title: "Organ bioprinting"
slug: "organ-bioprinting"
type: "technology"
status: "experimental"
horizon: "2040s"
trl: 4
categories: ["bodies"]
tags: ["bioprinting", "tissue engineering", "vascularization", "regenerative medicine", "transplantation", "biofabrication"]
summary: "The additive manufacture of living tissue by depositing cells and biomaterials in a programmed pattern, pursued as a route to transplantable organs but limited so far to simple structures."
updated: "2026-07-27"
humanEvidence: "A printed autologous ear cartilage implant has been placed in patients in a microtia trial and printed skin has reached early human work; no printed solid organ has been implanted in a person or supported an animal."
access: "Nothing to obtain: printed tissue is sold only as a research and drug-screening product, and the sole human implants are inside small company-run trials."
reversibility: "difficult"
issues: ["The 2022 3DBio ear implant and the microtia trial are described without a source."]
---

```infobox
{
  "caption": "Biofabrication technology",
  "rows": [
    { "label": "Type", "value": "Additive biofabrication" },
    { "label": "Main modalities", "value": "Extrusion, droplet, light-based" },
    { "label": "Feedstock", "value": "Cell-laden bioinks" },
    { "label": "Typical resolution", "value": "Tens to hundreds of micrometres" },
    { "label": "First clinical implant", "value": "Printed ear cartilage, 2022" },
    { "label": "Solid organs implanted", "value": "None" },
    { "label": "Principal bottleneck", "value": "Vascularization" },
    { "label": "Readiness", "value": "Laboratory and first trials" }
  ]
}
```

**Organ bioprinting** is the additive manufacture of living tissue: a machine deposits cells, hydrogels and sacrificial support materials in a programmed three-dimensional pattern, and the construct is then cultured until the cells remodel it into something with tissue-like structure and function. The technique borrows its motion control from industrial 3D printing and its biology from [[tissue-engineering]]. It has produced skin, cartilage, corneal stroma, vascular grafts and beating cardiac patches in the laboratory. It has not produced a solid organ that anyone has transplanted into a person, and the gap between those two statements is the subject of this article.

```keyfacts
[
  { "value": "~200 µm", "label": "Maximum viable distance from a capillary", "note": "beyond which cells outstrip passive oxygen diffusion" },
  { "value": "10⁸/mL", "label": "Cell density of native solid tissue", "note": "orders of magnitude above most printable bioinks" },
  { "value": "0", "label": "Printed solid organs implanted in humans", "note": "as of 2026" }
]
```

```figure
{"key": "organovo-bioprinter", "caption": "Depositing cells accurately has never been the bottleneck. Keeping printed tissue alive beyond a millimetre from a nutrient supply is."}
```

## How it works

A bioprinter converts a digital model, usually derived from a patient's CT or MRI scan, into a toolpath and then lays down material along it. Three deposition families dominate.

**Extrusion printing** pushes a viscous cell-laden gel through a nozzle by pneumatic pressure or a screw. It is the workhorse: cheap, tolerant of high cell densities, compatible with most hydrogels. Its weaknesses are resolution, typically a few hundred micrometres, and shear stress at the nozzle, which damages cells. Extrusion also cannot print an unsupported soft gel — a low-stiffness ink collapses under its own weight. The standard answer is embedded printing, in which the nozzle deposits into a supporting bath of granular gel that behaves as a solid at rest and yields locally around the moving needle. The FRESH method developed in Adam Feinberg's laboratory uses a gelatin microparticle slurry for this and was used to print collagen replicas of heart valves and a neonatal-scale ventricle that contracted when seeded with cardiomyocytes.[^lee2019]

**Droplet and inkjet printing** eject picolitre droplets of low-viscosity cell suspension. Resolution is better and cell damage lower, but the constructs are mechanically weak and the technique struggles with anything thick.

**Light-based printing** cures a photosensitive resin with patterned light. Digital light processing projects an entire layer at once, which is fast and precise; two-photon polymerization reaches sub-micrometre features but is far too slow for organ-scale volumes. Volumetric printing, in which tomographic light projections are rotated around a resin vial so that a whole object solidifies at once where the accumulated dose crosses a threshold, produces centimetre-scale constructs in tens of seconds and avoids layering artefacts altogether.[^bernal2019] A landmark demonstration used food-colouring dyes as biocompatible photoabsorbers to print hydrogels containing intertwined, topologically independent vascular networks, including an air-sac model that oxygenated human red blood cells flowing past it during cyclic ventilation.[^grigoryan2019]

### Bioinks

A bioink must satisfy contradictory requirements. It has to flow through a nozzle or cure under light, hold its shape afterwards, stay soft enough for cells to remodel, degrade at the rate new matrix is deposited, and be non-toxic throughout. Common bases are alginate, gelatin methacryloyl, fibrin, collagen I, hyaluronic acid, and decellularized extracellular matrix digested from donor tissue, which brings tissue-specific composition at the cost of batch variability. See [[decellularized-scaffolds]] for that material family.

The deeper problem is cell density. Native solid tissue packs cells at roughly a hundred million per millilitre; a printable gel usually carries one to two orders of magnitude fewer, because at high density the ink loses printability and the cells lose oxygen. One workaround inverts the usual arrangement: rather than printing cells in a gel, the SWIFT approach compacts organoid-derived cells into a dense living matrix and prints only the sacrificial vascular network into it, then washes the sacrificial ink out to leave perfusable channels.[^skylarscott2019] The tissue is the bulk; the print is the plumbing.

## Development history

```timeline
[
  { "year": "1988", "title": "Cells through a printer", "text": "Robert Klebe adapts an office inkjet to position cells on a substrate, coining 'cytoscribing' and establishing that printing hardware need not kill cells." },
  { "year": "2003", "title": "Thermal inkjet bioprinting", "text": "Thomas Boland's group at Clemson shows that modified commercial inkjet printers can deposit viable mammalian cells in patterns." },
  { "year": "2012", "title": "Sacrificial vasculature", "text": "Jordan Miller and colleagues print lattices of carbohydrate glass, cast tissue around them and dissolve them away, leaving open perfusable channels inside a cell-laden gel." },
  { "year": "2019", "title": "Three converging advances", "text": "Embedded collagen printing produces heart components; tomographic light projection prints centimetre-scale objects in seconds; entangled multivascular networks are printed in a single monolithic gel." },
  { "year": "2022", "title": "First printed implant in a trial", "text": "3DBio Therapeutics implants an ear reconstructed from a patient's own chondrocytes in a microtia clinical trial, the first bioprinted living implant placed in a person under regulatory oversight." },
  { "year": "2023–2026", "title": "Industrialization without organs", "text": "Bioprinted tissue moves into drug screening and into partnerships with pharmaceutical firms for implantable therapeutic tissues; no solid organ reaches the clinic." }
]
```

## What has actually been implanted

The honest inventory is short and consists of tissues that are thin, avascular, or both.

Printed cartilage has reached patients. An ear scaffold printed from a patient's expanded chondrocytes was implanted in a microtia trial beginning in 2022; cartilage is a natural first target because it is avascular, mechanically simple and immunologically quiet when autologous. Printed skin substitutes have been tested in animals and in early human work, including handheld and robot-mounted devices that deposit cells directly onto a wound bed. Printed bone-graft substitutes and airway splints exist, but most of these are acellular polymer devices produced by conventional 3D printing rather than bioprinting proper, and the distinction matters: a resorbable printed splint is a device, while a construct containing living cells is regulated as a biologic or a combination product.

Bioprinting hardware has also been flown to the International Space Station, on the reasoning that in microgravity a soft construct holds its shape without a support bath. The samples produced there are small, and the motivation overlaps with the broader interest in manufacturing medical supplies far from resupply described in [[space-medicine]].

Everything organ-shaped remains a laboratory object. A widely reported 2019 construct from Tal Dvir's group at Tel Aviv University, printed from a patient's cells in a bioink derived from their own omentum, was about the size of a cherry, had the gross anatomy of a heart, and did not pump.[^noor2019] The FRESH collagen ventricle contracted but produced negligible pressure. No printed kidney, liver or heart has supported an animal, let alone a person. Where whole-organ replacement has actually reached patients, it has done so through [[xenotransplantation]] or through mechanical substitutes such as the [[artificial-heart]], not through printing; the competing strategy of growing an organ from cells rather than depositing it is treated in [[lab-grown-organs]].

> [!caution] What "printed a heart" means
> Press coverage of bioprinting routinely reports the geometry as though it were the achievement. Printing the shape of an organ is a solved problem; a desktop machine can do it in a day. Producing tissue with the correct cell types, in the correct densities, arranged around a perfused vascular tree, at a mechanical and metabolic standard that keeps a human alive, is unsolved and not close to solved.

## The vascularization bottleneck

Every cell in a solid organ sits within roughly a hundred to two hundred micrometres of a capillary, because that is how far oxygen diffuses through tissue before consumption exhausts it. Any construct thicker than about a millimetre without internal perfusion develops a necrotic core within days. This single constraint explains the field's shape: the tissues that work clinically are sheets and small pieces, and the tissues that do not work are everything else.

Printed vascular networks solve part of the problem and expose the rest. Current printers resolve channels down to a few hundred micrometres reliably and to tens of micrometres in specialised light-based systems. Human capillaries are five to ten micrometres across, branch across roughly a dozen generations, and in a kidney or liver form a network whose total surface area is measured in tens of square metres. No printing process approaches that resolution at organ scale in an acceptable time. The prevailing strategy is therefore hybrid: print the large and medium vessels, seed them with endothelial cells, and rely on angiogenic sprouting from those vessels to build the capillary bed biologically. Whether self-assembly can reliably complete a network that a printer starts, and do so before the tissue starves, is the central open question.

Perfusion also has to be surgically connectable. A printed construct needs an inlet and an outlet that a surgeon can anastomose to the recipient's circulation, with wall strength sufficient to hold arterial pressure and an endothelial lining continuous enough not to trigger clotting. An incompletely endothelialized channel thromboses, which is the same failure mode that defeated recellularized donor scaffolds and that limits [[artificial-blood]] and mechanical circulatory support.

## Limitations and risks

**Cell supply.** A human-scale solid organ contains cells on the order of ten to a hundred billion. Sourcing them autologously from [[induced-pluripotent-stem-cells]] means months of expansion and differentiation per patient at high cost, on a timescale incompatible with acute organ failure. Allogeneic cells shorten the timeline but reintroduce immunosuppression, which is much of what a grown organ was supposed to avoid. Where the organ failed because of an inherited defect, the patient's own cells carry that defect and must be corrected before printing, adding the manufacturing burden of [[somatic-gene-therapy]] to the tissue problem.

**Maturation.** Printed tissue starts fetal-like. Cardiomyocytes derived from stem cells contract weakly and lack adult calcium handling; hepatocytes lose function within days in culture. Maturation requires mechanical and electrical conditioning in bioreactors over weeks, and it is incomplete. The same immaturity ceiling constrains [[organoids]].

**Quality control and regulation.** A printed organ is a patient-specific manufactured product, which sits awkwardly in regulatory frameworks built either for mass-produced devices or for standardized biologics. Regulators have issued guidance on additive manufacturing of devices while explicitly leaving cell-containing products outside its scope. Sterility, potency assays, batch release testing and shelf life all have to be defined for a product with a batch size of one.

**Tumorigenicity.** Constructs built from pluripotent-derived cells carry residual risk of undifferentiated cells forming teratomas, the same hazard that constrains [[epigenetic-reprogramming]] and stem-cell therapies generally.

**Cost and equity.** If a printed organ works, it would initially be expensive and produced by a small number of facilities. The distribution problem that follows is not technical, and [[access-and-inequality]] applies to it in full.

Measured on the scale set out in [[technology-readiness-level]], the field spans an unusual range: printed skin and cartilage sit near clinical validation, printed vascularized organs remain at proof of concept, and the two are routinely reported under the same heading.

## Outlook

The near-term commercial reality of bioprinting is not transplantation but tissue models: printed liver, kidney and tumour tissue used for toxicology and drug screening, a market that grew as regulators softened the requirement for animal testing in some drug categories. Therapeutic products under active development are small and encapsulated — printed islet constructs for type 1 diabetes, printed cartilage, printed nerve conduits, printed corneal tissue — where a few cubic centimetres of tissue can do clinical work without a vascular tree.

For solid organs, several groups now argue that printing will not be the fabrication method at all, but the scaffolding method: printing the vasculature and letting cells build the parenchyma. That reframing makes the problem tractable in principle and leaves it hard in practice, and it borrows its logic from developmental biology rather than manufacturing — the same logic that makes the study of animals capable of [[limb-regeneration]] relevant to a discipline built around machine tools. A useful benchmark for the field's honesty is that no printed construct has yet supported a large animal's physiology for a month. Until one does, timelines for a printed kidney rest on extrapolation rather than evidence, and the [[organ-shortage]] that motivates the whole enterprise continues to be addressed by donation reform, machine perfusion and edited pigs.

## See also

- [[tissue-engineering]]
- [[decellularized-scaffolds]]
- [[lab-grown-organs]]
- [[organoids]]
- [[organ-shortage]]
- [[xenotransplantation]]
- [[artificial-heart]]
- [[induced-pluripotent-stem-cells]]

## References

[^lee2019]: `paper` Lee, A. et al. "3D bioprinting of collagen to rebuild components of the human heart." *Science*, 2019.
[^bernal2019]: `paper` Bernal, P. N. et al. "Volumetric bioprinting of complex living tissue constructs within seconds." *Advanced Materials*, 2019.
[^grigoryan2019]: `paper` Grigoryan, B. et al. "Multivascular networks and functional intravascular topologies within biocompatible hydrogels." *Science*, 2019. {The oxygen-transfer demonstration used a hydrogel air-sac model perfused on the bench; it is a materials and geometry result, not living lung tissue.}
[^skylarscott2019]: `paper` Skylar-Scott, M. A. et al. "Biomanufacturing of organ-specific tissues with high cellular density and embedded vascular channels." *Science Advances*, 2019.
[^noor2019]: `paper` Noor, N. et al. "3D printing of personalized thick and perfusable cardiac patches and hearts." *Advanced Science*, 2019. {Reported worldwide as the first 3D-printed heart; the heart-shaped construct was cherry-sized and did not pump, and the contribution is the patient-derived bioink.}

==============================================================================
ARTICLE: organ-shortage
TITLE:   Organ shortage
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/organ-shortage
SOURCE:  https://futurehumanwiki.com/raw/organ-shortage
==============================================================================

---
title: "Organ shortage"
slug: "organ-shortage"
type: "concept"
status: "established"
horizon: "present"
categories: ["bodies", "society"]
tags: ["transplantation", "health policy", "donation", "allocation", "bioethics", "kidney"]
summary: "The persistent gap between the number of people who need a transplant and the number of usable donated organs, which drives most work on engineered and animal-derived replacements."
updated: "2026-07-27"
issues: ["Waiting-list size, the daily death figure and the 2023 transplant record are attributed to federal agencies but uncited.", "The claim that Iran eliminated its kidney waiting list needs a source."]
---

```infobox
{
  "caption": "Health system problem",
  "rows": [
    { "label": "US waiting list", "value": "Over 100,000 candidates" },
    { "label": "Largest component", "value": "Kidney, about 85–90%" },
    { "label": "US deaths on the list", "value": "About 17 per day" },
    { "label": "Deaths permitting donation", "value": "A small percentage" },
    { "label": "Highest national donation rate", "value": "Spain" },
    { "label": "US kidney discard rate", "value": "Roughly one in five" },
    { "label": "Global need met", "value": "Under one tenth" }
  ]
}
```

**Organ shortage** is the structural gap between the number of people who would benefit from a transplant and the number of organs available to transplant. It is the practical reason most of the technologies covered elsewhere on this wiki exist: [[xenotransplantation]], [[organ-bioprinting]], [[lab-grown-organs]] and mechanical support such as the [[artificial-heart]] are all responses to a supply constraint that donation policy has narrowed but never closed. The shortage is usually described with waiting-list statistics, which understate it, because being on a list requires being sick enough to need an organ and well enough to survive receiving one.

```keyfacts
[
  { "value": "~17/day", "label": "US deaths while waiting", "note": "federal figure; excludes those never listed" },
  { "value": "~1 in 5", "label": "Recovered US kidneys not transplanted", "note": "higher than most European systems" },
  { "value": "<10%", "label": "Share of global transplant need met", "note": "WHO Global Observatory estimate" }
]
```

## The arithmetic

More than a hundred thousand people are registered on the United States transplant waiting list at any time, the large majority of them waiting for a kidney, and federal agencies put the number dying each day while listed at around seventeen. The United States performed a record of more than forty-six thousand transplants in 2023 and volumes have continued to rise since, and yet the list has not shortened, because listing rates rise with capacity.

The list is a poor measure of need. In the United States more than half a million people are on dialysis, several times the number registered for a kidney; the rest are excluded by age, comorbidity, insurance status, or a referral that never happened. For livers, hearts and lungs the equivalent unlisted populations are smaller but real. Internationally the picture is worse: the World Health Organization's donation and transplantation observatory estimates that transplantation meets under a tenth of global need, and many countries have essentially no deceased-donor programme.

## Why supply is structurally limited

Deceased organ donation depends on a rare mode of death. Transplantable organs require perfusion until recovery, which in practice means the donor died in an intensive care unit, either after a determination of death by neurological criteria while a ventilator maintained circulation, or after planned withdrawal of life support with rapid recovery afterwards. Only a small percentage of all deaths occur in circumstances that permit either. Deaths at home, in the community, or after prolonged illness almost never yield organs.

This is what distinguishes organs from blood, the other donated tissue: a healthy adult can give blood repeatedly, which is why the response to blood scarcity has been recruitment and, more recently, the manufactured substitutes discussed in [[artificial-blood]], while the response to organ scarcity cannot be either.

That biological ceiling means consent rates, however improved, operate on a small denominator. Even a country converting every eligible family to donation would not produce enough kidneys to clear its dialysis population, which is why living donation and non-human sources matter.

Organ quality compounds the problem. Donors are older and more comorbid than they were a generation ago, and marginal organs are more likely to be declined. Roughly one in five kidneys recovered in the United States is discarded rather than transplanted, a rate substantially higher than in comparable European systems and one that reflects allocation logistics, cold ischaemia time accumulating during offers, and liability incentives as much as organ quality.[^stewart2017]

## Policy responses

**Consent systems.** Presumed consent, or opt-out, is the most discussed reform: donation is assumed unless a person has registered a refusal. Wales adopted it in 2015, England in 2020, Scotland and the Netherlands soon after, and Nova Scotia became the first North American jurisdiction to do so. The evidence that the legal default drives donation rates is weaker than the policy's popularity suggests, and systematic reviews find the association confounded by health-system capacity and mortality patterns.[^rithalia2009] Spain has the world's highest deceased-donor rate and an opt-out law, but its rates only rose sharply a decade after the law, following the creation of a national coordination organisation that placed trained transplant coordinators in intensive care units and standardised the family approach.[^matesanz2011] Practitioners generally attribute the Spanish result to that infrastructure rather than to the default rule, and in most opt-out countries families are still consulted and can refuse.

**Living donation.** Living donors supply a large share of kidneys and a smaller share of livers. Paired exchange programmes match incompatible donor-recipient pairs into chains, sometimes dozens long, initiated by a non-directed donor. These have expanded access substantially without new donors, by better matching the ones who exist. [[uterus-transplantation|Uterus transplantation]] draws on both living and deceased donors and is unusual in being planned as temporary: the graft is removed once it has supported a pregnancy, so that immunosuppression can stop.

**Allocation.** How organs are ranked among candidates is a distributional question with no neutral answer: weighting medical urgency, expected years of graft survival, waiting time, and geography produce different winners. Continuous distribution frameworks that replace hard geographic boundaries with weighted scores have been introduced for several organs in the United States. Oversight of the American system itself was restructured by 2023 legislation that ended the single-contractor model under which one organisation had run the national network since the 1980s.

**Incentives.** United States law prohibits transfer of organs for valuable consideration, and most countries have similar bans. Reform proposals separate removing disincentives — reimbursing lost wages, travel and childcare for living donors, which is now permitted in the United States — from paying donors, which is not. Iran operates the only legal regulated market in living kidney donation and reports having eliminated its waiting list; critics point to evidence that donors are disproportionately poor and that follow-up care is inadequate. The Declaration of Istanbul, adopted in 2008 and revised since, sets the international position against organ trafficking and transplant tourism.

> [!debate] Does paying for organs increase supply or corrupt it
> Proponents of regulated compensation argue that the current system already pays everyone in the operating room except the person supplying the organ, and that a legal market with price floors and follow-up care would be safer than the black market that exists anyway. Opponents argue that payment converts a gift relationship into a transaction that will fall hardest on the poor, that valuation of body parts is corrosive regardless of safeguards, and that the Iranian experience shows exploitation rather than a solution. The disagreement is about values, not primarily about elasticity of supply.

## Technology responses

**Machine perfusion.** The largest recent gain has come from keeping organs alive outside the body rather than cold on ice. Hypothermic oxygenated perfusion and normothermic machine perfusion, which maintains a liver or heart at body temperature and metabolically active, extend the viable window, permit assessment of organ function before implantation, and allow marginal organs to be rescued; a randomised European trial found normothermic liver preservation reduced graft injury and discard compared with static cold storage.[^nasralla2018] Portable normothermic systems made routine donation after circulatory death feasible for hearts, an organ that had been recovered almost exclusively from brain-dead donors. The effect on supply has been material and immediate, and unlike engineered organs it required no new biology.

Normothermic regional perfusion, which restores circulation to the donor's abdominal and thoracic organs in situ after circulatory death while occluding the vessels supplying the brain, has raised objections in the United States about whether restoring circulation is compatible with the determination of death that preceded it.

**Reducing discard.** Better allocation software, earlier offers, and objective viability assessment during perfusion address the fifth of recovered kidneys that is thrown away, which is the cheapest available supply.

**New sources.** Pig organs from animals edited to remove antigens and add human regulatory proteins have entered clinical trials, with the arithmetic of the kidney list as the explicit rationale. Engineered organs remain preclinical, for the reasons given in [[tissue-engineering]] and [[decellularized-scaffolds]]. Cell-scale products arrive sooner than organ-scale ones: transplanted islet clusters, built by methods close to those behind [[organoids]], can restore a specific function without replacing an anatomical structure, and reduce demand for whole-pancreas transplantation before they touch the kidney list.

## The demand side

Supply-side framing obscures that transplant demand is substantially preventable. The dominant causes of kidney failure are diabetes and hypertension; the dominant causes of liver failure now include metabolic dysfunction-associated steatotic liver disease, with viral hepatitis in decline since the arrival of curative hepatitis C therapy. Drugs that slow chronic kidney disease progression, and the metabolic effects of [[glp-1-receptor-agonists|newer weight and glucose therapies]], may bend demand curves more than any donation reform. Curative treatment of inherited metabolic liver disease by [[somatic-gene-therapy]] removes a smaller but real share of the paediatric transplant burden.

The furthest version of the demand-side argument belongs to the [[geroscience-hypothesis]]: organ failure is concentrated in later life, so an intervention that slowed the underlying aging process would reduce the incidence of every form of it at once. That claim is unproven in humans, and its economic version is set out in [[longevity-dividend]]. It also cuts the other way, since populations that live longer accumulate more organ failure in absolute terms unless morbidity compresses as well, the disputed proposition examined in [[compression-of-morbidity]]. A transplant system whose list stops growing because fewer organs fail would be a better outcome than one that grows its supply to match.

Whether that happens is the open question. Every previous expansion of supply — expanded criteria donors, donation after circulatory death, machine perfusion — has been absorbed by increased listing rather than a shorter list, which is what makes the shortage structural rather than a temporary deficit, and what keeps [[access-and-inequality]] central to any discussion of who receives a manufactured organ if one ever arrives.

## See also

- [[xenotransplantation]]
- [[lab-grown-organs]]
- [[organ-bioprinting]]
- [[artificial-heart]]
- [[decellularized-scaffolds]]
- [[access-and-inequality]]
- [[tissue-engineering]]
- [[artificial-blood]]

## References

[^stewart2017]: `paper` Stewart, D. E. et al. "Diagnosing the decades-long rise in the deceased donor kidney discard rate in the United States." *Transplantation*, 2017.
[^rithalia2009]: `paper` Rithalia, A. et al. "Impact of presumed consent for organ donation on donation rates: a systematic review." *BMJ*, 2009.
[^matesanz2011]: `paper` Matesanz, R. et al. "Spanish experience as a leading country: what kind of measures were taken?" *Transplant International*, 2011. {Written from inside the Spanish national transplant organisation about its own programme; it is a descriptive account, not a controlled comparison with other countries.}
[^nasralla2018]: `paper` Nasralla, D. et al. "A randomized trial of normothermic preservation in liver transplantation." *Nature*, 2018. {A multicentre European randomized trial whose primary endpoint was a biochemical marker of graft injury; patient survival was not what it was designed to test.}

==============================================================================
ARTICLE: organoids
TITLE:   Organoids
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/organoids
SOURCE:  https://futurehumanwiki.com/raw/organoids
==============================================================================

---
title: "Organoids"
slug: "organoids"
type: "technology"
status: "emerging"
horizon: "present"
trl: 5
categories: ["bodies"]
tags: ["stem cells", "disease modelling", "brain organoids", "drug screening", "regenerative medicine", "bioethics"]
summary: "Self-organizing tissue cultures grown from stem cells that reproduce some of the architecture and function of an organ, used mainly for disease modelling rather than transplantation."
updated: "2026-07-27"
humanEvidence: "Organoids grown from a patient's own biopsy are used in the Netherlands to predict that patient's response to cystic fibrosis drugs; organoid-derived cells have been transplanted only in early trials, and no organoid has replaced an organ."
access: "A laboratory product rather than a treatment: the culture reagents are sold commercially, and the only patient-facing uses are the Dutch cystic fibrosis drug-response assay and a few early transplantation trials."
reversibility: "difficult"
issues: ["The Japanese ulcerative-colitis organoid trial is described without a source.", "The comparison of organoid oscillations to preterm-infant EEG is unattributed."]
---

```infobox
{
  "caption": "Cell culture technology",
  "rows": [
    { "label": "First modern example", "value": "Intestinal organoid, 2009" },
    { "label": "Cell sources", "value": "Adult stem cells, iPSCs" },
    { "label": "Typical size", "value": "0.1–4 mm" },
    { "label": "Common matrix", "value": "Tumour-derived basement membrane gel" },
    { "label": "Main use", "value": "Disease modelling, drug screening" },
    { "label": "Key limits", "value": "No vasculature, fetal maturity" },
    { "label": "Status", "value": "Standard research tool" }
  ]
}
```

**Organoids** are three-dimensional cell cultures, grown from stem cells, that self-organize into structures reproducing some of the cell types, spatial arrangement and function of a real organ. They are not miniature organs. A cerebral organoid has cortical layers and firing neurons but no blood supply, no immune cells, no thalamic input and no consistent body plan; an intestinal organoid has crypts and absorptive cells but no nerves and no microbiome. What makes them valuable is that they are human, three-dimensional, and generated in numbers large enough to screen against.

## How they are made

Two derivation routes dominate. The adult-stem-cell route takes tissue-resident stem cells — from an intestinal biopsy, say — and embeds them in a laminin-rich basement membrane gel with a defined cocktail of niche factors, typically including R-spondin, epidermal growth factor, and inhibitors of BMP and TGF-β signalling. The cells proliferate and spontaneously build crypt-villus architecture. These organoids are epithelial only, genetically stable, and can be passaged for years.

The pluripotent route starts from embryonic stem cells or from [[induced-pluripotent-stem-cells]] generated with the [[yamanaka-factors]], and applies a timed sequence of differentiation cues that mimics embryonic patterning: germ-layer induction, then regional specification, then self-organization. This route produces organoids containing multiple germ layers, including neurons and glia, but is slower, more variable, and yields tissue of fetal maturity.

Both depend on a permissive matrix. The standard reagent is a basement-membrane extract derived from a mouse tumour line, which is chemically undefined and varies between lots. Replacing it with synthetic hydrogels of tuned stiffness and defined adhesion ligands has been a persistent goal, partly for reproducibility and partly because an undefined animal-derived product is a poor foundation for anything clinical. Matrix stripped from donor tissue, described in [[decellularized-scaffolds]], is a third option that trades definition for biological fidelity.

## Development history

```timeline
[
  { "year": "1907", "title": "Self-organization observed", "text": "Henry Van Peters Wilson shows that dissociated sponge cells reaggregate into a functioning organism, the first demonstration that tissue architecture can arise from cells alone." },
  { "year": "2009", "title": "Intestinal organoids", "text": "Toshiro Sato and Hans Clevers show that a single Lgr5-positive intestinal stem cell builds crypt-villus structures in a gel without any supporting mesenchyme." },
  { "year": "2011", "title": "Optic cup", "text": "Yoshiki Sasai's group reports that mouse embryonic stem cells spontaneously form a stratified optic cup in culture, establishing that complex organ morphogenesis can be intrinsic." },
  { "year": "2013", "title": "Cerebral organoids and liver buds", "text": "Madeline Lancaster and Jürgen Knoblich use cerebral organoids to model microcephaly, and Takanori Takebe's group generates vascularizable liver buds from iPSCs." },
  { "year": "2015", "title": "Kidney organoids", "text": "Melissa Little's group produces organoids containing nephrons with segmented tubules, along with endothelium and stroma." },
  { "year": "2022", "title": "Integration into a host brain", "text": "Human cortical organoids transplanted into the somatosensory cortex of newborn rats grow, receive sensory input and can influence the animals' behaviour." }
]
```

## What they are actually used for

The dominant applications are diagnostic and pharmacological rather than therapeutic.

**Disease modelling.** Organoids derived from a patient carry that patient's genome, which makes them a testbed for monogenic disease. Cerebral organoids from microcephaly patients were the first widely cited case; they have since been used to study Zika virus neurotropism, and organoids of intestine, lung, liver and kidney model cystic fibrosis, polycystic kidney disease and metabolic disorders.

**Functional drug testing.** The clearest clinical use is in cystic fibrosis. Rectal biopsy organoids from an individual patient swell in response to forskolin only if CFTR is functional, and the magnitude of swelling predicts response to CFTR modulator drugs.[^dekkers2013] Health systems in the Netherlands have used this assay to grant access to expensive modulators for patients whose mutations are too rare to have been included in registration trials. This is personalised medicine in the literal sense: the test is run on the patient's own tissue rather than on a population average. Patient organoids were also among the first human tissues in which a disease mutation was corrected by [[crispr-cas9]], when the CFTR defect was repaired in intestinal organoids from cystic fibrosis patients and the corrected cultures regained function.[^schwank2013]

**Cancer biology.** Tumour organoid biobanks preserve the genetic and histological diversity of patient tumours far better than immortalized cell lines, and are used for drug sensitivity screening and for studying resistance.

**Toxicology.** Liver, cardiac and kidney organoids and organ-on-chip devices are being adopted for safety screening, a shift encouraged by regulatory moves away from mandatory animal testing for some product classes.

```compare
{
  "columns": ["Organoid", "Organ-on-chip", "Animal model"],
  "rows": [
    { "label": "Species", "values": ["Human", "Human", "Non-human"] },
    { "label": "Self-organizing architecture", "values": ["Yes", "Imposed by device", "Native"] },
    { "label": "Perfusion", "values": ["None", "Engineered flow", "Native circulation"] },
    { "label": "Immune and neural input", "values": ["Absent", "Usually absent", "Present"] },
    { "label": "Throughput", "values": ["High", "Moderate", "Low"] },
    { "label": "Maturity of tissue", "values": ["Fetal", "Fetal to variable", "Adult"] }
  ]
}
```

## The ceilings

**No vasculature.** An organoid grows until diffusion fails, which for most tissue types means a diameter of a few hundred micrometres to a couple of millimetres before a necrotic core develops. This caps size, distorts internal gradients, and prevents the perfusion-dependent maturation that real organs undergo. Approaches include co-culture with endothelial cells, transplantation into a host animal so that host vessels invade, and assembly into printed vascular frameworks as described in [[organ-bioprinting]].

**Fetal maturity.** Transcriptionally and functionally, most organoids resemble first- or second-trimester tissue. Cerebral organoids do not develop adult cortical identity, and analyses have reported that cells in cortical organoids carry a metabolic stress signature that impairs proper subtype specification, a caution against reading them as faithful models of the mature brain.[^bhaduri2020] Prolonged culture improves maturity slowly and unevenly.

**Variability.** Batch-to-batch and line-to-line differences are substantial, particularly for the pluripotent route, where the same protocol can yield organoids of different regional identity. This limits statistical power and complicates use as a screening platform.

**Missing components.** Organoids typically lack resident immune cells, vasculature, innervation and mechanical loading. Assembloids, in which separately patterned organoids are fused, restore some of what is missing — cortical and subpallial spheroids fused together permit interneuron migration to be observed directly — but the reconstruction is partial.

## Brain organoids and the consciousness question

Cerebral organoids generate spontaneous electrical activity, and cultured long enough they produce synchronized network oscillations. One widely discussed report compared the developmental trajectory of these oscillations to electroencephalographic features seen in preterm infants, a comparison the authors themselves qualified and that other researchers regarded as an artefact of the analysis rather than evidence of anything experiential.

The mainstream position is that current brain organoids are extremely unlikely to be conscious. They lack sensory input, lack the thalamocortical loops that most theories of consciousness treat as necessary, lack a body to act on, and are structurally disorganized compared to a real cortex. That said, the question is not empty, because there is no accepted test. Assessing organoid sentience runs into exactly the problem described in [[neural-correlates-of-consciousness]] and [[machine-consciousness]]: behavioural evidence is unavailable and theory-based indicators disagree with one another. Ethicists have argued for developing criteria in advance rather than after a disputed result — an application of the [[precautionary-principle]] to a research programme rather than to a deployed technology — and professional guidance so far treats organoid work as requiring standard rather than special oversight while flagging the issue as one to revisit.[^farahany2018] The situation differs from [[whole-brain-emulation]], where the object under discussion is a model of a specific existing brain; an organoid is an anonymous piece of developing tissue with no history and no prior mental content.

Transplantation sharpens it. Human cortical organoids grafted into rat cortex have been shown to integrate into host circuits, respond to whisker stimulation, and drive learned behaviour, which raises questions about the moral status of the resulting chimeric animal rather than of the organoid alone.[^revah2022] Related questions attach to [[synthetic-embryos]], where the definitional problem is what counts as an embryo rather than what counts as a mind.

> [!note] Terminology
> "Organoid" is used loosely. It covers epithelial-only cultures with strict lineage fidelity, multi-lineage constructs from pluripotent cells, and assemblies fused from several pieces. Claims about what organoids can do rarely generalize across these categories. Cell collectives that self-organize into forms with no organ counterpart, such as the frog-cell constructs described in [[xenobots]], fall outside the term entirely.

## Outlook

The therapeutic use of organoids is beginning at the margins, and it does not look like transplanting an organoid as such. An early Japanese trial has transplanted intestinal organoid-derived cells into the ulcerated bowel of patients with ulcerative colitis, using the culture as a way to expand a patient's own epithelium rather than as a structure. Pancreatic islet clusters grown from stem cells are built by comparable methods and have relieved insulin dependence in early type 1 diabetes trials. In both cases a few million cells placed in the right site do useful work, and the three-dimensional architecture that defines an organoid in the laboratory is largely incidental to the therapy.

The larger ambition, using organoids as building blocks for [[lab-grown-organs]], depends on solving perfusion and maturation, the same two problems that limit [[tissue-engineering]] generally. Little in the current work is likely to bear on the [[organ-shortage]] within this decade. What organoids have already changed is the epistemics of human biology: for the first time, a laboratory can run a controlled experiment on human tissue of a specified genotype, repeatedly, without a patient in the room.

## See also

- [[tissue-engineering]]
- [[organ-bioprinting]]
- [[lab-grown-organs]]
- [[induced-pluripotent-stem-cells]]
- [[synthetic-embryos]]
- [[machine-consciousness]]
- [[decellularized-scaffolds]]
- [[limb-regeneration]]

## References

[^dekkers2013]: `paper` Dekkers, J. F. et al. "A functional CFTR assay using primary cystic fibrosis intestinal organoids." *Nature Medicine*, 2013.
[^bhaduri2020]: `paper` Bhaduri, A. et al. "Cell stress in cortical organoids impairs molecular subtype specification." *Nature*, 2020.
[^schwank2013]: `paper` Schwank, G. et al. "Functional repair of CFTR by CRISPR/Cas9 in intestinal stem cell organoids of cystic fibrosis patients." *Cell Stem Cell*, 2013. {The correction was made in cultured stem cells taken from patients; nothing was returned to a patient, and no clinical benefit was tested.}
[^farahany2018]: `paper` Farahany, N. A. et al. "The ethics of experimenting with human brain tissue." *Nature*, 2018.
[^revah2022]: `paper` Revah, O. et al. "Maturation and circuit integration of transplanted human cortical organoids." *Nature*, 2022. {Human tissue grafted into newborn rats; the behavioural result is a property of the rat, and the paper makes no claim about organoid experience.}

==============================================================================
ARTICLE: osseointegration
TITLE:   Osseointegration
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/osseointegration
SOURCE:  https://futurehumanwiki.com/raw/osseointegration
==============================================================================

---
title: "Osseointegration"
slug: "osseointegration"
type: "technology"
status: "established"
horizon: "present"
trl: 8
categories: ["bodies"]
tags: ["prosthetics", "implants", "titanium", "amputation", "bone", "biomaterials"]
summary: "The direct structural fusion of living bone to a titanium implant, the biological principle underlying dental implants and skeletally anchored limb prostheses."
updated: "2026-07-27"
humanEvidence: "Dental implants have decades of routine clinical use in millions of patients; bone-anchored limbs rest on uncontrolled patient series reporting longer prosthesis wear time and better mobility, not on randomized trials."
access: "Dental implants are a routine commercial procedure worldwide; bone-anchored limbs are performed at a small number of specialist centres, and the US screw-fixated system entered under a humanitarian exemption that caps volume."
reversibility: "difficult"
issues: ["The claim that most recipients have at least one superficial infection carries no source."]
---

```infobox
{
  "caption": "Biomaterial phenomenon and surgical technique",
  "rows": [
    { "label": "Discovered", "value": "1952, Lund" },
    { "label": "Discovered by", "value": "Per-Ingvar Brånemark" },
    { "label": "First dental patient", "value": "1965" },
    { "label": "First bone-anchored limb", "value": "1990, Sweden" },
    { "label": "Standard material", "value": "Commercially pure titanium" },
    { "label": "US limb approval", "value": "2015, humanitarian exemption" },
    { "label": "Principal complication", "value": "Skin-penetration site infection" },
    { "label": "Readiness", "value": "TRL 9 dental, TRL 8 limb" }
  ]
}
```

**Osseointegration** is the formation of a direct, load-bearing connection between living bone and
the surface of an implanted material, without intervening fibrous tissue. It is the reason a dental
implant can carry chewing forces for decades, and it is the enabling principle for prostheses
anchored to the skeleton rather than strapped to it. The phenomenon was found by accident, in an
experiment about blood flow, and its extension from teeth to limbs took a further quarter of a
century.

## How it works

Bone ordinarily walls off foreign material with a fibrous capsule, which under load becomes a
loosening interface. Titanium behaves differently. A spontaneous, self-repairing oxide layer a few
nanometres thick forms on its surface, adsorbs proteins in a configuration osteoblasts can attach
to, and provokes little inflammatory response. Bone then forms directly against the
oxide.[^branemark1983]

Two processes contribute. In distance osteogenesis, new bone grows from the existing surface toward
the implant; in contact osteogenesis, osteogenic cells migrate onto the implant surface and lay down
matrix outward from it. Surface topography strongly influences which dominates, which is why modern
implants are roughened by grit-blasting, acid-etching, or anodizing rather than left machined
smooth.

Fixation has two phases that are often confused. *Primary stability* is mechanical: the press-fit or
screw-thread grip of the implant in prepared bone at the moment of insertion. *Secondary stability*
is biological: the bone that forms and remodels around the implant over subsequent months. Primary
stability declines as the surgical trauma resolves while secondary stability rises, and the dip
between them is why loading protocols are staged over weeks to months. An implant loaded too early
or too heavily during that window develops micromotion, fibrous encapsulation, and
failure.[^albrektsson1981]

Stress shielding sets a longer-term constraint. Bone remodels according to the loads it carries, and
a stiff implant that carries load the bone used to carry causes local resorption. Implant geometry
and stiffness are chosen to keep enough load in the bone to maintain it.

## Discovery and development

```timeline
[
  { "year": "1952", "title": "Accidental discovery", "text": "Per-Ingvar Brånemark implants titanium optical chambers in rabbit bone to study microcirculation and finds at the end of the study that the chambers cannot be removed; bone has fused to the metal." },
  { "year": "1965", "title": "First dental application", "text": "Brånemark places titanium fixtures in the jaw of Gösta Larsson, who had a cleft palate and no lower teeth; the implants remain functional for the rest of his life." },
  { "year": "1977–1980s", "title": "Acceptance and extension", "text": "The dental establishment, initially sceptical, adopts the technique after long-term series are published; bone-anchored hearing aids and craniofacial prostheses follow." },
  { "year": "1990", "title": "First bone-anchored limb", "text": "Rickard Brånemark performs the first transfemoral osseointegrated prosthesis procedure in Gothenburg, using a screw-fixated design later commercialized as OPRA." },
  { "year": "1999–2010s", "title": "Press-fit designs", "text": "German and Australian groups develop press-fit implants that shorten the two-stage protocol and broaden the eligible population." },
  { "year": "2015", "title": "US regulatory entry", "text": "The screw-fixated limb system receives a humanitarian device exemption in the United States, limiting the number of procedures but establishing a legal route." },
  { "year": "2020", "title": "Integration with neural interfaces", "text": "Self-contained arm prostheses combine skeletal attachment with implanted electrodes for muscle recording and nerve stimulation, used at home over years." }
]
```

## Bone-anchored limb prostheses

A conventional prosthetic limb transmits load through a socket that grips the soft tissue of the
residual limb. Soft tissue is a poor structural interface. It changes volume through the day and
across seasons, it develops pressure ulcers, folliculitis, and heat rash, and the pistoning motion
between socket and bone wastes energy and blurs the wearer's sense of where the limb is. Sockets are
the single most common reason for prosthesis dissatisfaction.

Skeletal attachment replaces the socket with a metal fixture implanted in the medullary canal of the
femur, tibia, or humerus, connected through a percutaneous abutment to the external prosthesis. Two
design families dominate: screw-fixated implants requiring two operations several months apart, and
press-fit implants with a porous or roughened surface that can be done in one or two stages with a
shorter protocol.

The functional gains are consistent across series. Prosthesis wear time increases, walking distance
and mobility scores improve, sitting is more comfortable, and hip range of motion is
unrestricted.[^hagberg2009]
Users also report **osseoperception**: mechanical vibration transmitted through the implant into
bone produces a crude but genuine sense of the ground and of contact with objects, a form of
feedback no socket provides and that no commercial [[myoelectric-prosthetics]] system delivers
electronically. It is the one channel in prosthetics where information flows back into the body
without any decoding step of the kind [[neural-decoding]] research is built around.

The interface also solves a problem for [[neuroprosthetics]]. A percutaneous abutment provides a
protected route for wires, allowing electrodes on nerves and muscles inside the residual limb to
connect to external electronics without a separate skin-crossing site. Long-term implanted arm
prostheses using this route have combined direct skeletal loading, myoelectric control from
implanted electrodes, and nerve stimulation for touch.[^ortizcatalan2020]

## Risks at the skin interface

The permanent breach in the skin is the technology's defining hazard. Skin does not seal against
metal the way bone bonds to it; the epithelium migrates down the abutment and the site remains a
potential entry route for bacteria for the life of the implant. Superficial infection requiring oral
antibiotics is common enough that most recipients experience at least one episode. Deep infection
and osteomyelitis are much less common but can require implant removal and further bone loss.

Other complications include periprosthetic fracture — a fall transmits force directly into the
femur rather than being absorbed by a socket, so a designed mechanical fuse in the connector is
standard — abutment breakage, aseptic loosening, and soft-tissue redundancy around the stoma
requiring revision.

Candidacy is restricted. Adequate residual bone length and quality are required, and poorly
controlled diabetes, active infection, peripheral vascular disease, and heavy smoking are common
exclusions. That removes much of the world's amputee population, since dysvascular amputation
dominates in high-income countries and conflict-related amputation dominates in settings where the
surgery is unavailable at all, an instance of the pattern described in
[[access-and-inequality]]. High-impact activity is generally discouraged, which limits the technique
for younger, athletic amputees who would otherwise benefit most and keeps it distinct from the
performance question raised in [[enhancement-in-sport]]. The procedure also remains contested among
some amputees who regard socket problems as manageable and permanent hardware as an unnecessary
risk, a disagreement of the kind catalogued under [[disability-rights-and-enhancement]].

> [!caution] Not a solved interface
> Bone-to-metal integration is reliable. Skin-to-metal is not. Every proposal for permanently
> percutaneous hardware — limb anchors, transcutaneous power leads, the skull pedestals used in
> academic [[brain-computer-interface]] research — inherits the same unsolved problem, and it is a
> principal reason implanted systems are pushed toward fully wireless designs.

## Beyond limbs

Dental implants remain by far the largest application, performed in the millions annually. Bone-anchored
hearing devices route sound through the skull to the cochlea, bypassing a damaged outer or middle
ear, and sit alongside the [[cochlear-implant]] in the hearing-restoration toolkit for a distinct
patient group. Craniofacial prostheses — ears, orbits, noses — are retained on osseointegrated
abutments where adhesive retention fails. Orthopaedic joint replacements rely on the same
bone-implant biology, using porous coatings for cementless fixation. Skeletal anchorage has also
been proposed as a load path for a wearable [[exoskeleton]], transferring force into bone rather
than compressing soft tissue; no such system has been built for a person.

The principle also constrains alternatives. Engineered bone grafts under
[[tissue-engineering]] and acellular matrices of the kind described in
[[decellularized-scaffolds]] aim at biological rather than metallic integration, and would remove
the percutaneous problem entirely if a limb could be regrown; nothing in
[[limb-regeneration]] research suggests that is available for mammals.

## Outlook

Incremental work is directed at the skin interface: porous or textured abutment surfaces intended to
allow soft-tissue attachment, antibacterial coatings, silver and iodine surface treatments, and
implant geometries that reduce motion at the stoma. None has yet converted infection from a managed
recurring problem into a solved one.

The broader trajectory is convergence. Skeletal attachment, implanted neural electrodes, and powered
actuators are being combined into single systems, so that the prosthesis is loaded through bone,
controlled by nerve, and provides feedback through both. That combination is what makes a prosthetic
limb behave less like a tool and more like a limb, and it also sharpens the question raised under
[[morphological-freedom]] and [[human-enhancement]]: whether people without amputations should ever
be offered a permanent skeletal port, and on what evidence a surgeon could justify one.

## See also

- [[myoelectric-prosthetics]]
- [[neuroprosthetics]]
- [[exoskeleton]]
- [[limb-regeneration]]
- [[cochlear-implant]]
- [[tissue-engineering]]
- [[disability-rights-and-enhancement]]
- [[access-and-inequality]]

## References

[^branemark1983]: `paper` Brånemark, P.-I. "Osseointegration and its experimental background." *Journal of Prosthetic Dentistry*, 1983. {The discoverer's own account of his laboratory and clinical experience, written for a dental audience; it is a synthesis rather than an independent test of the phenomenon.}

[^albrektsson1981]: `paper` Albrektsson, T., Brånemark, P.-I., Hansson, H.-A., Lindström, J. "Osseointegrated titanium implants: requirements for ensuring a long-lasting, direct bone-to-implant anchorage in man." *Acta Orthopaedica Scandinavica*, 1981.

[^hagberg2009]: `paper` Hagberg, K., Brånemark, R. "One hundred patients treated with osseointegrated transfemoral amputation prostheses: rehabilitation perspective." *Journal of Rehabilitation Research and Development*, 2009. {An uncontrolled series from the centre that developed the implant, with patient-reported mobility and comfort as the outcomes.}

[^ortizcatalan2020]: `paper` Ortiz-Catalan, M. et al. "Self-contained neuromusculoskeletal arm prostheses." *New England Journal of Medicine*, 2020.

==============================================================================
ARTICLE: overpopulation-and-longevity
TITLE:   Overpopulation and life extension
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/overpopulation-and-longevity
SOURCE:  https://futurehumanwiki.com/raw/overpopulation-and-longevity
==============================================================================

---
title: "Overpopulation and life extension"
slug: "overpopulation-and-longevity"
type: "concept"
status: "contested"
horizon: "2050s+"
categories: ["society", "longevity"]
tags: ["demography", "aging", "fertility", "resources", "policy", "ethics"]
summary: "The objection that extending human lifespan would cause unsustainable population growth, and the demographic modelling that finds the effect smaller than assumed."
updated: "2026-07-27"
issues: ["The claim that the top income decile accounts for about half of emissions is uncited."]
---

```infobox
{
  "caption": "Objection in the ethics of life extension",
  "rows": [
    { "label": "Claim", "value": "Longer lives, unsustainable population" },
    { "label": "Dominant demographic term", "value": "Fertility, not mortality" },
    { "label": "Global fertility rate", "value": "About 2.2, falling" },
    { "label": "Projected peak population", "value": "Mid-2080s, UN 2024 projection" },
    { "label": "Modelled effect of defeating aging", "value": "Modest, not explosive" },
    { "label": "Status", "value": "Widely made, weakly supported" }
  ]
}
```

**Overpopulation and life extension** concerns the most frequently raised objection to radical longevity research: that if people stopped dying of age-related causes, population would grow beyond what the planet can support. The objection is intuitive, is almost never made against curing any specific disease, and has been examined by demographers, whose conclusion is that the effect is real but far smaller than the intuition suggests — because population dynamics are driven by births far more than by deaths at old ages.

## The arithmetic

A population's growth rate depends on how many children are born and how many survive to have children of their own. Deaths after the end of the reproductive span remove people from the population without affecting the number of descendants they produce. Eliminating mortality at seventy therefore adds to the stock of living people without altering the reproductive dynamics at all.

The consequence is a change in the shape of growth, not its rate. Under a regime with constant births and no deaths, population grows linearly — adding roughly the same number each year — rather than exponentially, which is what compounding reproduction produces. Linear growth over a century is a large number; it is not the runaway curve the objection assumes.

The second term is fertility, and it dominates. A demographic analysis by Leonid Gavrilov and Natalia Gavrilova modelled scenarios including the complete elimination of aging and found that under fertility rates at or below replacement, the resulting population trajectories were considerably more moderate than the objection anticipates.[^gavrilov2010] The result is not a claim that population would not grow. It is that mortality is the smaller lever.

This holds a fortiori for the interventions actually under development. Nothing that targets the [[hallmarks-of-aging]] — not [[senolytics]], not the candidate geroprotectors reviewed under the [[geroscience-hypothesis]] — is expected to eliminate old-age mortality. The realistic near-term outcome is a compression or extension of the final decades, and whether it is compression or extension is the question [[compression-of-morbidity]] examines.

> [!stat] Where the growth actually comes from
> Most projected population growth between now and 2100 comes from two sources: the momentum of young age structures in populations where fertility has recently fallen, and continued high fertility in a small number of countries. Neither is affected by geroprotective medicine.

## The fertility trend

The demographic context has changed substantially since the objection was formulated in the 1990s and 2000s, when global population was still growing at close to its peak rate.

The United Nations' 2024 projection puts world population at roughly 8.2 billion, growing more slowly than at any point since the mid-twentieth century, and expects it to peak around 10.3 billion in the mid-2080s before declining.[^unwpp2024] The global total fertility rate is approximately 2.2 and falling; more than half of all countries and areas are already below the replacement level of about 2.1. South Korea's rate, the world's lowest, fell below 0.75 — a figure that implies each generation being roughly a third the size of the one before it if sustained.

This inverts the policy problem in most high-income and many middle-income countries. Their concern is not that people will live too long but that too few will be born to support those who do, which is the anxiety behind the pension arithmetic examined in [[longevity-dividend]]. In that context a geroprotective intervention that extends working life is a partial remedy for a demographic problem rather than a cause of one.

The historical precedent supports caution about extrapolation in either direction. Predictions of mass famine from population growth made in the late 1960s failed, partly because agricultural yields rose faster than forecast and partly because fertility fell faster than forecast. The same forecasting difficulty applies to the projections above.

## What would actually change the picture

Three scenarios would make the objection bite, and they are distinguishable.

**Reproductive extension.** If the interventions described in [[reproductive-longevity]] substantially widened the fertile window, the total number of children a person could have would rise, which affects the term that matters. This is a plausible near-term development and receives almost no attention in the overpopulation debate, which focuses on the end of life rather than the middle.

**Very large lifespan gains sustained over centuries.** Linear growth is manageable over decades and unmanageable over millennia. Any scenario approaching [[longevity-escape-velocity]] would eventually require fertility well below replacement, indefinitely — a constraint on reproductive choice that its proponents rarely spell out. The same trade-off appears in miniature in [[generation-ship-biology]], where a fixed carrying capacity forces an explicit birth policy. Nothing in the record of extreme longevity described in [[maximum-human-lifespan]] suggests such gains are near, but the objection is about the scenario, not the timeline.

**Uneven adoption.** If life extension were available only in low-fertility high-consumption countries, aggregate population effects would be small and aggregate resource effects would be larger than the headcount suggests.

## Resources and consumption

The objection is often stated in terms of headcount when the operative variable is consumption, which is distributed far more unequally than population. Estimates from consumption-based emissions accounting attribute roughly half of global emissions to the highest-earning tenth of the world's population. A billion additional people at the consumption level of the global median would have a smaller environmental effect than a much smaller increase concentrated among the affluent — which is the group that would receive expensive longevity medicine first, on the analysis in [[access-and-inequality]].

Estimates of Earth's human carrying capacity are not converging. A survey of the published literature by Joel Cohen found estimates spanning more than three orders of magnitude, clustering loosely around ten to twelve billion but with no agreement on the assumptions that generate any particular figure.[^cohen1995] The spread reflects the fact that carrying capacity is a function of technology, diet, and distribution rather than a property of the planet.

## What the objection is really doing

Several writers have noted that the overpopulation argument is applied selectively. It is not raised against paediatric vaccination, cancer treatment, or cardiac surgery, all of which add life-years to the population by the same mechanism. Aubrey de Grey's standard reply is that the objection reduces to a claim that some people should die so that others may be born, which few of its users would endorse when stated directly; [[aubrey-de-grey]] covers his framing of the argument.

The more charitable reading is that the objection is a proxy for three concerns it states badly. One is distributive: a worry about who receives the extra years, which is a real problem and is not about population size. A second is about quality — that added years will be years of dependency rather than of the [[healthspan]] the field promises, which is an argument about the intervention rather than about the population, and which shades into the pressures examined in [[right-to-die-and-right-to-live]]. The third is about stagnation — a society in which the same people hold positions, property, and authority for centuries, closing off the generational turnover through which institutions change. That is a substantive worry and has nothing to do with resources.

> [!debate] The generational-turnover version
> Max Planck's observation that science advances one funeral at a time is the compressed form of this argument. Whether institutional renewal genuinely depends on mortality, or merely on the mobility that mortality happens to produce, is an open question that the demographic literature does not address and that the longevity field has largely ignored.

## Where it stands

The demographic case against the objection is reasonably strong, and it depends on an assumption worth stating plainly: that fertility remains at or below current levels. Fertility is a behavioural variable, responsive to policy and to circumstance, and its recent decline has surprised demographers repeatedly. A society that solved aging and then chose to reverse its fertility decline would face exactly the problem the objection describes, on a timescale of centuries rather than decades. Nothing in the geroscience literature engages with that scenario, and nothing in the demographic literature engages with a population in which the mortality assumption no longer holds.

## See also

- [[longevity-escape-velocity]]
- [[longevity-dividend]]
- [[maximum-human-lifespan]]
- [[compression-of-morbidity]]
- [[reproductive-longevity]]
- [[access-and-inequality]]
- [[right-to-die-and-right-to-live]]
- [[geroscience-hypothesis]]

## References

[^gavrilov2010]: `paper` Gavrilov, L.A. and Gavrilova, N.S. "Demographic Consequences of Defeating Aging." *Rejuvenation Research*, 2010. {A modelling exercise, not a projection: its moderate trajectories depend on fertility remaining at or below replacement, which is the assumption doing the work.}
[^cohen1995]: `book` Cohen, J.E. *How Many People Can the Earth Support?* W. W. Norton, 1995.
[^unwpp2024]: `report` United Nations Department of Economic and Social Affairs, Population Division. *World Population Prospects 2024*. United Nations, 2024. {A projection with wide uncertainty intervals; the 2024 revision lowered the expected peak relative to the preceding one as fertility fell faster than forecast.}

==============================================================================
ARTICLE: pantropy
TITLE:   Pantropy
PORTAL:  Space & Extreme Environments
URL:     https://futurehumanwiki.com/wiki/pantropy
SOURCE:  https://futurehumanwiki.com/raw/pantropy
==============================================================================

---
title: "Pantropy"
slug: "pantropy"
type: "concept"
status: "speculative"
horizon: "2050s+"
categories: ["space", "enhancement"]
tags: ["space settlement", "germline", "adaptation", "terraforming", "science fiction", "ethics"]
summary: "The proposal to engineer humans to suit other worlds rather than engineering those worlds to suit humans, and the ethical problems of designing offspring for an environment."
updated: "2026-07-27"
issues: ["Bernal 1929 and Mason 2021 are discussed in the text but carry no citations."]
---

```infobox
{
  "caption": "Speculative approach to space settlement",
  "rows": [
    { "label": "Coined by", "value": "James Blish, 1950s" },
    { "label": "Etymology", "value": "Greek pan + trope" },
    { "label": "Opposed to", "value": "Terraforming" },
    { "label": "Mechanism required", "value": "Heritable genetic modification" },
    { "label": "Demonstrated", "value": "No" },
    { "label": "Nearest real analogue", "value": "High-altitude human adaptation" }
  ]
}
```

**Pantropy** is the proposal to modify humans so that they can live on other worlds without extensive artificial habitats, in contrast to terraforming, which modifies the world to suit unmodified humans. The word was coined by the science fiction writer James Blish, whose *The Seedling Stars* stories imagined "adapted men" engineered for environments as hostile as the surface of Ganymede and as small as a freshwater puddle. As a research programme it does not exist. As a framing device it recurs constantly in discussions of Mars settlement, and it is best understood as [[germline-editing|heritable genome editing]] with a destination attached.

## Origins

The idea predates the word. In 1929, J.D. Bernal argued in *The World, the Flesh and the Devil* that humans would eventually rebuild their own bodies to suit conditions off Earth, treating the biological form as an engineering constraint rather than a given. Three decades later, Manfred Clynes and Nathan Kline coined the term "cyborg" for precisely this purpose: their 1960 paper proposed augmenting human physiology so that astronauts could function in space without carrying a terrestrial environment with them.[^clynes1960] That the word now associated with [[brain-computer-interface|neural implants]] began as a space-medicine proposal is a useful reminder of how the two literatures are connected.

Blish supplied the term and the contrast. Writing in the same decade that Jack Williamson's coinage of "terraforming" entered circulation, he set the two approaches against each other as rival colonisation strategies, and noted the asymmetry that still defines the debate: terraforming is enormously expensive but leaves people unchanged, while pantropy is cheap by comparison and changes them permanently.

```timeline
[
  { "year": "1929", "title": "Bernal's proposal", "text": "J.D. Bernal argues that space settlement would require redesigning the human body, not merely transporting it." },
  { "year": "1942", "title": "Terraforming named", "text": "Jack Williamson coins the opposing term in a science fiction story, establishing the world-modifying alternative." },
  { "year": "1952–1957", "title": "Blish's adapted men", "text": "James Blish's pantropy stories, collected as The Seedling Stars, describe humans engineered for alien environments and coin the term." },
  { "year": "1960", "title": "Cyborgs and Space", "text": "Clynes and Kline propose pharmacological and mechanical augmentation of astronauts, introducing the word cyborg for the concept." },
  { "year": "2014", "title": "A natural worked example", "text": "The Tibetan EPAS1 high-altitude haplotype is traced to Denisovan introgression, demonstrating that human adaptation to an extreme environment can be genetically specific." },
  { "year": "2018", "title": "Terraforming Mars questioned", "text": "An inventory of accessible Martian carbon dioxide concludes there is not nearly enough to thicken the atmosphere with foreseeable technology." },
  { "year": "2021", "title": "A modern argument", "text": "Christopher Mason's The Next 500 Years makes an explicit case that engineering humans for other worlds is not merely permissible but obligatory." }
]
```

## Pantropy against terraforming

The comparison has become less academic. An inventory of carbon dioxide accessible in Martian polar caps, dust and minerals concluded that even releasing all of it would raise atmospheric pressure only a fraction of the way toward habitability, and that no foreseeable technology can supply the rest.[^jakosky2018] Terraforming Mars, on that analysis, is not a matter of engineering effort but of missing inventory.

That result strengthens the pantropist's argument in one respect and does nothing for it in another. It removes the option of waiting for a habitable Mars. It does not establish that human biology could be modified to tolerate a surface with under one per cent of Earth's atmospheric pressure, unbreathable composition, no magnetic field, and lethal ultraviolet flux. No plausible edit makes a human survivable in a vacuum. Pantropy on Mars therefore reduces to something far more modest than Blish imagined: adaptation to partial gravity, chronic radiation, an altered light cycle, and life indoors — a set of pressures, not a new organism.

> [!debate] The technology-first objection
> Humans already occupy Antarctica, the deep ocean and orbit without any genetic change, using suits, habitats and supply chains. Critics argue this is the decisive precedent: engineering an environment is fast, reversible, and improves with each iteration, while engineering a genome is slow, irreversible in a population, and improves only across generations. Pantropists reply that habitat dependence is itself a permanent tax, and that a population which can never step outside is not a settlement but a permanently sustained expedition.

## What natural adaptation shows

Human populations have adapted genetically to extreme environments, which is the strongest empirical support pantropy has, and also the clearest indication of its scale.

Tibetans carry a variant of *EPAS1*, acquired by introgression from Denisovans, that keeps haemoglobin concentration low at altitude and avoids the chronic polycythaemia that afflicts unadapted populations.[^huerta2014] Andean highlanders solve the same problem differently, with elevated haemoglobin, showing that a single environmental pressure admits multiple genetic solutions. Bajau divers of Southeast Asia show enlarged spleens associated with a specific variant, plausibly increasing the oxygen reservoir available during breath-hold dives.[^ilardo2018] Arctic populations carry fatty-acid desaturase variants suited to a high marine-fat diet.

Each of these took on the order of thousands of years, involves a small number of loci, and produces a change in physiological tuning rather than a new capability. Nothing in the record of human adaptation supports the expectation that editing could deliver tolerance of vacuum, kilogray radiation doses, or a nitrogen-free atmosphere. Natural selection has already explored the accessible space, and it is narrow.

## Candidate modifications

Proposals that survive contact with biology are correspondingly limited.

**Radiation.** The most-discussed target, treated in [[radiation-hardening-humans]]. Candidates include the tardigrade damage-suppressor protein described in [[tardigrade-genes]] and additional tumour-suppressor copies of the kind found in elephants. Both remain cell-culture results, and the second carries a documented tradeoff with accelerated aging phenotypes in mice.

**Skeletal maintenance in low gravity.** If partial gravity turns out to be insufficient to maintain bone, sclerostin or myostatin pathway modification could raise the set point; [[myostatin-inhibition]] is already a clinical target for sarcopenia, and the reservations there — added mass without proportionate function — apply here too. The relevant physiology is covered in [[microgravity-adaptation]].

**Circadian entrainment.** The Martian sol runs about thirty-nine minutes longer than an Earth day, at the edge of what the human circadian system entrains to. Mission operations staff working Mars time have shown that most people can adapt with light management, which suggests this problem is solvable behaviourally and does not require genetics — a useful check on the impulse to reach for editing first. Where the analysis leads elsewhere is in the literature on [[sleep-reduction|engineered sleep need]], which faces the same question of whether a set point is compressible at all.

**Metabolic and dietary tolerance.** Modifying vitamin D synthesis, oxidative stress handling, or microbiome composition to suit a closed food system is technically nearer than the others, and near-useless without the food system itself; see [[closed-loop-life-support]].

## Objections

The technical objection is that nobody can specify the edits. Every trait proposed is polygenic, pleiotropic, or both, and the difficulties catalogued in [[genetic-enhancement-of-intelligence]] apply with equal force to radiation tolerance and bone maintenance. Germline modification of complex traits in humans is not a matter of waiting for better delivery; the target list does not exist.

The safety objection is the one raised against all heritable editing: mosaicism, off-target changes, and unknown interactions, discussed in [[crispr-off-target-effects]]. An edit made for Mars would be evaluated against an environment nobody has lived in, with no possibility of a control group.

The ethical objection is specific to pantropy and sharper than the general enhancement debate covered in [[bioethics-of-enhancement]]. A child engineered for low gravity and high radiation is engineered *away* from Earth. If the adaptation is significant, returning may be uncomfortable, unhealthy, or impossible. The child's range of available lives has been narrowed by a decision made before conception, in a way that [[morphological-freedom]] arguments — which concern what adults may do to themselves — do not license. [[procreative-beneficence]] is an awkward fit as well: it is not obvious that engineering a child for Mars serves the child's interests rather than the settlement's.

> [!caution] Where the settlement's interest and the child's diverge
> Most enhancement debates assume the modified person benefits. Pantropy assumes a colony benefits from having members suited to it. Those coincide only if the person wants to stay. The disability-rights literature summarised in [[disability-rights-and-enhancement]] has long argued that framing a body as suited or unsuited to an environment obscures who set the environment, and the argument transfers directly.

## Outlook

Pantropy is not an active research programme and no institution funds it as such. What exists is a set of adjacent capabilities — germline editing, [[embryo-selection]], transgene expression in human cells — that could in principle be pointed at the problem, and a rhetorical tradition that treats doing so as the natural endpoint of settlement.

The near-term question is whether any modification is needed at all, and that depends on data nobody has: whether Martian gravity is sufficient to maintain a human skeleton, whether shielded habitats can hold radiation dose within acceptable limits, and whether children can develop normally at 0.38 g. If the answers are favourable, pantropy remains fiction. If they are not, the first generation born off Earth would face the choice in a form that no [[future-of-humanity|long-horizon]] argument made on Earth can settle for them, and the people making it would not be the people affected.

## See also

- [[radiation-hardening-humans]]
- [[germline-editing]]
- [[space-medicine]]
- [[generation-ship-biology]]
- [[human-enhancement]]
- [[posthuman]]
- [[tardigrade-genes]]
- [[bioethics-of-enhancement]]

## References

[^clynes1960]: `paper` Clynes, M.E. and Kline, N.S. "Cyborgs and Space." *Astronautics*, 1960. {A short proposal for drug-infusing implants in astronauts rather than a research programme; the word cyborg entered use through it.}
[^jakosky2018]: `paper` Jakosky, B.M. and Edwards, C.S. "Inventory of CO2 available for terraforming Mars." *Nature Astronomy*, 2018. {The inventory covers carbon dioxide reachable with present-day technology; it does not address volatiles imported from elsewhere.}
[^huerta2014]: `paper` Huerta-Sánchez, E. et al. "Altitude adaptation in Tibetans caused by introgression of Denisovan-like DNA." *Nature*, 2014.
[^ilardo2018]: `paper` Ilardo, M.A. et al. "Physiological and Genetic Adaptations to Diving in Sea Nomads." *Cell*, 2018.

==============================================================================
ARTICLE: parabiosis-and-young-blood
TITLE:   Parabiosis and young blood
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/parabiosis-and-young-blood
SOURCE:  https://futurehumanwiki.com/raw/parabiosis-and-young-blood
==============================================================================

---
title: "Parabiosis and young blood"
slug: "parabiosis-and-young-blood"
type: "intervention"
status: "experimental"
horizon: "2030s"
trl: 3
categories: ["longevity"]
tags: ["parabiosis", "plasma", "aging", "gdf11", "rejuvenation", "clinical trials"]
summary: "Experiments joining the circulation of young and old animals, and the attempts to translate their rejuvenating effects into plasma-based treatments for humans."
updated: "2026-07-27"
humanEvidence: "Human work is limited to small plasma-infusion and plasma-exchange studies reporting feasibility, biomarker shifts, and one debated subgroup result; every rejuvenation and lifespan finding is from mice."
access: "Not approved for aging anywhere; plasma products are licensed only for medical indications, and clinics in loosely regulated jurisdictions sell young-donor infusions of the kind the FDA warned against in 2019."
reversibility: "reversible"
issues: ["The 2019 FDA warning on young-donor plasma is described without a citation."]
---

```infobox
{
  "caption": "Experimental approach in biogerontology",
  "rows": [
    { "label": "Technique", "value": "Surgical parabiosis, plasma transfer" },
    { "label": "First performed", "value": "1864 (Paul Bert, rats)" },
    { "label": "Modern revival", "value": "2005" },
    { "label": "Key disputed factor", "value": "GDF11" },
    { "label": "Human evidence", "value": "Small, inconclusive trials" },
    { "label": "Commercial clinics", "value": "Subject of FDA warning, 2019" },
    { "label": "Readiness", "value": "TRL 3" }
  ]
}
```

**Parabiosis and young blood** refers to a line of experiments in which the circulatory systems of a
young and an old animal are joined, and to the effort to extract from those experiments a treatment
that can be given to people. Heterochronic parabiosis in mice improves tissue repair, neurogenesis,
and several molecular measures of age in the older partner, which is among the more striking results
in biogerontology. The translation has been troubled: the best-known candidate factor is disputed, an
alternative interpretation attributes the effect to dilution rather than to anything youthful, and the
human market ran far ahead of the science.

## The experiments

Surgical parabiosis, in which two animals are joined along a flank incision so that a shared
vasculature develops, dates to Paul Bert's rat experiments in 1864 and was used through the
twentieth century to study endocrine and metabolic signals. Mid-century work reported that old rats
joined to young partners survived longer than unjoined controls, in studies with small numbers and
considerable surgical mortality.

The modern field begins in 2005, when Conboy and colleagues showed that exposing an old mouse to a
young circulation restored the activation of muscle satellite cells and liver progenitors, largely by
recovering Notch signalling.[^conboy2005] The aged cells themselves were not irreversibly damaged;
they were responding to their environment. That reframing connected the result directly to
[[stem-cell-exhaustion]], and suggested that a share of what looks like intrinsic aging is imposed by
circulating signals — a systemic account of the [[hallmarks-of-aging]] rather than a cell-autonomous
one. The candidate signals include the secreted products of [[cellular-senescence|senescent cells]]
and the chronic immune activation of [[inflammaging]], which is why senescent-cell burden and plasma
composition are often studied together.

Subsequent work extended the finding to the brain. Old blood was shown to impair neurogenesis in
young mice, implicating specific circulating chemokines, and young plasma alone was reported to
improve synaptic plasticity and performance on memory tasks in old mice.[^villeda2011] Umbilical cord
plasma produced similar effects, with the metalloproteinase inhibitor TIMP2 identified as one
mediator.[^castellano2017] A 2023 study reported that a period of heterochronic parabiosis followed
by separation left old mice with reduced epigenetic-age estimates and a modest increase in remaining
lifespan.[^zhang2023]

## The GDF11 dispute

The clearest example of the field's difficulties is GDF11, a circulating TGF-β family protein.
Beginning in 2013, one group reported that GDF11 declines with age and that restoring it reverses
cardiac hypertrophy, improves skeletal muscle regeneration, and enhances brain
vasculature.[^loffredo2013] The papers appeared in leading journals and drove substantial commercial
interest.

A rebuttal from an industrial group in 2015 reported the opposite: that the antibodies and assays
used could not distinguish GDF11 from the closely related protein myostatin, that GDF11 does not
decline with age in the way claimed, and that administering it impairs rather than improves muscle
regeneration.[^egerman2015] The exchange has never been fully resolved. It bears directly on
[[myostatin-inhibition]], since the two proteins share receptor pathways and much of the reagent
problem.

> [!debate] Young factors or old factors?
> The intuitive reading of parabiosis is that young blood supplies something restorative. The
> alternative is that old blood contains inhibitory factors and the young partner simply dilutes
> them. Replacing about half the plasma of old mice with saline and albumin (no young blood at all)
> produced improvements comparable to parabiosis in several tissues.[^mehdipour2020] If dilution is
> the mechanism, the therapeutic target is removal, not transfusion.

## Confounds

Parabiosis is a crude instrument. Joined animals share far more than plasma: the young partner's
kidneys, liver, lungs, spleen, and bone marrow filter and service both bodies. An old animal in
parabiosis effectively acquires a second set of young organs, and improvements in its tissues may
reflect better clearance of metabolic waste rather than the delivery of any signalling molecule.
Immune cells cross between partners. Feeding, activity, and thermoregulation change. Results from
plasma transfer alone, which removes most of these confounds, are consistently weaker than results
from surgical joining — a pattern that argues the organ-sharing component is substantial.

The relevant circulating factors are also not necessarily "youth" factors in any general sense.
Plasma from exercised mice transfers benefits to sedentary ones, and exercise changes the same broad
classes of protein. The distinction between a rejuvenation signal and an ordinary
[[exercise-and-aging|exercise response]] has not been cleanly drawn.

## Human attempts

Human work has taken three forms, none of which has produced convincing evidence of benefit.

*Plasma infusion trials.* A small study infusing plasma from young donors into patients with
Alzheimer's disease established feasibility and tolerability without addressing efficacy, and the
company behind it moved toward defined plasma fractions rather than whole plasma before being
acquired. A separate line of work using therapeutic plasma exchange with albumin replacement in
Alzheimer's reported slowed decline in a subgroup, a result that remains debated.

*Commercial young-plasma clinics.* A US startup sold transfusions of young donor plasma to paying
customers for general anti-aging purposes. In 2019 the FDA issued a public statement warning that
infusions of plasma from young donors for aging, memory, or other conditions have no proven benefit
and carry the known risks of transfusion, including allergic reaction, circulatory overload, and
infection. The clinic model has recurred in less regulated jurisdictions since.

*Plasma exchange in healthy adults.* Small studies of therapeutic plasma exchange in older
volunteers have reported changes in inflammatory and proteomic panels and in
[[epigenetic-clock|clock-based]] age estimates. These are surrogate measures, and a shift in a
[[biological-age|biological-age estimate]] is not evidence of functional improvement. The most
publicized instance, a technology entrepreneur exchanging plasma with his teenage son as one arm of
a heavily measured [[quantified-self]] regimen, was discontinued after the participants reported no
discernible benefit, which is a fair summary of the public human record.

## Risks

Plasma is a biological product with an established risk profile: transfusion reactions, transfusion-
associated circulatory overload, transfusion-related acute lung injury, and residual infectious risk.
These risks are accepted in the treatment of bleeding or immune disease and are not obviously
acceptable in a healthy person pursuing an unproven benefit. Repeated large-volume exchange also
removes clotting factors, antibodies, and drugs along with whatever the target is, and requires
replacement fluids that carry their own costs. The donor-supply implications of a large market for
young plasma, and its interaction with existing blood-donation systems, are a distributional problem
in their own right, related to those discussed under [[access-and-inequality]] and to the wider
search for [[artificial-blood|manufactured blood products]].

## Outlook

The productive residue of this line of work is not transfusion. It is the identification of specific
circulating proteins that change with age and have measurable effects when manipulated — a target
list that feeds into [[gene-therapy-for-aging|systemic gene therapy]], recombinant protein programmes,
and the broader project of [[aging-biomarkers|measuring systemic aging]]. Klotho, TIMP2, and
exercise-induced factors are being pursued individually, where dose and mechanism can be controlled
in a way that whole plasma never allows.

Removal is the mirror strategy. If old plasma carries inhibitory factors, the therapeutic act is
subtraction — targeted apheresis for specific proteins, or clearance of the cells producing them,
which is the logic of [[senolytics]]. Either version would have to be tested against
[[healthspan|functional endpoints]] rather than the biomarker panels that dominate current reports.

The open question is whether the parabiosis effect will decompose into a small number of tractable
molecules or whether it depends on the wholesale exchange of a circulatory environment. If the latter,
there may be no drug on the other side of these experiments at all — only the demonstration that aged
tissue retains more capacity than it displays, which is the same premise driving
[[epigenetic-reprogramming]].

## See also

- [[stem-cell-exhaustion]]
- [[epigenetic-reprogramming]]
- [[myostatin-inhibition]]
- [[senolytics]]
- [[aging-biomarkers]]
- [[gene-therapy-for-aging]]
- [[hallmarks-of-aging]]
- [[inflammaging]]

## References

[^conboy2005]: `paper` Conboy, I.M. et al. "Rejuvenation of aged progenitor cells by exposure to a young systemic environment." *Nature*, 2005. {Surgically joined mice share organs as well as plasma, so the result does not isolate any circulating factor.}
[^villeda2011]: `paper` Villeda, S.A. et al. "The ageing systemic milieu negatively regulates neurogenesis and cognitive function." *Nature*, 2011.
[^castellano2017]: `paper` Castellano, J.M. et al. "Human umbilical cord plasma proteins revitalize hippocampal function in aged mice." *Nature*, 2017.
[^zhang2023]: `paper` Zhang, B. et al. "Multi-omic rejuvenation and life span extension on exposure to youthful circulation." *Nature Aging*, 2023. {Mice, and the exposure is surgical joining followed by separation, which has no human equivalent; the reported lifespan gain was modest.}
[^loffredo2013]: `paper` Loffredo, F.S. et al. "Growth differentiation factor 11 is a circulating factor that reverses age-related cardiac hypertrophy." *Cell*, 2013. {A later report from another group could not distinguish GDF11 from myostatin with comparable assays and found the opposite trend with age.}
[^egerman2015]: `paper` Egerman, M.A. et al. "GDF11 increases with age and inhibits skeletal muscle regeneration." *Cell Metabolism*, 2015.
[^mehdipour2020]: `paper` Mehdipour, M. et al. "Rejuvenation of three germ layers tissues by exchanging old blood plasma with saline-albumin." *Aging*, 2020.

==============================================================================
ARTICLE: partial-reprogramming
TITLE:   Partial reprogramming
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/partial-reprogramming
SOURCE:  https://futurehumanwiki.com/raw/partial-reprogramming
==============================================================================

---
title: "Partial reprogramming"
slug: "partial-reprogramming"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 3
categories: ["longevity", "genetics"]
tags: ["reprogramming", "rejuvenation", "epigenetics", "aging", "gene therapy"]
summary: "Transient expression of pluripotency factors that resets age-associated epigenetic marks in a cell while stopping short of erasing its differentiated identity."
updated: "2026-07-28"
humanEvidence: "No human trial has been published as of 2026; the human results are in cultured cells, and every in vivo result is in mice, most of them transgenic lines carrying an inducible factor cassette."
access: "Not available to patients: no approved product and no published human trial as of 2026, though companies have stated an intention to begin one in an eye indication."
reversibility: "context"
issues: ["The stated company intention to run a first-in-human eye trial is unsourced."]
---

```infobox
{
  "caption": "Experimental rejuvenation technique",
  "rows": [
    { "label": "Also called", "value": "Transient reprogramming" },
    { "label": "Factors used", "value": "OSK, sometimes OSKM", "link": "/wiki/yamanaka-factors" },
    { "label": "First in vivo result", "value": "2016, progeroid mice" },
    { "label": "Typical readout", "value": "Methylation age, transcriptome" },
    { "label": "Human data", "value": "None published as of 2026" },
    { "label": "Readiness", "value": "TRL 3" }
  ]
}
```

**Partial reprogramming** is the deliberate interruption of the reprogramming process partway through, so that a cell sheds age-associated epigenetic marks without becoming a stem cell. It exists because two things happen at different speeds when pluripotency factors are switched on: the epigenome starts changing almost immediately, while the cell's committed identity survives for considerably longer. Everything about the technique is an attempt to exploit that gap.

```keyfacts
[
  { "value": "2 on, 5 off", "label": "Days of induction per week in the original protocol", "note": "Ocampo and colleagues, 2016" },
  { "value": "OSK", "label": "Factor set now standard", "note": "c-Myc dropped to reduce tumour risk" },
  { "value": "None", "label": "Published human efficacy trials", "note": "as of mid-2026" }
]
```

## The reprogramming trajectory

When the [[yamanaka-factors]] are expressed in a fibroblast, the cell begins a journey that ends, in a small minority of cases, at pluripotency. The journey has recognisable stages. An early phase involves loss of the somatic transcriptional programme and, in fibroblasts, a mesenchymal-to-epithelial transition. A maturation phase follows in which some cells activate the endogenous pluripotency network. Only then does the cell become an [[induced-pluripotent-stem-cells|induced pluripotent stem cell]] and stop depending on the exogenous factors.

Methylation age, as read by an [[epigenetic-clock]], falls steadily from the beginning of this process rather than dropping at the end. Nadia Olova and colleagues mapped the two curves and found that epigenetic age declines well before somatic identity is lost, which established that a usable window exists.[^olova2019] Diljeet Gill and colleagues in Wolf Reik's laboratory pushed human fibroblasts into the maturation phase and then withdrew the factors, obtaining cells that were substantially younger by transcriptomic and methylation measures while remaining fibroblasts, with improved collagen production and migration.[^gill2022]

## Dosing regimens

Three schedules are in use, and the choice among them is the main design decision.

**Cyclic induction** switches the factors on for short intervals separated by longer rest periods. The original in vivo protocol used two days on and five days off, indefinitely. This is the most-tested approach and the one with the most safety data.

**Single-pulse transient induction** applies the factors once, for days rather than hours, then removes them permanently. The maturation-phase work in human cells uses this schedule. It reaches deeper into the trajectory and relies on the cell re-establishing its identity afterwards.

**Continuous low-dose expression** keeps the factors on at a level too low to drive full reprogramming. It has been sustained in mice for months without gross toxicity but concentrates the risk of slow drift toward dedifferentiation.

> [!note] What "partial" refers to
> The term describes how far along the trajectory the cell is taken, not what fraction of cells are treated. In practice both matter, and most in vivo experiments cannot distinguish a modest change in every cell from a large change in a few.

## In vivo results

A decade of mouse work has produced a rough dose-response map rather than one headline finding. Read as a set, the experiments vary four things: the factor combination, the schedule, the total exposure, and the tissue. What each result contributes is a coordinate on that map.

The upper bound came first. Cyclic OSKM induction in mice carrying a [[progeria|progeria-causing]] lamin A mutation extended median lifespan, while the same construct left on continuously killed the animals.[^ocampo2016] Everything since has been an attempt to work inside the gap those two arms defined. Wild-type animals on the cyclic schedule regenerated better after injury without a reported lifespan benefit, so the progeroid number cannot be carried across to normal aging.

The lower bound is less well defined. Continuous low-level induction sustained over months in middle-aged and old mice shifted age-associated transcriptional and methylation signatures in several tissues without evident tumour formation, which suggests the tolerated exposure depends on intensity and duration together rather than on either alone.[^browder2022] A single transient cycle, at the opposite extreme of the schedule space, produced smaller multi-omic changes in naturally aged tissue.[^chondronasiou2022] No lifespan result in normal, genetically unmodified mice has been replicated by an independent laboratory, so the progeroid finding remains the only well-established survival benefit.

Only one tissue has yielded a functional rather than molecular endpoint. AAV-delivered OSK restored visual function after optic nerve injury, in a glaucoma model and in aged animals, and the effect disappeared when the TET demethylases were knocked down.[^lu2020] That dependence is the closest the field has come to showing that the benefit runs through demethylation rather than through some general consequence of perturbing the cell. The broader case for reprogramming as a rejuvenation platform, and the money behind it, are set out under [[epigenetic-reprogramming]].

## What is and is not rejuvenated

Partial reprogramming reliably changes methylation profiles, transcriptional signatures, and markers of [[cellular-senescence]]. Reports also describe improved [[autophagy|autophagic]] flux, restored [[proteostasis]] markers, and better function in aged [[stem-cell-exhaustion|stem cell compartments]].

It does not repair somatic mutations, which persist through any number of epigenetic resets. It does not remove extracellular matrix crosslinks or amyloid deposits. It does not eliminate senescent cells, which is why some groups combine it with [[senolytics]] rather than treating it as a substitute. Of the [[hallmarks-of-aging]], it addresses one directly and several indirectly, and leaves others untouched.

An unresolved mechanistic question is whether the benefit comes from epigenetic reset at all. Reprogramming stresses cells, kills some of them, and stimulates proliferation in others. A tissue that looks younger afterwards may have been partly repopulated rather than rejuvenated, and few experiments have been designed to distinguish the two.

## Safety

The open safety question is not whether the factors are dangerous, which is settled, but whether a dose exists that is large enough to reset a cell and small enough for the tissue to tolerate. Sustained OSKM expression throughout a mouse produces pluripotent cells in multiple organs and lethal teratomas, which fixes the ceiling.[^abad2013] Nothing establishes that the ceiling and the effective dose are separated by a comfortable margin in every tissue, and the margin appears to differ substantially between skin and eye on one side and intestine, pancreas and liver on the other. That asymmetry is the main argument for treating a confined compartment first.

Two failure modes sit inside the tolerated window and are largely invisible to the assays usually run. The first is functional loss without visible dedifferentiation: a hepatocyte can score younger on a clock while metabolising drugs worse, and few studies pair a molecular readout with a tissue-function assay in the same animal. The second is clonal rather than tissue-level. One cell pushed too far can seed a tumour long after the treatment ends, an event that no group-average measurement taken weeks later would detect. Dropping c-Myc lowers this risk without removing it, since Klf4 is also a proto-oncogene and dedifferentiation is itself a step toward malignancy.

## Delivery and control

Every result above depends on a control system with no human equivalent. A doxycycline-inducible transgene sits in the germline of the mouse, is present in every cell at a known copy number, and stops when the drug is taken out of the drinking water. None of those properties survives translation: a patient has no transgene, cannot be transduced uniformly, and has no way to have the construct removed if it misbehaves.

The available options therefore trade control against reach. [[aav-vectors]] carrying an inducible cassette persist for the life of the transduced cell, so the off-switch must remain reliable for decades in tissue nobody can sample. mRNA in [[lipid-nanoparticles]] is self-limiting by construction, which solves the off-switch and leaves redosing and tropism beyond the liver unsolved. Confining treatment to a compartment such as the eye sidesteps both and forfeits any systemic effect. These are the constraints on [[somatic-gene-therapy]] generally, sharpened by a payload whose failure mode is a tumour.

A separate line of work removes the payload rather than controlling it, screening cell-type-specific transcription factors for youthful expression within a fixed identity so that pluripotency is never approached. The cost is that the screen has to be repeated for every cell type. [[newlimit]] has built its programme on that trade; [[altos-labs]] works on both routes, while [[retro-biosciences]] has concentrated on engineering the pluripotency factors themselves rather than replacing them.

## Outlook

The evidence base is animal, largely mouse, and concentrated in a small number of transgenic lines. A first-in-human trial in an eye indication would supply the safety information the field most lacks, and companies have stated intentions to run one. Until then, the claim that partial reprogramming rejuvenates tissue rests on measures whose relationship to [[biological-age]] is itself unsettled, and any human timeline offered for systemic rejuvenation should be read as advocacy rather than forecast.

## See also

- [[epigenetic-reprogramming]]
- [[yamanaka-factors]]
- [[epigenetic-clock]]
- [[gene-therapy-for-aging]]
- [[induced-pluripotent-stem-cells]]
- [[aging-biomarkers]]
- [[senolytics]]

## References

[^olova2019]: `paper` Olova, N., Simpson, D.J., Marioni, R.E., Chandra, T. "Partial reprogramming induces a steady decline in epigenetic age before loss of somatic identity." *Aging Cell*, 2019.
[^gill2022]: `paper` Gill, D. et al. "Multi-omic rejuvenation of human cells by maturation phase transient reprogramming." *eLife*, 2022. {Human cells in culture rather than human subjects; the endpoints are transcriptomic, methylation-based, and fibroblast function.}
[^ocampo2016]: `paper` Ocampo, A. et al. "In Vivo Amelioration of Age-Associated Hallmarks by Partial Reprogramming." *Cell*, 2016. {The lifespan result is in mice carrying a progeria mutation; wild-type mice on the same schedule showed no reported lifespan benefit.}
[^browder2022]: `paper` Browder, K.C. et al. "In vivo partial reprogramming alters age-associated molecular changes during physiological aging in mice." *Nature Aging*, 2022.
[^chondronasiou2022]: `paper` Chondronasiou, D. et al. "Multi-omic rejuvenation of naturally aged tissues by a single cycle of transient reprogramming." *Aging Cell*, 2022.
[^lu2020]: `paper` Lu, Y. et al. "Reprogramming to recover youthful epigenetic information and restore vision." *Nature*, 2020.
[^abad2013]: `paper` Abad, M. et al. "Reprogramming in vivo produces teratomas and iPS cells with totipotency features." *Nature*, 2013. {Sustained whole-body factor expression in mice; it fixes the ceiling on exposure and says nothing about where the tolerated dose sits.}

==============================================================================
ARTICLE: personal-identity-and-continuity
TITLE:   Personal identity and continuity
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/personal-identity-and-continuity
SOURCE:  https://futurehumanwiki.com/raw/personal-identity-and-continuity
==============================================================================

---
title: "Personal identity and continuity"
slug: "personal-identity-and-continuity"
type: "concept"
status: "contested"
horizon: "indefinite"
categories: ["minds", "society"]
tags: ["identity", "philosophy of mind", "parfit", "survival", "consciousness", "bioethics"]
summary: "The question of what makes a person at one time the same person as at another, and what — if anything — must be preserved for that person to survive."
updated: "2026-07-28"
---

```infobox
{
  "caption": "Problem in philosophy and bioethics",
  "rows": [
    { "label": "Classical statement", "value": "Locke, 1694" },
    { "label": "Main criteria", "value": "Psychological, biological, narrative" },
    { "label": "Pivotal modern work", "value": "Parfit, 1984" },
    { "label": "Hard case", "value": "Fission" },
    { "label": "Practical bearing", "value": "Advance directives, uploading, cryonics" },
    { "label": "Status", "value": "No settled consensus" }
  ]
}
```

**Personal identity and continuity** is the problem of what makes a person existing at one time the
same person as one existing at another, and what has to be preserved for that person to survive a
change. It is an old question in philosophy that acquires practical force whenever a technology
proposes to interrupt, copy, or reconstruct a mind — which is why it recurs across
[[mind-uploading]], [[cryonics]], gradual neural replacement, and advance directives in dementia.
Nothing in neuroscience answers it, because it is not a question about what happens but about how to
describe what happens.

## The classical positions

John Locke separated the person from both the soul and the animal. A person, he wrote, is a thinking
intelligent being that can consider itself as itself across time, and personal identity extends only
as far as consciousness — effectively, memory — extends backwards.[^locke1694] The proposal explains
why a person is responsible for past acts they remember and severs identity from bodily persistence.

Two objections followed quickly. Joseph Butler charged the account with circularity: to remember
doing something is already to remember *oneself* doing it, so memory presupposes identity rather
than constituting it. Thomas Reid's brave officer case sharpened it: an old general remembers being
a young officer, the young officer remembered being a flogged schoolboy, but the general remembers
nothing of the schoolboy. Memory is not transitive; identity is. The standard repair replaces direct
memory with overlapping chains of psychological connection — *psychological continuity* — which
restores transitivity at the cost of the intuitive simplicity Locke had.

## Parfit's reductionism

Derek Parfit's *Reasons and Persons* reframed the field.[^parfit1984] Parfit argues that persons are
not separately existing entities over and above brains, bodies, and interrelated mental events;
personal identity consists in facts describable without presupposing it. From this he draws a
conclusion most readers find harder than the metaphysics: identity is not what matters in survival.
What matters is Relation R — psychological connectedness and continuity — and R can hold to a degree,
can hold with more than one future person, and can hold without identity.

The practical upshot is that some questions have no answer because they are questions about how to
describe a case, not about a hidden fact. Parfit's *combined spectrum* imagines a series in which one
person's brain and body are progressively replaced by another's; somewhere along it, asking whether
the resulting person is the original is, on his view, an empty question.

> [!debate] Why this is the crux for uploading
> If Parfit is right, an upload that inherits full psychological continuity carries everything worth
> caring about, and its being a copy is not a further loss. If he is wrong and identity requires
> something more — a continuously running physical process, say — then destructive scanning is death
> with a convincing survivor. The engineering cannot decide between these. See
> [[teleportation-problem]].

## Fission and the branching problem

The case that breaks most theories is division. Suppose each hemisphere of a brain is transplanted
into a different body, and both resulting people have the original's memories and character. They
cannot both be the original, since they are two and identity is one-to-one. They cannot be neither,
since each has as good a claim as an ordinary survivor. Choosing one arbitrarily is unmotivated.

The available responses divide neatly. Parfit accepts that identity fails while what matters is preserved
twice over. David Lewis, taking persons to be four-dimensional worms of person-stages, holds that
two people were present all along, sharing stages before the split.[^lewis1976] Robert Nozick's
closest-continuer theory makes identity depend on whether a candidate is the best continuer, with
the awkward result that whether a person survives can depend on what happens to someone
else.[^nozick1981] Every option gives up something ordinarily assumed about persons.

Fission is not purely hypothetical. Commissurotomy patients, whose corpus callosum has been severed
to control epilepsy, show hemispheric dissociations under controlled testing that Thomas Nagel took
as evidence that the unity of consciousness may not be all-or-nothing.[^nagel1971] The clinical
picture has since become more complicated, and researchers dispute how divided such patients really
are.

## Rivals: animalism and narrative

Animalism, defended by Eric Olson, holds that each of us is a human animal, and that a person
persists exactly as long as the organism does.[^olson1997] Its main argument is the
too-many-thinkers problem: if a person is distinct from the animal sitting in a chair, then two
thinkers occupy the chair. Animalism denies that psychological continuity is even necessary for
survival — a patient in a persistent vegetative state is still the same animal — and denies that any
upload could be a person's continuation, since the animal remains behind or dies.

Narrative accounts, associated with Marya Schechtman, treat identity as constituted by an ongoing
self-told story that organizes experience into a life.[^schechtman1996] These accounts are better at
capturing what people mean when they say someone is "not the same person" after a trauma, and worse
at delivering the sharp yes-or-no answers that law and metaphysics ask for.

## What people actually believe

Experimental work suggests folk judgements track something narrower than psychological continuity in
general. Nina Strohminger and Shaun Nichols found that participants judging cases of neural change
treated moral traits — honesty, cruelty, compassion — as the most identity-defining part of a mind,
more so than memories or personality broadly.[^strohminger2014] Related studies find that intuitions
about survival shift with how a case is framed, and that people give inconsistent answers to
structurally identical scenarios.

This is relevant to technologies that alter mental function without ending life. Reports of
personality change after [[deep-brain-stimulation]] were widely cited as an identity problem,
though several reviews have since argued that the empirical basis is thinner than the philosophical
literature assumed and that much of it rests on small numbers of case reports. Similar questions
attach to a [[memory-prosthesis]], to pharmacological [[moral-enhancement]], and more broadly to the
debates surveyed in [[bioethics-of-enhancement]].

## Practical stakes

**Advance directives and dementia.** Ronald Dworkin argued that a person's earlier, competent wishes
should govern their later demented self, on the grounds that a life has critical interests that
outlive the capacity to express them.[^dworkin1993] Rebecca Dresser replied that the later person's
present experiential interests are real and may conflict, and that precedent autonomy can amount to
one person disposing of another. The disagreement is exactly the psychological-versus-biological
criterion applied at a bedside.

**Preservation and revival.** Whether [[brain-preservation]] preserves a person depends on which
criterion holds and on what is physically retained; the two questions are usually run together.
Under animalism, revival of a preserved brain does not obviously return the same animal. Under
psychological criteria, it depends entirely on whether the information supporting continuity
survived, which is an empirical matter addressed by [[connectomics]] and not yet settled.

**Very long lives.** A person who lives for centuries under something like
[[longevity-escape-velocity]] may retain no direct psychological connection to their earliest self.
On Parfit's account the strength of Relation R fades with distance, which weakens ordinary reasons
for prudential concern about one's own remote future and complicates arguments about consent to
continued existence taken up in [[right-to-die-and-right-to-live]].

**Emulation and copying.** If [[substrate-independence]] holds and [[whole-brain-emulation]] works,
copies become cheap, and every institution built around one-person-one-body — property, criminal
responsibility, voting, consent — encounters cases it was not designed for. The weak-form products marketed as
[[digital-immortality]] raise a smaller version of the same problem now, since a convincing
imitation of a dead person has no continuity claim at all but is treated by users as though it did.

## Open problems

The deepest difficulty is that the competing theories agree on all the physical facts. No
measurement distinguishes a world in which Relation R is what matters from one in which strict
identity is, and no result from [[neural-correlates-of-consciousness]] research bears on it. That
leaves the question to be settled, if at all, by which set of intuitions people are willing to give
up — and by what legal systems decide when a case arrives that cannot be postponed.

## See also

- [[teleportation-problem]]
- [[mind-uploading]]
- [[substrate-independence]]
- [[brain-preservation]]
- [[cryonics]]
- [[digital-immortality]]
- [[whole-brain-emulation]]
- [[machine-consciousness]]
- [[head-transplant]]

## References

[^locke1694]: `book` Locke, J. *An Essay Concerning Human Understanding*, 2nd edition, Book II, Chapter XXVII, "Of Identity and Diversity," 1694.

[^parfit1984]: `book` Parfit, D. *Reasons and Persons*. Oxford University Press, 1984.

[^lewis1976]: `book` Lewis, D. "Survival and Identity." In A. Rorty (ed.), *The Identities of Persons*, University of California Press, 1976.

[^nozick1981]: `book` Nozick, R. *Philosophical Explanations*. Harvard University Press, 1981.

[^nagel1971]: `paper` Nagel, T. "Brain Bisection and the Unity of Consciousness." *Synthese*, 1971.

[^olson1997]: `book` Olson, E. T. *The Human Animal: Personal Identity Without Psychology*. Oxford University Press, 1997.

[^schechtman1996]: `book` Schechtman, M. *The Constitution of Selves*. Cornell University Press, 1996.

[^strohminger2014]: `paper` Strohminger, N. and Nichols, S. "The essential moral self." *Cognition*, 2014. {Vignette experiments with lay participants: evidence about what people judge identity to consist in, not about what it does consist in.}

[^dworkin1993]: `book` Dworkin, R. *Life's Dominion: An Argument About Abortion, Euthanasia, and Individual Freedom*. Knopf, 1993.

==============================================================================
ARTICLE: polygenic-embryo-screening
TITLE:   Polygenic embryo screening
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/polygenic-embryo-screening
SOURCE:  https://futurehumanwiki.com/raw/polygenic-embryo-screening
==============================================================================

---
title: "Polygenic embryo screening"
slug: "polygenic-embryo-screening"
type: "technology"
status: "contested"
horizon: "present"
trl: 7
categories: ["reproduction", "genetics", "society"]
tags: ["polygenic scores", "ivf", "embryo selection", "genomics", "bioethics", "reproduction"]
summary: "The ranking of IVF embryos by polygenic risk scores computed from their inferred genotypes, sold commercially since 2019 and rejected as unvalidated by professional genetics bodies."
updated: "2026-07-27"
humanEvidence: "Children have been born after polygenic ranking of embryos, but no study has followed selected children to test the predicted gain; the scores are validated in adult cohorts, mostly of European ancestry."
access: "Sold in the United States as a laboratory-developed test, an add-on to a self-funded IVF cycle at a small number of clinics; the United Kingdom does not permit it and several European states restrict it."
reversibility: "irreversible"
issues: ["The 2024-2025 trait-screening timeline entry cites no source."]
---

```infobox
{
  "caption": "Reproductive genetic technology",
  "rows": [
    { "label": "Also known as", "value": "PGT-P" },
    { "label": "Basis", "value": "Polygenic scores" },
    { "label": "First offered", "value": "2019" },
    { "label": "Requires", "value": "IVF and embryo biopsy", "link": "/wiki/embryo-selection" },
    { "label": "Typical predicted gain", "value": "Small fraction of a standard deviation" },
    { "label": "UK status", "value": "Not permitted" },
    { "label": "Professional position", "value": "Not recommended" }
  ]
}
```

**Polygenic embryo screening**, formally preimplantation genetic testing for polygenic conditions (PGT-P), ranks the embryos produced in an [[embryo-selection|IVF cycle]] by polygenic scores calculated from hundreds of thousands of common genetic variants. Clinics and direct-to-consumer companies use it to report an embryo's relative predicted risk for conditions such as type 2 diabetes, coronary artery disease and schizophrenia, and in some cases for non-disease traits including height and [[genetic-enhancement-of-intelligence|cognitive test performance]]. The laboratory procedure is a modest extension of routine embryo genotyping. The dispute is about what the resulting rankings are worth.

```keyfacts
[
  { "value": "≈2.5 cm", "label": "Predicted height gain", "note": "selecting the top-scoring of ten embryos, Karavani et al., 2019" },
  { "value": "≈2.5 pts", "label": "Predicted IQ-score gain", "note": "same selection scenario; within-family, not population, prediction" },
  { "value": "2019", "label": "First commercial offering", "note": "Genomic Prediction's embryo health score" }
]
```

## How it works

PGT-P attaches to an existing clinical workflow rather than replacing it. Five or six days after fertilisation, an embryologist biopsies five to ten cells from the trophectoderm, the outer layer of the blastocyst that will become placenta rather than fetus. The DNA in that sample is measured in picograms, so it is first amplified across the whole genome, then read on a SNP array or by low-pass sequencing.

The resulting genotype is sparse and noisy. To make a polygenic score usable, the laboratory imputes the missing variants: it genotypes both parents, reconstructs which parental haplotypes each embryo inherited, and fills in the rest by reference to a haplotype panel. An embryo's score is therefore not measured but reconstructed, and imputation error propagates into the ranking. Companies then sum the embryo's variant dosages weighted by effect sizes taken from a published genome-wide association study, and report each embryo's position relative to its siblings or to a population distribution.

Some providers have moved to sequencing the amplified biopsy at higher depth in order to call rare variants alongside common ones, which brings PGT-P closer to a whole-genome report on a sample of a few cells. That does not change the underlying statistical problem, which lies in the effect-size weights rather than the genotyping.

> [!note] Terminology
> PGT-A tests for aneuploidy, PGT-M for a known single-gene disorder, and PGT-SR for structural rearrangements. All three ask a categorical question about one locus or chromosome. PGT-P asks a probabilistic question about the whole genome, which is why it sits in a different evidentiary category.

## Development history

```timeline
[
  { "year": "2007–2018", "title": "GWAS reach useful scale", "text": "Consortium studies of hundreds of thousands of participants produce polygenic scores with non-trivial predictive power for height, education and several common diseases." },
  { "year": "2017–2019", "title": "First commercial service", "text": "Genomic Prediction, co-founded by physicist Stephen Hsu, begins offering an embryo health score to IVF clinics in the United States." },
  { "year": "2019", "title": "Quantitative critique", "text": "Karavani and colleagues model the expected gain from selecting the top-scoring embryo and conclude it is small for both height and cognitive scores." },
  { "year": "2020–2021", "title": "First reported birth and formal objections", "text": "A child selected using a polygenic score is publicly reported, and a group of behaviour geneticists and statisticians set out the problems with the practice in the New England Journal of Medicine." },
  { "year": "2022", "title": "European societies object", "text": "A joint statement from European genetics and reproductive medicine societies calls the practice unproven and unethical." },
  { "year": "2024–2025", "title": "Trait screening goes public", "text": "Journalistic investigations document companies offering embryo ranking on cognitive traits, and at least one firm markets such ranking openly, drawing broad criticism from geneticists." }
]
```

## What the gain actually is

The size of any benefit is set by three quantities: how well a polygenic score predicts the trait *between siblings*, how much genetic variation exists among the embryos of one couple, and how many embryos there are to choose from. All three are less favourable than intuition suggests.

Siblings differ only by the recombination and segregation of their parents' chromosomes, so the spread of true genetic values within a sibship is narrower than in the population at large. Selecting the highest-scoring of ten embryos captures only the upper tail of that narrow distribution, and only in proportion to the score's accuracy. Modelling this explicitly, Karavani and colleagues estimated an expected gain of roughly two and a half centimetres of height, or a comparable number of IQ points, when choosing the best of ten embryos.[^karavani2019] With a more realistic number of viable embryos the expected gain falls further, and the variance around it is wide enough that a substantial fraction of couples would see no gain at all.

For disease, the arithmetic differs because relative risk reduction can be large where absolute risk is small. Lencz and colleagues showed that the apparent utility depends heavily on the selection strategy: excluding embryos in the top few per cent of risk for a highly heritable disorder gives a different picture from ranking all embryos and taking the best.[^lencz2021] A large relative reduction in the risk of a condition affecting one person in a hundred still leaves the great majority of the couple's expected outcome untouched.

The comparison that matters clinically is with what selection already does well. Testing for a known single-gene disorder converts a one-in-four risk into something close to certainty of avoidance, because the underlying genetics is categorical. Polygenic testing offers a probabilistic nudge across many conditions at once, and the two are often presented to patients on the same report. Nothing in the laboratory output distinguishes the reliable number from the speculative one, which is the practical complaint clinical geneticists make most often.

> [!debate] Relative versus absolute
> A report telling a couple that one embryo carries, say, sixty per cent less risk of schizophrenia than another describes a shift in a probability that was already low, computed from a model that has never been validated against the outcome of a selected birth. Proponents argue that a real if small expected benefit is still a benefit. Critics reply that the reported percentages imply a precision the underlying scores do not have.

## Why prediction weakens where it matters

### Within-family versus population prediction

Polygenic scores are trained on unrelated individuals, where they absorb more than direct genetic effects. Population-scale associations also capture assortative mating, residual population structure, and indirect genetic effects from parents' genotypes acting through the environment. None of these differentiate one embryo from its sibling. Within-sibship analyses consistently find that predictive power drops when the comparison is made between siblings, and the drop is largest for socially patterned traits such as educational attainment.[^okbay2022] A score advertised with its population R² therefore overstates what it can do inside a family, which is the only setting in which embryo selection operates.

### Ancestry portability

Almost all large discovery cohorts are of European ancestry. Because linkage disequilibrium patterns and allele frequencies differ across populations, scores lose accuracy when applied to people of other ancestries, often by a factor of two or more.[^martin2019] Embryos of mixed ancestry are worse still, since no reference panel matches them. A technology whose accuracy tracks the ancestry composition of biobanks distributes its benefit unevenly by construction, a point that connects PGT-P to broader arguments about [[access-and-inequality]] and [[genetic-discrimination]].

### Pleiotropy and joint optimisation

Variants do not act on one trait. Genetic correlations between traits mean that pushing an embryo's score up for one outcome moves other scores, sometimes in unwanted directions. Companies address this by combining several scores into a single index, which requires the couple, or the company, to assign relative weights to incommensurable outcomes. That weighting is a value judgement dressed as a computation.[^turley2021]

### Embryo count

Selection intensity is the binding constraint. A typical stimulated cycle yields a modest number of blastocysts, and fewer still are chromosomally normal; for older patients, whose situation is discussed in [[reproductive-longevity]], the usable number is often one or two. In that case the ranking has nothing to rank. This is why [[in-vitro-gametogenesis]], which would in principle supply large numbers of embryos, is the development that would most change the calculus.

## The professional response

No major professional body endorses PGT-P for clinical use. A joint statement from European human genetics and reproductive medicine societies described the practice as unproven and unethical, citing the absence of clinical validity data, the misleading precision of the reports, and the burden placed on prospective parents asked to interpret them.[^forzano2022] The ethics committee of the American Society for Reproductive Medicine reached a similarly cautious conclusion, holding that the technology is not ready for routine offering and that clinics providing it must disclose its unvalidated status.[^asrm2024] American medical genetics bodies have issued comparable points-to-consider documents.

The objections are not only statistical. Reporting a score for a trait such as cognitive ability treats the trait as an optimisation target, which critics connect to the expressivist objection developed in [[disability-rights-and-enhancement]] and to the older arguments catalogued in [[bioconservatism]]. Supporters, several of them associated with [[julian-savulescu]]'s work on [[procreative-beneficence]], argue that a parent who would accept a small expected health benefit from prenatal vitamins has no principled reason to refuse a small expected benefit from selection.

There is also a counselling problem that neither side disputes. A PGT-P report presents multiple percentile figures across several conditions, derived from models most clinicians cannot interrogate, to patients already making a decision under time pressure and emotional strain. Genetic counsellors have argued that no realistic consultation can convey the difference between a validated Mendelian result and a polygenic percentile, and that the format of the report itself does persuasive work the evidence does not support. Because PGT-P is ordered as an add-on rather than as a distinct procedure, the discussion is often shorter than the one preceding [[germline-editing|any proposal to alter]] an embryo's genome, despite resting on weaker evidence.

## Regulation and access

In the United States, PGT-P reaches patients as a laboratory-developed test, a category historically subject to limited premarket review; no regulator has assessed the clinical validity of any embryo polygenic score. The United Kingdom takes the opposite approach: the Human Fertilisation and Embryology Authority licenses preimplantation testing only for specified serious conditions, and polygenic screening is not among them. That statutory list is the same mechanism by which Britain authorised [[mitochondrial-replacement-therapy]] under strict conditions, and it illustrates the difference between approving a technique and approving an indication. Several European jurisdictions restrict embryo testing more tightly still, and a few permit it only for conditions on a statutory list. The practical consequence is a market concentrated in a small number of American clinics serving an international clientele, at a price that sits on top of an IVF cycle most health systems do not fund. The resulting picture is the one anticipated in debates over the [[governance-of-genome-editing]]: a technology governed less by law than by where the clinic is.

## Outlook

Two developments would change the argument. The first is a within-family validated score, trained on sibling comparisons at sufficient scale that its coefficients reflect direct genetic effects rather than the social correlates of ancestry and family background. The second is a large increase in the number of embryos available per couple, which only laboratory-derived gametes plausibly deliver. Together they would move PGT-P from a service selling a fractional expected shift to one capable of a change large enough to argue about on its merits, which is roughly the situation that [[designer-babies|popular discussion]] already assumes exists and that the arguments in [[bioethics-of-enhancement]] were developed to address.

Neither development supplies the missing evidence. Establishing that a selected child actually enjoys the predicted advantage requires following selected and unselected children for decades, against a counterfactual sibling who was never born. No such study is being run, and it is not obvious how one could be. The field is therefore likely to keep arguing about a technology whose central claim is, by its own design, close to untestable.

## See also

- [[embryo-selection]]
- [[designer-babies]]
- [[genetic-enhancement-of-intelligence]]
- [[in-vitro-gametogenesis]]
- [[procreative-beneficence]]
- [[germline-editing]]
- [[genetic-discrimination]]
- [[disability-rights-and-enhancement]]

## References

[^karavani2019]: `paper` Karavani, E. et al. "Screening Human Embryos for Polygenic Traits Has Limited Utility." *Cell*, 2019. {The headline figures assume a choice among ten embryos, more than a typical cycle yields, and scores of present-day accuracy.}
[^lencz2021]: `paper` Lencz, T. et al. "Utility of polygenic embryo screening for disease depends on the selection strategy." *eLife*, 2021.
[^turley2021]: `paper` Turley, P. et al. "Problems with Using Polygenic Scores to Select Embryos." *New England Journal of Medicine*, 2021.
[^okbay2022]: `paper` Okbay, A. et al. "Polygenic prediction of educational attainment within and between families from genome-wide association analyses in 3 million individuals." *Nature Genetics*, 2022.
[^martin2019]: `paper` Martin, A. R. et al. "Clinical use of current polygenic risk scores may exacerbate health disparities." *Nature Genetics*, 2019. {Measures how score accuracy falls across ancestries in adult cohorts; embryos were not studied and no birth outcome was followed.}
[^forzano2022]: `paper` Forzano, F. et al. "The use of polygenic risk scores in pre-implantation genetic testing: an unproven, unethical practice." *European Journal of Human Genetics*, 2022. {A joint position statement from professional societies rather than a study; it reports no new outcome data.}
[^asrm2024]: `paper` Ethics Committee of the American Society for Reproductive Medicine. "The use of preimplantation genetic testing for polygenic risk scores (PGT-P): an Ethics Committee opinion." *Fertility and Sterility*, 2024.

==============================================================================
ARTICLE: posthuman
TITLE:   Posthuman
PORTAL:  Foundational Concepts
URL:     https://futurehumanwiki.com/wiki/posthuman
SOURCE:  https://futurehumanwiki.com/raw/posthuman
==============================================================================

---
title: "Posthuman"
slug: "posthuman"
type: "concept"
status: "speculative"
horizon: "indefinite"
categories: ["foundations", "society"]
tags: ["transhumanism", "posthumanism", "identity", "enhancement", "philosophy", "critical theory"]
summary: "A hypothetical being whose central capacities exceed the human maximum, and, in a separate scholarly tradition, a critique of the humanist subject itself."
updated: "2026-07-27"
---

```infobox
{
  "caption": "Concept in philosophy and futures studies",
  "rows": [
    { "label": "Two traditions", "value": "Transhumanist, critical" },
    { "label": "Term in use since", "value": "1970s" },
    { "label": "Canonical transhumanist text", "value": "Bostrom, 2008", "link": "/wiki/nick-bostrom" },
    { "label": "Canonical critical text", "value": "Hayles, 1999" },
    { "label": "Central capacities", "value": "Healthspan, cognition, emotion" },
    { "label": "Instances demonstrated", "value": "None" }
  ]
}
```

**Posthuman** names a being whose general capacities exceed the maximum attainable by any current human, to a degree that would put it outside the category "human" as that category is currently used. The word carries a second, largely unrelated meaning in the humanities, where posthumanism is a critique of the humanist picture of the self-contained rational subject rather than a forecast about future beings. The two usages share a prefix and almost nothing else, and the resulting confusion is a persistent feature of the literature.

## Transhuman, posthuman, and the gradient

In transhumanist writing the two terms mark positions on a continuum. A *transhuman* is a person in transition: someone using [[human-enhancement|enhancement technologies]] but still recognisably a member of the species. A posthuman lies beyond that, past whatever threshold makes the species label stop applying. [[fm-2030]] introduced the transitional sense in the 1970s, describing transhumans as evolutionary intermediaries identifiable by their reproductive choices, prostheses, mobility and rejection of traditional social forms.

The difficulty is that nobody has specified where the threshold sits, and most attempts to place it collapse under examination. A person with a [[cochlear-implant]] and a pacemaker who lives to 110 is not posthuman on anyone's account. A [[whole-brain-emulation|brain emulation]] running at a thousand times biological speed would be on almost everyone's. Between those poles the concept provides no principled cut, which is why careful writers treat "posthuman" as a direction rather than a status.

## Origins of the term

```timeline
[
  { "year": "1957", "title": "Huxley's transhumanism", "text": "Julian Huxley argues that the human species can transcend itself deliberately, and proposes 'transhumanism' as the name for the project." },
  { "year": "1973–1989", "title": "FM-2030's transitional human", "text": "Fereidoun Esfandiary describes transhumans as the earliest manifestations of a new evolutionary stage, and posthumans as what follows." },
  { "year": "1977", "title": "Posthumanism named", "text": "Literary theorist Ihab Hassan asks whether Western culture is entering a posthumanist phase in which the humanist subject no longer holds." },
  { "year": "1985", "title": "The cyborg as figure", "text": "Donna Haraway's cyborg manifesto uses the human–machine hybrid to attack the boundaries between organism, machine and gender." },
  { "year": "1990", "title": "Extropian formulation", "text": "Max More's founding transhumanist essays set out posthumanity as an explicit goal rather than a diagnosis of the present." },
  { "year": "1999", "title": "How We Became Posthuman", "text": "N. Katherine Hayles traces how information came to be treated as separable from its material substrate, and argues against that separation." },
  { "year": "2008", "title": "A capacity-based definition", "text": "Nick Bostrom defines a posthuman as a being with at least one general capacity far beyond any unaided human's, naming healthspan, cognition and emotion." }
]
```

## The transhumanist definition

The most-cited technical definition is [[nick-bostrom]]'s. A *posthuman capacity* is a general central capacity greatly exceeding the maximum attainable by any current human without recourse to new technological means; a posthuman is a being with at least one such capacity.[^bostrom2008] Bostrom names three central capacities: **healthspan**, the capacity to remain fully healthy and active; **cognition**, general intellectual capacities including memory and abstract reasoning; and **emotion**, the capacity to enjoy life and to respond appropriately to people and situations.

Three features of this definition do real work. It is *capacity-based*, so it says nothing about what a posthuman is made of; a biologically modified human, an emulation, and an engineered organism could all qualify. It is *disjunctive*, requiring only one capacity to cross the line, which is why radical [[longevity-escape-velocity|life extension]] alone would count. And it is *comparative to the current maximum*, not to the average, which sets a high bar: a drug that lifts a typical person to the level of an unusually healthy one is enhancement, not posthumanity.

> [!note] Terminology
> "Posthuman" and "posthumanism" are not the same word with different endings. Posthuman is a noun for a possible being; posthumanism is an intellectual position about the concept of the human. A scholar can be a posthumanist and regard posthumans as a naive fantasy, and many are.

The definition also deliberately avoids the question of moral status. Bostrom's essay is a reply to the charge that posthuman existence would be undesirable or unintelligible, and argues that healthspan, cognition and emotion are goods whose extreme versions people already recognise as valuable when they encounter unusually healthy, clever or emotionally resilient people. The argument concerns desirability, not feasibility, and Bostrom is explicit that it establishes nothing about whether such beings can be built.

## Critical posthumanism

The humanities tradition begins from a different question: not what comes after humans, but what was wrong with the concept of the human in the first place. Ihab Hassan's 1977 essay put the question in circulation; Donna Haraway's cyborg gave it a figure.[^haraway1985] The target is the humanist subject inherited from the Enlightenment — autonomous, disembodied in its reasoning, implicitly male, European and able-bodied — and the boundaries it maintains against animals, machines and the environment.

N. Katherine Hayles's *How We Became Posthuman* is the pivotal text for this wiki's subject matter because it is directly about the ideas covered here.[^hayles1999] Hayles traces how mid-century cybernetics came to treat information as a pattern that could be abstracted from any particular material carrier, and argues that this move — not any technology — is what created the posthuman imaginary. Her objection to [[mind-uploading]] is not that the engineering is hard but that the premise of [[substrate-independence]] smuggles in a fantasy of escape from the body which the science does not support. She ends the book arguing for a posthuman that recognises embodiment as constitutive rather than incidental.

Cary Wolfe pushes the point further, arguing that transhumanism is not a break with humanism but its intensification: it takes the humanist ideals of autonomy, mastery and rational self-improvement and proposes to achieve them more completely.[^wolfe2010] Rosi Braidotti's version shifts the emphasis to anti-anthropocentrism, extending moral consideration across species and questioning why human flourishing should be the organising value at all.

## Two traditions, one word

The traditions rarely engage. Transhumanist writing on posthumanity is largely analytic philosophy and applied futurism, concerned with capacities, probabilities and [[bioethics-of-enhancement|ethical permissibility]]. Critical posthumanism is largely continental theory, concerned with subject formation, representation and power. Each occasionally treats the other as a straw man: the first reads the second as wordplay, the second reads the first as Californian ideology with a philosophy department attached.

A third usage confuses matters further. Francis Fukuyama's *Our Posthuman Future* uses "posthuman" pejoratively, to name the outcome bioconservatives fear: a world in which biotechnology has dissolved the shared human nature he takes to ground equal political rights.[^fukuyama2002] That argument is set out in [[bioconservatism]], and it borrows the transhumanist sense of the word while reversing its valence.

> [!debate] Is the posthuman a threshold or a direction?
> Bostrom's definition implies a crossing point, since a capacity either exceeds the current human maximum or does not. Critics inside transhumanism note that capacities are continuous and technology-relative: literacy, spectacles and antibiotics have all raised human maxima, so the reference class keeps moving and the threshold recedes as it is approached.

## Objections

The definitional objection is that "greatly exceeding" is unquantified and does no work. A being that lives to 200 and one that lives to 20,000 both satisfy it, though almost nothing else about them is comparable.

The empirical objection is that no central capacity has yet been pushed past the human maximum by any technology. Assistive devices restore function; [[nootropics]] produce small and inconsistent effects in healthy adults; [[genetic-enhancement-of-intelligence]] is limited by the polygenic architecture of the traits involved; [[brain-computer-interface|brain–computer interfaces]] currently return a fraction of natural capability to people who have lost it. Every proposed route to posthumanity is at an earlier stage than its advocates' timelines assume, and this wiki's per-article readiness assessments reflect that gap.

The identity objection asks whether the posthuman would be the same person as the human who became one — the problem treated in [[personal-identity-and-continuity]]. Gradual modification looks survivable on most theories; a single discontinuous transformation does not.

The political objection, made from both the left and the disability-rights tradition, is that positing a superior future being encodes a judgement about present ones. If a posthuman is defined by capacities, then existing people are ranked by how far short they fall, a structure examined in [[disability-rights-and-enhancement]] and one reason [[morphological-freedom]] is framed as a right to self-modification rather than an obligation to improve.

## Outlook

No agreed criterion exists for recognising a posthuman if one appeared, and the candidates are not converging. A capacity threshold makes the category depend on a moving human maximum. A substrate criterion excludes biologically enhanced humans and includes any sufficiently capable machine, which collapses the term into [[artificial-general-intelligence]]. A phenomenological criterion would require settling the questions raised in [[machine-consciousness]], which remain open. The unresolved issue is whether "posthuman" picks out a natural kind at all, or whether it functions as a placeholder for whatever lies past the current limits of imagination about human capability — a placeholder that each generation refills with the technologies it happens to have.

## See also

- [[transhumanism]]
- [[human-enhancement]]
- [[nick-bostrom]]
- [[bioconservatism]]
- [[substrate-independence]]
- [[morphological-freedom]]
- [[future-of-humanity]]
- [[technological-singularity]]

## References

[^bostrom2008]: `book` Bostrom, N. "Why I Want to Be a Posthuman When I Grow Up." In Gordijn, B. and Chadwick, R. (eds), *Medical Enhancement and Posthumanity*. Springer, 2008. {An argument that posthuman existence would be desirable and worth wanting, not a claim that it is achievable.}
[^haraway1985]: `paper` Haraway, D. "A Manifesto for Cyborgs: Science, Technology, and Socialist Feminism in the 1980s." *Socialist Review*, 1985. {A political essay using the cyborg as a figure for broken boundaries; it is not a claim about human-machine hybrids as engineering.}
[^hayles1999]: `book` Hayles, N. K. *How We Became Posthuman: Virtual Bodies in Cybernetics, Literature, and Informatics*. University of Chicago Press, 1999.
[^wolfe2010]: `book` Wolfe, C. *What Is Posthumanism?* University of Minnesota Press, 2010.
[^fukuyama2002]: `book` Fukuyama, F. *Our Posthuman Future: Consequences of the Biotechnology Revolution*. Farrar, Straus and Giroux, 2002.

==============================================================================
ARTICLE: exoskeleton
TITLE:   Powered exoskeletons
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/exoskeleton
SOURCE:  https://futurehumanwiki.com/raw/exoskeleton
==============================================================================

---
title: "Powered exoskeletons"
slug: "exoskeleton"
type: "technology"
status: "emerging"
horizon: "present"
trl: 8
categories: ["bodies", "enhancement"]
tags: ["robotics", "rehabilitation", "mobility", "wearables", "human augmentation", "actuators"]
summary: "Wearable powered frames that apply force in parallel with the wearer's own limbs, used for gait rehabilitation, industrial load support, and proposed strength augmentation."
updated: "2026-07-27"
humanEvidence: "People with spinal cord injury or stroke walk in these devices under supervision, though reviews find the evidence too thin to show an advantage over conventional gait training; the clearest measured gains are metabolic savings in healthy walkers."
access: "Cleared for clinical use and, in the United States since 2014, for personal use; medical devices cost tens of thousands of dollars or more with limited payer coverage, while passive industrial exosuits are sold outright."
reversibility: "reversible"
issues: ["The systematic-review conclusions in the rehabilitation section carry no citations."]
---

```infobox
{
  "caption": "Class of wearable robot",
  "rows": [
    { "label": "First powered prototype", "value": "1960s (Hardiman, Belgrade)" },
    { "label": "Main clinical use", "value": "Gait rehabilitation" },
    { "label": "First US personal-use clearance", "value": "2014" },
    { "label": "Actuation", "value": "Electric, hydraulic, pneumatic, passive spring" },
    { "label": "Binding constraint", "value": "Actuator and battery power density" },
    { "label": "Best lab metabolic saving", "value": "Roughly 15–20% for walking" },
    { "label": "Readiness", "value": "TRL 8 (medical), TRL 5 (full-body augmentation)" }
  ]
}
```

**Powered exoskeletons** are wearable machines whose structure runs alongside the body's own
skeleton and whose actuators apply torques across the wearer's joints. They are used clinically to
move the legs of people with paralysis, industrially to offload weight from a worker's back and
shoulders, and militarily — mostly in prototype — to carry loads. The engineering problem is
deceptively hard, because an exoskeleton must produce more benefit than the metabolic and
mechanical cost of carrying it. Unlike the implanted devices grouped under
[[neuroprosthetics]], or the skeletal anchorage described in [[osseointegration]], an exoskeleton
touches the body only at straps and cuffs, which is both its principal safety advantage and the
source of most of its engineering difficulty.

```figure
{"key": "exoskeleton-hal", "caption": "HAL reads surface electromyography at the thigh and drives the joint the wearer is already trying to move."}
```

## How they work

Three subsystems define a device. The *structure* transmits load, ideally into the ground rather than
into the wearer, and must keep its joints aligned with human joints that are not simple hinges;
misalignment produces shear at the attachment points and is a common source of discomfort and
abandonment. The *actuators* supply torque, most often through electric motors with harmonic or
ball-screw transmissions, sometimes hydraulically for high force, and sometimes not at all in
passive designs that store energy in springs. The *controller* decides when and how much to assist.

Intent detection is the control problem. Options range from simple state machines triggered by
weight shift and joint angle, through force sensors in the shoe, to surface electromyography that
reads the wearer's muscle activity before movement begins — the approach taken by Japan's HAL
system, which uses skin-surface bioelectrical signals as its control input and shares its signal
chain with [[myoelectric-prosthetics]]. Cortical control through a
[[brain-computer-interface]] has been demonstrated in single participants and adds surgical risk to
a problem that peripheral sensors mostly solve; the same intent-estimation methods used in
[[neural-decoding]] apply, but the signal is easier to obtain at the muscle.

The distinction between *rehabilitation* devices, which move a limb the wearer cannot move, and
*augmentation* devices, which assist a limb the wearer can move, matters more than the hardware
suggests. The first must be safe when the wearer contributes nothing; the second must interfere with
nothing when the wearer needs no help, a much harder requirement, because an actuator that fights
the user is worse than no actuator at all.

## Development history

```timeline
[
  { "year": "1965–1971", "title": "Hardiman", "text": "General Electric builds a full-body hydraulic master-slave suit weighing around 680 kg. Only one arm is ever operated, and it moves violently enough to be considered unsafe with a person inside." },
  { "year": "Late 1960s", "title": "Active exoskeletons in Belgrade", "text": "Miomir Vukobratović's group at the Mihajlo Pupin Institute builds legged exoskeletons for rehabilitating paraplegic patients, developing the zero-moment-point criterion that later underpins bipedal robotics." },
  { "year": "2000–2008", "title": "DARPA funding and BLEEX", "text": "A US programme on exoskeletons for human performance augmentation funds the Berkeley Lower Extremity Exoskeleton and spawns Ekso Bionics and Lockheed's HULC." },
  { "year": "2011–2016", "title": "Clinical clearances", "text": "Medical exoskeletons for spinal cord injury and stroke gain regulatory clearance in Japan, Europe, and the United States; the first personal-use clearance in the US comes in 2014." },
  { "year": "2015–2022", "title": "Soft suits and optimized assistance", "text": "Textile-based exosuits and human-in-the-loop optimization of assistance profiles produce the largest measured reductions in the metabolic cost of walking." },
  { "year": "2019–2025", "title": "Military programmes retrench", "text": "The US special operations powered-armour programme ends without a fielded suit, and several full-body industrial exoskeleton efforts are scaled back, while passive industrial exosuits find a durable market." }
]
```

## Evidence in rehabilitation

Overground exoskeletons let many people with complete or incomplete spinal cord injury stand and
walk in a supervised setting, and the devices are safe in trained hands. What is not established is
that they produce better functional outcomes than conventional gait training. Systematic reviews of
exoskeleton-assisted walking after spinal cord injury and stroke repeatedly conclude that the
evidence base is small, heterogeneous, and insufficient to demonstrate superiority; secondary
benefits often claimed — bowel function, spasticity, bone density, cardiovascular fitness — are
supported mainly by small studies and self-report.

Practical constraints compound this. Walking speeds achieved with lower-limb exoskeletons are
typically far below what is needed for community ambulation, nearly all devices require forearm
crutches and therefore intact upper-limb function, donning takes minutes and often requires
assistance, and users cannot navigate the stairs, kerbs, and uneven ground that make up ordinary
environments. For most people with paraplegia a wheelchair remains faster, cheaper, and more
reliable, which is not a failure of the technology so much as a statement of the standard it has to
beat. Disability advocates have also questioned the emphasis on restoring upright walking over
improving wheelchair access, an argument set out under
[[disability-rights-and-enhancement]]. Implanted alternatives that stimulate the user's own spinal
circuitry attack the same problem from inside the body, and epidural stimulation trials have
produced volitional stepping in participants who could not achieve it in a powered frame.

Where exoskeletons appear more clearly useful is as therapy delivery: a device that lets one
therapist provide high-repetition, high-dose gait practice instead of two or three therapists
manually moving a patient's legs. That is a labour-productivity argument rather than a functional
one.

Soft exosuits are the most promising rehabilitation variant. A textile suit applying modest
assistance to ankle plantarflexion and hip flexion improved walking symmetry and reduced metabolic
cost in people with post-stroke hemiparesis.[^awad2017] They are lighter, do not require joint
alignment, and cannot lock a limb, which reduces the safety burden.

## Industrial and military use

Industrial exoskeletons have quietly become the largest deployed category, and most are passive.
Spring-loaded shoulder supports for overhead work and back supports that store energy during
forward bending reduce measured muscle activity in the assisted region, which is well established.
Whether they reduce injury rates over years is not, and there is a recognized concern that load
removed from one joint reappears at another — a back support can increase demand on the hips or
alter gait. Long-term field trials are scarce.

Military programmes have a consistent record of ambition followed by retrenchment. The US special
operations "tactical assault light operator suit" effort was terminated in 2019 without producing a
fielded system, and several powered load-carriage exoskeletons have been discontinued. The
recurring failure is not strength but endurance: a soldier's mission profile demands hours of
untethered operation, and no battery chemistry supplies that at acceptable weight. Lighter,
narrower-purpose devices such as powered knee assists for load carriage have fared better.

The military case also raises the coordination problem examined under
[[enhancement-arms-race]]: capabilities adopted because an adversary might adopt them, with no
individual actor able to stop unilaterally. Unlike pharmacological or genetic augmentation, an
exoskeleton is removable, which makes it the least ethically fraught form of
[[human-enhancement]] and, for the same reason, the one likely to be normalized first. Competitive
sport has already banned assistive devices outright rather than attempting to regulate their degree,
a simpler line than the one drawn for prosthetic athletes under
[[enhancement-in-sport]].

Space agencies are a distinct customer. Crews returning from long missions lose bone and muscle at
rates described under [[space-medicine]], and exoskeletons have been proposed both as inflight
resistive-exercise devices and as ground-support equipment during readaptation; the countermeasure
problem in [[microgravity-adaptation]] remains dominated by conventional exercise hardware.

## The power and actuator problem

The physics is unforgiving. An exoskeleton that assists walking must overcome the metabolic penalty
of its own mass, and mass carried distally — at the ankle or foot — costs several times more than
mass at the waist. Electric motors have high power density but low torque density, so they need
gearboxes, which add mass, friction, and backdrive resistance. Hydraulics deliver torque but need a
pump, fluid, and heat rejection. Batteries are the hard ceiling: lithium-ion cells store on the
order of a few hundred watt-hours per kilogram, well over an order of magnitude below liquid fuel,
so untethered full-body augmentation trades run time against weight in a loop that has no good
solution.

This is why the strongest results come from minimal devices. An entirely unpowered ankle exoskeleton
using a spring and a mechanical clutch reduced the metabolic cost of walking by roughly seven
percent, with no battery at all.[^collins2015] Portable powered ankle exoskeletons whose assistance
profile is optimized to the individual wearer through iterative real-world testing have produced
reductions on the order of fifteen to twenty percent, along with increased walking
speed.[^slade2022] Both results concern healthy walking on flat ground, and neither approaches the
popular image of powered armour.

> [!key] Assistance has to be personalized
> The largest metabolic gains come not from stronger actuators but from tuning the timing and shape
> of assistance to the individual, discovered automatically by measuring the wearer's energy use
> across many candidate profiles. A generic assistance pattern can make walking harder rather than
> easier.

## Outlook

Three trajectories are plausible and largely independent. Passive and lightly powered industrial
exosuits could keep spreading on economics alone, since their case does not depend on any advance in
actuators or batteries. Medical exoskeletons are likely to stay a rehabilitation-clinic technology
unless payers accept them for personal use at scale, a process that has begun in some health systems
but proceeds slowly and unevenly, with the access pattern described in [[access-and-inequality]].

The demographic argument is the most consequential and least discussed. Age-related loss of muscle
mass and power is the mechanism behind most mobility loss in old age, and no drug reliably reverses
it; the best evidence remains for training itself, discussed under
[[exercise-and-aging]], and mobility is one of the functional domains any serious measure of
[[healthspan]] has to capture. A light, cheap, comfortable hip or ankle assist that lets an eighty-year-old
keep walking would affect far more people than any spinal cord injury device, and it is a much
easier engineering target than powered armour. Whether such a device can be made unobtrusive enough
that people will actually wear it is a design question, not a physics one, and it is where the
field's near-term value most likely lies.

## See also

- [[myoelectric-prosthetics]]
- [[osseointegration]]
- [[neuroprosthetics]]
- [[human-enhancement]]
- [[enhancement-in-sport]]
- [[disability-rights-and-enhancement]]
- [[exercise-and-aging]]
- [[brain-computer-interface]]

## References

[^awad2017]: `paper` Awad, L. N. et al. "A soft robotic exosuit improves walking in patients after stroke." *Science Translational Medicine*, 2017. {A small study of ambulatory stroke survivors measuring the immediate effect of assistance, not the outcome of a training programme.}

[^collins2015]: `paper` Collins, S. H., Wiggin, M. B., Sawicki, G. S. "Reducing the energy cost of human walking using an unpowered exoskeleton." *Nature*, 2015. {The saving is in healthy adults walking on level ground in a laboratory, and says nothing about impaired gait.}

[^slade2022]: `paper` Slade, P., Kochenderfer, M. J., Delp, S. L., Collins, S. H. "Personalizing exoskeleton assistance while walking in the real world." *Nature*, 2022.

==============================================================================
ARTICLE: precautionary-principle
TITLE:   Precautionary principle
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/precautionary-principle
SOURCE:  https://futurehumanwiki.com/raw/precautionary-principle
==============================================================================

---
title: "Precautionary principle"
slug: "precautionary-principle"
type: "concept"
status: "contested"
horizon: "present"
categories: ["society", "foundations"]
tags: ["governance", "risk", "regulation", "biotechnology", "law", "uncertainty"]
summary: "The regulatory doctrine that scientific uncertainty about a serious or irreversible harm is not by itself a reason to permit the activity that might cause it."
updated: "2026-07-27"
issues: ["The WTO 2006 panel, 2018 CJEU ruling and 2024 Philippine decision carry no citations."]
---

```infobox
{
  "caption": "Doctrine in risk regulation and environmental law",
  "rows": [
    { "label": "Origin", "value": "German Vorsorgeprinzip, 1970s" },
    { "label": "Canonical weak form", "value": "Rio Declaration, 1992" },
    { "label": "Canonical strong form", "value": "Wingspread Statement, 1998" },
    { "label": "Legal force", "value": "EU treaty law; several conventions" },
    { "label": "Best-known critic", "value": "Cass Sunstein" },
    { "label": "Transhumanist counter", "value": "Proactionary principle", "link": "/wiki/max-more" }
  ]
}
```

**The precautionary principle** holds that where an activity threatens serious or irreversible harm, the absence of scientific certainty about that harm is not a sufficient reason to allow the activity to proceed unrestricted. It exists in weak versions that merely license regulators to act under uncertainty and strong versions that shift the burden of proof onto whoever proposes the activity. The difference between those versions is not a matter of degree, and most arguments about the principle are really arguments about which one is being invoked.

## Formulations

```compare
{
  "columns": ["Weak form (Rio, 1992)", "Strong form (Wingspread, 1998)"],
  "rows": [
    { "label": "Trigger", "values": ["Threat of serious or irreversible damage", "Any threat of harm to health or environment"] },
    { "label": "Effect", "values": ["Uncertainty is not grounds for postponing measures", "Precautionary measures should be taken"] },
    { "label": "Burden of proof", "values": ["Left with the regulator", "Shifted to the proponent"] },
    { "label": "Cost test", "values": ["Measures must be cost-effective", "None specified"] },
    { "label": "Typical venue", "values": ["Treaty and trade law", "Advocacy and civil-society statements"] }
  ]
}
```

The weak form authorises action; the strong form requires it. A third and stronger reading, sometimes attributed to the principle by its critics and rarely defended in print, would prohibit any activity whose safety cannot be affirmatively demonstrated. That reading is genuinely incoherent, since no activity can be shown safe in that sense and inaction has consequences of its own — but it is a straw man of the versions that appear in law.

## Origins and legal status

The German *Vorsorgeprinzip* — roughly, the principle of taking care in advance — entered environmental legislation in the 1970s as a rationale for acting on forest damage and air pollution before causal chains were established. The idea spread through North Sea protection conferences in the 1980s and reached its most-quoted formulation as Principle 15 of the 1992 Rio Declaration.

In European Union law the principle has treaty status as one of the bases of environmental policy, and a European Commission communication in 2000 set out how it should be applied: measures must be proportionate, non-discriminatory, consistent with comparable measures elsewhere, based on an examination of costs and benefits, and subject to review as evidence accumulates.[^ec2000] That document is the most operationally careful statement of the doctrine in existence, and it is considerably more restrained than the principle's reputation suggests.

The Cartagena Protocol on Biosafety, adopted in 2000, is the main instrument applying precaution to biotechnology specifically, permitting parties to restrict imports of living modified organisms where scientific information is insufficient. UNESCO's world commission on the ethics of scientific knowledge issued a report in 2005 attempting a general definition and application framework.

## How it operates in biotechnology

The principle's record in this field is genuinely mixed, and both camps cite it selectively.

**Where precaution has been vindicated.** The 1975 [[asilomar-conference]] is the field's founding case: researchers imposed a moratorium on recombinant DNA work under acknowledged uncertainty, developed containment levels, and resumed. The moratorium lasted under a year and produced a containment framework that is still the template for self-governance in the field. The requirement that drug sponsors demonstrate safety and efficacy before marketing — a burden-shifting rule of exactly the strong-form type — is uncontroversial and followed a series of harms including thalidomide.

**Where precaution has been costly.** The European Union's de facto moratorium on approvals of genetically modified crops from 1999 was found by a World Trade Organization panel in 2006 to have caused undue delay. A 2018 Court of Justice ruling brought organisms produced by genome editing within the EU's GMO directive, subjecting edits indistinguishable from natural mutations to the full approval regime; the Commission subsequently proposed a separate regime for new genomic techniques. In 2024 a Philippine appellate court revoked biosafety permits for Golden Rice and Bt eggplant on precautionary grounds, a decision that was appealed. In each case, the costs of delay fell on people who would have benefited and are not counted in the risk assessment.

**Where the argument is live.** [[gene-drive]] releases are the clearest current test, because a self-propagating edit cannot be recalled and field trials are the only way to obtain the missing information. Proposals for [[germline-editing]] are governed by an explicitly precautionary international consensus, discussed in [[governance-of-genome-editing]], which the [[he-jiankui-affair]] hardened rather than created. The 2024 call by a large group of scientists to halt work toward [[mirror-life]] is a precautionary argument in its purest modern form: the authors argue that chirally inverted organisms could evade immune recognition across many species, that no adequate defence is known, and that the research should stop before the capability exists.[^adamala2024]

> [!key] Why the biotechnology cases differ
> Most environmental applications of the principle concern harms that are gradual and in principle reversible. The cases that dominate this wiki — self-spreading edits, heritable modification, novel replicators — involve irreversibility as a structural feature. That is the condition under which even the principle's critics tend to concede it has force.

## Criticisms

The central objection, argued at length by Cass Sunstein, is that the principle is incoherent as a decision rule because it forbids the very steps it requires.[^sunstein2005] Regulation has risks; so does non-regulation. Banning a pesticide risks crop loss; approving it risks toxicity. If the principle counsels avoiding uncertain serious harm, it counsels both action and inaction simultaneously, and it appears to give determinate guidance only because attention has been selectively directed at one side of the ledger. Sunstein attributes this selectivity to known features of risk perception: availability, loss aversion, and greater aversion to harms caused by human action than to statistically larger harms caused by omission.

A second objection concerns *risk–risk tradeoffs*: interventions displace risk rather than removing it, and the displaced risk is systematically undercounted because it lands on different people at a later time.

A third objection is that the principle has no stopping rule. Since certainty is unavailable in principle, any threshold for "sufficient evidence" is a judgement imported from outside the principle, which means the principle is not doing the work its invocation implies.

A fourth is that it is applied asymmetrically to novel technologies. Existing practices with well-characterised large harms are grandfathered, while novel practices with speculative small harms face the burden — a pattern visible in the regulatory treatment of [[xenotransplantation]] relative to the ongoing mortality documented in [[organ-shortage]].

## The proactionary reply

[[max-more]] proposed the *proactionary principle* in 2004 as an explicit counterweight, articulating a presumption in favour of proceeding: freedom to innovate is valuable, restriction has costs that must be counted symmetrically, and evaluation should use the best available evidence rather than worst cases, weighting probability rather than mere possibility.[^more2013] The formulation has roots in [[extropianism]] and has been developed at book length by Steve Fuller and Veronika Lipinska.

Its weakness mirrors the precautionary principle's. Stated weakly it is a plea for symmetric accounting, which almost everyone accepts. Stated strongly it presumes in favour of action under uncertainty, which is as arbitrary as presuming against. The strand of [[effective-accelerationism]] that treats technological advance as intrinsically good takes the strong reading, generally without the accounting.

## Asymmetric-risk versions

The most defensible modern formulations abandon the attempt at a general rule and specify the conditions under which precaution is rational. Nassim Taleb and colleagues argue that precaution is warranted where potential harm is *systemic* and *ruinous* — where the affected system is coupled enough that damage propagates and the downside includes irreversible collapse — and unwarranted for risks that are local and bounded, however alarming.[^taleb2014] On their analysis, an activity with a small chance of unbounded harm repeated indefinitely has a probability of ruin approaching one, which is an argument from the structure of the payoff rather than from uncertainty as such.

Sunstein himself endorses a narrower "anti-catastrophe" version on similar grounds: where a harm is potentially irreversible and catastrophic, and the option to wait has value, paying to preserve future flexibility can be justified even under deep uncertainty. [[nick-bostrom]]'s work on [[existential-risk]] arrives at the same structure from a different direction, and [[differential-technological-development]] can be read as an attempt to convert that structure into a policy that sequences technologies rather than blocking them.

## Outlook

The principle's practical future in biotechnology depends on a question it cannot answer internally: who counts the costs of delay. Approval regimes make the risks of action visible and attributable, while the harms of inaction — untreated disease, forgone yield, deaths on transplant waiting lists — are diffuse and statistical. Proposals to fix the asymmetry range from mandatory counting of forgone benefits in regulatory impact assessment to conditional approval with mandatory post-market surveillance, the model already used for accelerated drug approvals. None has been adopted broadly enough to test whether it changes outcomes, and the technologies now arriving in [[synthetic-genomes]] and self-propagating systems are precisely the ones where getting the accounting wrong in either direction is expensive.

## See also

- [[bioconservatism]]
- [[differential-technological-development]]
- [[existential-risk]]
- [[dual-use-research]]
- [[governance-of-genome-editing]]
- [[mirror-life]]
- [[max-more]]
- [[asilomar-conference]]

## References

[^ec2000]: `report` European Commission. *Communication from the Commission on the Precautionary Principle*, COM(2000) 1. Brussels, 2000. {A Commission communication is guidance on how the principle should be applied, not a binding instrument in itself.}
[^adamala2024]: `paper` Adamala, K. P. et al. "Confronting risks of mirror life." *Science*, 2024. {A policy argument signed by a large group of researchers, not an experimental result; no mirror organism exists to have been studied.}
[^sunstein2005]: `book` Sunstein, C. R. *Laws of Fear: Beyond the Precautionary Principle*. Cambridge University Press, 2005.
[^more2013]: `book` More, M. "The Proactionary Principle: Optimizing Technological Outcomes." In More, M. and Vita-More, N. (eds), *The Transhumanist Reader*. Wiley-Blackwell, 2013.
[^taleb2014]: `preprint` Taleb, N. N., Read, R., Douady, R., Norman, J. and Bar-Yam, Y. "The Precautionary Principle (with Application to the Genetic Modification of Organisms)." Extreme Risk Initiative working paper, NYU School of Engineering, 2014. {A working paper rather than a reviewed article; its argument is about payoff structure, and it takes genetically modified crops as the worked example.}

==============================================================================
ARTICLE: prime-editing
TITLE:   Prime editing
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/prime-editing
SOURCE:  https://futurehumanwiki.com/raw/prime-editing
==============================================================================

---
title: "Prime editing"
slug: "prime-editing"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 6
categories: ["genetics"]
tags: ["crispr", "genome editing", "gene therapy", "reverse transcriptase", "biotechnology"]
summary: "A genome editing method in which a nicking Cas9 fused to a reverse transcriptase writes a new DNA sequence specified by an extended guide RNA."
updated: "2026-07-27"
humanEvidence: "Human use is one ex vivo trial in chronic granulomatous disease, which reported restored immune-cell function in a treated patient in 2025; in vivo correction has been shown only in mice."
access: "Not obtainable as a therapy: no prime editing product is approved anywhere as of 2026 and human use is confined to early-phase trials; the reagents are sold for research."
reversibility: "irreversible"
issues: ["The 2025 chronic granulomatous disease result is stated without a citation."]
---

```infobox
{
  "caption": "Genome editing platform",
  "rows": [
    { "label": "Type", "value": "Nickase–reverse transcriptase fusion" },
    { "label": "First described", "value": "2019" },
    { "label": "Developed by", "value": "David Liu laboratory" },
    { "label": "Guide molecule", "value": "pegRNA" },
    { "label": "Double-strand break", "value": "No" },
    { "label": "Edits possible", "value": "All 12 substitutions, small indels" },
    { "label": "Readiness", "value": "TRL 6" }
  ]
}
```

**Prime editing** installs a chosen DNA sequence at a chosen site by nicking one strand of the double helix and then writing new bases onto the exposed end from an RNA template carried by the guide itself. Described in 2019, it removes two constraints at once: it needs no double-strand break, and unlike [[base-editing]] it is not restricted to particular chemical conversions.[^anzalone2019] Its practical difficulty is the reverse of that generality — the machine is large, the design space for each target is wide, and efficiency varies from excellent to negligible depending on the site.

## How it works

Three things have to happen in order, and each can fail independently: the editor must find and nick the target, the reverse transcriptase must copy the intended sequence onto the exposed strand, and the cell must keep the new sequence rather than discard it. The third step is the one the designer controls least.

### The pegRNA

The editor is a Cas9 nickase, disabled in one of its two nuclease domains so that it cuts only the strand containing the PAM, fused to an engineered reverse transcriptase. The guide is extended into a **prime editing guide RNA**, or pegRNA, which carries three functional parts: a spacer that specifies the target, a primer binding site complementary to the nicked strand, and a reverse transcription template encoding the desired new sequence.

After the nick, the freed 3' end of the DNA anneals to the primer binding site. The reverse transcriptase then extends that end, copying the template into DNA. The result is a flap of newly written DNA carrying the edit, competing with the original flap for reincorporation into the duplex.

### Resolving the flap

Whether the edit survives depends on how the cell resolves that competition. Excision of the original flap and ligation of the new one installs the change on one strand; the resulting mismatch must then be resolved in the edit's favour rather than corrected back. Mismatch repair frequently reverses prime edits, which is why an early and effective improvement was transient suppression of that pathway.[^chen2021] A second strategy nicks the unedited strand as well, marking it as the one to be rewritten. That raises efficiency but reintroduces a modest risk of indels, since two nearby nicks approximate a double-strand break.

```timeline
[
  { "year": "2019", "title": "Prime editing described", "text": "Anzalone and colleagues report search-and-replace editing without double-strand breaks or donor DNA, demonstrating all twelve base substitutions and small insertions and deletions in human cells." },
  { "year": "2021", "title": "Mismatch repair identified as the brake", "text": "Transient suppression of mismatch repair substantially raises editing efficiency, defining the PE4 and PE5 systems." },
  { "year": "2022", "title": "Guide RNA stabilised", "text": "Structured RNA motifs added to the pegRNA 3' end protect it from degradation, and paired-pegRNA designs enable larger replacements and integrase-mediated insertions." },
  { "year": "2023–2024", "title": "Compact and stabilised editors", "text": "Evolved reverse transcriptases shrink the construct, and fusion of an RNA-binding protein that recognises the pegRNA improves performance across cell types." },
  { "year": "2025", "title": "First clinical data", "text": "Initial results are reported from the first prime editing trial, an ex vivo therapy for chronic granulomatous disease." }
]
```

## Development history

The original report defined a numbered progression that the field still uses. PE1 paired the nickase with an unmodified viral reverse transcriptase and worked poorly. PE2 used an engineered enzyme and worked well enough to be usable. PE3 added the second nick. Later systems addressed mismatch repair, pegRNA stability, and the size of the reverse transcriptase itself, with each generation typically improving efficiency several-fold in the cell types where it had previously been weakest.

Two extensions widened the scope beyond small edits. Paired pegRNAs that write complementary flaps allow replacement of a whole segment rather than a few bases. Coupling prime editing to a serine integrase — using the editor to install a short recombinase landing site and the integrase to insert a payload into it — permits insertions of many kilobases, closing part of the gap between editing and gene addition.[^yarnall2023] That capability connects prime editing to the larger-scale ambitions described in [[synthetic-genomes]] and [[recoded-organisms]], where the goal is to rewrite rather than repair.

## Current state

Prime editing is a standard research tool for installing precise variants in cell lines, [[induced-pluripotent-stem-cells]] and [[organoids]], where the ability to write an exact patient mutation makes disease modelling far cleaner than nuclease-based approaches allow. In animals, prime editing has corrected the causal mutation in mouse models of metabolic liver disease and several other monogenic conditions, in each case in mice rather than in humans.[^bock2022]

Clinical translation began recently. The first prime editing therapy to reach patients treats chronic granulomatous disease by correcting a mutation in blood stem cells outside the body; initial results reported in 2025 indicated restoration of the missing immune-cell function in a treated patient. The company behind it subsequently restructured and redirected effort toward liver targets, a reminder that in this field the binding constraint is often capital rather than biology. No prime editing product is approved anywhere as of 2026.

> [!note] Terminology
> "Search and replace" is the phrase from the original paper and is apt for the mechanism, but it invites a word-processor analogy that overstates reliability. There is no undo, no confirmation step, and no guarantee the replacement is installed in every cell.

## Limitations

Delivery is the acknowledged bottleneck. Fusing a reverse transcriptase to a Cas9 nickase makes a protein substantially larger than Cas9 alone, and larger again than a base editor; with its pegRNA the construct sits well beyond the packaging capacity of a single [[aav-vectors|adeno-associated virus]]. Workarounds include splitting the protein across two vectors and reassembling it inside the cell, packaging messenger RNA in [[lipid-nanoparticles]], and delivering preformed protein in engineered virus-like particles. Each adds manufacturing complexity, and none has yet been shown to work efficiently in tissues outside the liver and blood.

Efficiency is also site-dependent in ways that remain partly empirical. The length of the primer binding site and of the template, the local chromatin state, and the cell's mismatch repair activity all matter, so optimising a new target still involves screening dozens of pegRNA designs. Reported efficiencies in primary human cells commonly sit well below those in immortalised cell lines, and papers do not always make the distinction obvious.

## Risks

Because installing an edit requires three separate base-pairing events — spacer, primer binding site and template — prime editing is intrinsically more discriminating than a single-nick or single-cut system, and measured [[crispr-off-target-effects|off-target editing]] at guide-similar sites is generally lower than for [[crispr-cas9]]. The characteristic unintended products are different in kind: fragments of the pegRNA scaffold copied into the target site, tandem duplications, and small indels where the second nick was used.

For [[somatic-gene-therapy]], the risk profile is favourable enough that the main open questions concern delivery vectors rather than the editor. For heritable applications, prime editing is sometimes advanced as the tool that would make [[germline-editing]] technically defensible. That argument runs into the same objections as its predecessors: efficiency in embryos is not the issue, mosaicism and verification are, and the international frameworks catalogued in [[governance-of-genome-editing]] are not written in terms of which enzyme is used — a posture that has not moved since the [[he-jiankui-affair]]. The practical alternative for most heritable disease risk remains embryo selection rather than editing.

## Outlook

Prime editing's ceiling is set by a design problem as much as a chemical one. Whether pegRNA design becomes predictable depends on whether rules learned from large screening datasets transfer to primary cells and to loci nobody has tested. If they do, retargeting turns into a lookup and the platform scales; if they do not, every new mutation stays a small research project, which is a poor foundation for a technology whose stated purpose is treating rare mutations one at a time.

The comparison that will settle the technology's role is with the simpler editors rather than with nucleases. Where a transition substitution would fix the mutation, a base editor is smaller, better characterised in humans, and further along in trials; prime editing earns its complexity only at targets the simpler chemistries cannot reach. Compared with [[casgevy]] and the other approved editing therapies, its repertoire is wider and the clinical evidence behind it far thinner. A secondary question, shared with [[gene-therapy-for-aging]] and with the ambitions of [[access-and-inequality|equitable access]] generally, is whether a platform whose selling point is one bespoke correction per patient can ever be manufactured and approved at a cost that matches the size of each patient population.

## See also

- [[base-editing]]
- [[crispr-cas9]]
- [[epigenome-editing]]
- [[crispr-off-target-effects]]
- [[aav-vectors]]
- [[somatic-gene-therapy]]
- [[synthetic-genomes]]
- [[germline-editing]]

## References

[^anzalone2019]: `paper` Anzalone, A. V. et al. "Search-and-replace genome editing without double-strand breaks or donor DNA." *Nature*, 2019. {The demonstrations are in cultured cells, mostly human cell lines; efficiencies in primary cells and in living tissue are generally much lower.}
[^chen2021]: `paper` Chen, P. J. et al. "Enhanced prime editing systems by manipulating cellular determinants of editing outcomes." *Cell*, 2021.
[^bock2022]: `paper` Böck, D. et al. "In vivo prime editing of a metabolic liver disease in mice." *Science Translational Medicine*, 2022. {Mice, and in liver, the tissue current delivery vehicles reach most easily; the result does not by itself extend to other organs or to people.}
[^yarnall2023]: `paper` Yarnall, M. T. N. et al. "Drag-and-drop genome insertion of large sequences without double-strand DNA cleavage using CRISPR-directed integrases." *Nature Biotechnology*, 2023.

==============================================================================
ARTICLE: procreative-beneficence
TITLE:   Procreative beneficence
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/procreative-beneficence
SOURCE:  https://futurehumanwiki.com/raw/procreative-beneficence
==============================================================================

---
title: "Procreative beneficence"
slug: "procreative-beneficence"
type: "concept"
status: "contested"
horizon: "present"
categories: ["society", "reproduction"]
tags: ["bioethics", "reproduction", "selection", "eugenics", "philosophy"]
summary: "Julian Savulescu's principle that prospective parents have moral reason to select, among the children they could have, the one expected to live the best life."
updated: "2026-07-27"
---

```infobox
{
  "caption": "Principle in reproductive ethics",
  "rows": [
    { "label": "Proposed by", "value": "Julian Savulescu", "link": "/wiki/julian-savulescu" },
    { "label": "First stated", "value": "2001" },
    { "label": "Venue", "value": "Bioethics" },
    { "label": "Form", "value": "Prima facie reason, not a duty" },
    { "label": "Applies to", "value": "Selection, not coercion" },
    { "label": "Status", "value": "Widely discussed, widely rejected" }
  ]
}
```

**Procreative beneficence** is the principle, stated by [[julian-savulescu]] in 2001, that couples who select which of several possible children to bring into existence have significant moral reason to choose the one expected to have the best life, or at least as good a life as the alternatives, on the basis of the relevant information available — including information about traits that are not diseases.[^sav2001] It is one of the most cited claims in contemporary reproductive ethics and one of the most widely rejected. Its practical bite has grown as [[embryo-selection]] has moved from testing for single-gene conditions to scoring embryos on statistical predictors of common traits.

## What the principle says

Three features are load-bearing and routinely misread.

It is a principle about **selection**, not modification. It says nothing about [[germline-editing]] and applies wherever a choice is made among possible children — which embryo to transfer, which gamete donor to use, and in a weaker sense when to conceive.

It generates a **reason, not an enforceable duty**. Savulescu is explicit that the obligation is prima facie and that it does not license coercion, state direction, or the withdrawal of reproductive liberty. On his account a couple who decline to test, or who transfer an embryo with a lower expected wellbeing, act against a reason they have; they do not thereby become subject to interference.

It ranges over **wellbeing, not health**. The principle is deliberately not restricted to disease, because Savulescu regards the therapy line criticised in [[bioethics-of-enhancement]] as arbitrary. If a trait makes a life go better, it counts, which places the principle squarely inside the argument about [[human-enhancement]] rather than beside it. In later work with Guy Kahane the formulation was revised to the child with the best chance of the best life, acknowledging that prediction is probabilistic.[^sk2009]

> [!note] The relation to eugenics
> Historical eugenics was state-directed, coercive, and aimed at population-level genetic quality. Procreative beneficence is individual, non-coercive, and aimed at the welfare of a particular child. Critics accept the distinction and argue that the aggregate of many uncoordinated private choices can produce the population-level effects the state was condemned for pursuing, without anyone having intended them.

## The non-identity problem

The principle's philosophical foundation, and its central vulnerability, is the non-identity problem set out by Derek Parfit.[^parfit1984] Choosing embryo A over embryo B does not make anyone better or worse off, because the alternative for A's life is not a worse life but no life. Nobody is harmed by not being selected, and the child who results cannot complain of a choice without which they would not exist.

This blocks the obvious person-affecting justification for selection and forces the principle onto impersonal ground: the reason to select the better-off child is that it is better for the world to contain a better life, not better for anyone. Critics regard this as the weak point. Rebecca Bennett has argued that once the impersonal move is made, procreative beneficence loses its intuitive force — the claim that one has moral reason to produce a somewhat happier person, rather than a differently constituted equally real person, is not obviously true and is not established by the intuitions the principle recruits.[^bennett2009]

The same structure creates a symmetrical difficulty for critics. The expressivist objection holds that selecting against a trait sends a demeaning message about people who have it. But if no particular person is harmed by non-selection, the objection must also be about meaning rather than harm, and both sides end up arguing about what choices express rather than about what they do. [[disability-rights-and-enhancement]] develops that objection on its own terms.

## Objections

**It presupposes a theory of the good life.** To rank possible children one needs a metric. Savulescu appeals to general-purpose means — health, intelligence, self-discipline — on the ground that they help under most conceptions of a good life. Michael Parker's reply is that reasonable parents disagree profoundly about what makes a life go well, and that a principle requiring them to select on a metric they reject is not neutral but substantive.[^parker2007]

**It over-generates.** Taken literally, the principle applies to every reproductive decision, including timing, partner choice, and whether to reproduce at all. Few of its defenders accept the implications at that scope.

**It licenses selection on socially advantageous rather than intrinsically good traits.** If a trait improves a life only because society rewards it — height, or a majority ethnic appearance — then selecting for it adapts children to injustice rather than correcting it. Savulescu accepts this consequence in some cases, which several critics treat as a reductio. Robert Sparrow has pressed the point hardest, arguing that consistently applied the principle recommends selection profiles that its proponents would not defend in public.[^sparrow2011] The same structure recurs in the popular framing of [[designer-babies]], where the traits imagined as purchasable are almost all positional.

**It sits awkwardly with the author's other commitments.** Savulescu's case for [[moral-enhancement]] rests on the claim that human dispositions are dangerously ill-suited to modern conditions. If that is right, the traits a parent should select for are not the ones that make a child's own life go best, and the principle points in a different direction from its author's other work.

**It assumes the prediction works.** This is the objection that has strengthened most since 2001.

## The prediction problem

When Savulescu wrote, selection meant testing for single-gene conditions with near-deterministic effects. The principle's modern application is [[polygenic-embryo-screening]], which scores embryos on aggregates of thousands of common variants, and the technical difficulties there constrain what the principle can actually recommend.

Polygenic scores are validated mostly by comparing unrelated people, where they capture population structure and environmental confounding along with genetic effect. Selection among siblings is a within-family comparison, in which those confounds are removed and predictive power falls substantially. Modelling work has put the expected gain from choosing the top-scoring embryo out of ten at a few centimetres of height or a few points of [[genetic-enhancement-of-intelligence|measured cognitive ability]], with wide uncertainty around each individual case, and considerably less for embryos whose ancestry is not represented in the training data.[^karavani2019] A commentary in a major medical journal set out the same limitations for clinicians.[^turley2021] Professional genetics societies in Europe and North America have advised against clinical use of polygenic embryo scoring on these grounds.

The practical ceiling is not only statistical. A stimulation cycle yields a limited number of viable embryos — often fewer than five for older patients — and the gain from selection grows only very slowly as that number rises. [[in-vitro-gametogenesis]], if it ever works in humans, would remove that constraint and is the development that would make procreative beneficence a decision of consequence rather than a philosophical exercise.

> [!debate] Small effects, large implications
> Defenders argue that a small expected benefit is still a benefit, and that the principle does not require large effects to hold. Opponents reply that a principle whose real-world application yields an uncertain fraction of a standard deviation, while restructuring how parents relate to their children, is a poor trade — and that the moral cost is incurred whether or not the prediction pays off.

## Where it stands

Procreative beneficence has not been adopted by any professional body or embedded in any regulation. UK law instead directs clinics to consider the welfare of the child, a weaker and vaguer standard that does not rank possible children against one another. Where selection is regulated at all, it is regulated by listing permitted indications rather than by any general principle about wellbeing.

Its influence has been on the shape of the debate. It forced defenders of selection-for-disease to explain why the same reasoning stops at disease, and it forced opponents of enhancement, including the writers surveyed under [[bioconservatism]], to say what is wrong with wanting a child's life to go well. It also supplied the ethical vocabulary now used by companies marketing embryo scoring directly to consumers, generally without the qualifications Savulescu attached — a commercial setting in which the [[access-and-inequality|distribution]] of the service determines who can act on the principle at all. Whether a principle can be responsible for the uses made of it in a commercial setting it did not anticipate is a question its author has addressed and his critics do not consider settled.

## See also

- [[julian-savulescu]]
- [[polygenic-embryo-screening]]
- [[embryo-selection]]
- [[designer-babies]]
- [[disability-rights-and-enhancement]]
- [[bioethics-of-enhancement]]
- [[in-vitro-gametogenesis]]
- [[moral-enhancement]]

## References

[^sav2001]: `paper` Savulescu, J. "Procreative Beneficence: Why We Should Select the Best Children." *Bioethics*, 2001.
[^sk2009]: `paper` Savulescu, J. and Kahane, G. "The Moral Obligation to Create Children with the Best Chance of the Best Life." *Bioethics*, 2009.
[^parfit1984]: `book` Parfit, D. *Reasons and Persons*. Oxford University Press, 1984.
[^bennett2009]: `paper` Bennett, R. "The Fallacy of the Principle of Procreative Beneficence." *Bioethics*, 2009.
[^parker2007]: `paper` Parker, M. "The Best Possible Child." *Journal of Medical Ethics*, 2007.
[^sparrow2011]: `paper` Sparrow, R. "A Not-So-New Eugenics: Harris and Savulescu on Human Enhancement." *Hastings Center Report*, 2011.
[^karavani2019]: `paper` Karavani, E. et al. "Screening Human Embryos for Polygenic Traits Has Limited Utility." *Cell*, 2019. {The modelled gains assume a choice among ten embryos, more than a typical stimulation cycle yields.}
[^turley2021]: `paper` Turley, P. et al. "Problems with Using Polygenic Scores to Select Embryos." *New England Journal of Medicine*, 2021. {A critique setting out the limitations of embryo polygenic scores rather than reporting a new study of them.}

==============================================================================
ARTICLE: proteostasis
TITLE:   Proteostasis collapse
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/proteostasis
SOURCE:  https://futurehumanwiki.com/raw/proteostasis
==============================================================================

---
title: "Proteostasis collapse"
slug: "proteostasis"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity"]
tags: ["aging", "protein folding", "chaperones", "proteasome", "neurodegeneration", "mechanisms"]
summary: "The age-related failure of the chaperone, proteasome and lysosomal systems that keep proteins correctly folded and turned over, and its link to aggregation diseases."
updated: "2026-07-28"
issues: ["Lecanemab and donanemab approvals are stated without a source."]
---

```infobox
{
  "caption": "Hallmark of aging",
  "rows": [
    { "label": "Tier", "value": "Primary hallmark" },
    { "label": "Term coined", "value": "2008" },
    { "label": "Coined by", "value": "Balch, Morimoto, Dillin, Kelly" },
    { "label": "Main arms", "value": "Chaperones, proteasome, lysosome" },
    { "label": "Master regulator", "value": "HSF1 (heat-shock response)" },
    { "label": "Associated diseases", "value": "Alzheimer's, Parkinson's, ALS" }
  ]
}
```

**Proteostasis collapse** is the progressive age-related failure of the systems that fold proteins correctly, refold them when they are damaged, and destroy them when they cannot be repaired. Proteostasis — protein homeostasis — is maintained by a network of several hundred chaperones, the ubiquitin–proteasome system, and lysosomal degradation, working under the control of inducible stress responses.[^balch2008] Its loss is a primary hallmark in the [[hallmarks-of-aging]] framework and the mechanistic link between aging and the major neurodegenerative diseases.

## The proteostasis network

```timeline
[
  { "year": "1962", "title": "Heat-shock response observed", "text": "Ferruccio Ritossa notices new chromosomal puffing patterns in Drosophila salivary glands after a temperature shift, the first sign of an inducible protein-protection programme." },
  { "year": "1972", "title": "Sequence determines structure", "text": "Christian Anfinsen receives the Nobel Prize in Chemistry for showing that a protein's amino acid sequence encodes its folded conformation, framing misfolding as a solvable problem." },
  { "year": "2004", "title": "Ubiquitin system recognized", "text": "Aaron Ciechanover, Avram Hershko and Irwin Rose share the Nobel Prize in Chemistry for the discovery of ubiquitin-mediated protein degradation." },
  { "year": "2008", "title": "Proteostasis named", "text": "Balch, Morimoto, Dillin and Kelly propose 'proteostasis' for the integrated network and argue it is a druggable target across folding diseases." },
  { "year": "2009–2010", "title": "Collapse mapped in a model organism", "text": "Work in C. elegans locates the loss of folding capacity in early adulthood and shows that hundreds of proteins become insoluble with age." }
]
```

A newly made polypeptide must reach one conformation out of an enormous number of possible ones, in a cytoplasm containing several hundred milligrams of protein per millilitre. Molecular chaperones — the HSP70 and HSP90 families, the small heat-shock proteins, and the chaperonin complex that folds actin and tubulin — bind exposed hydrophobic surfaces and prevent inappropriate association while folding proceeds. Only part of the network does folding work; the rest handles triage and disposal. Predicting a native structure from sequence, the task [[ai-protein-design|machine-learning models]] now perform well, is a different problem from predicting how a chain reaches that structure or how often it fails to, which is what the network exists to manage.

Two degradation systems handle proteins that cannot be rescued. The ubiquitin–proteasome system tags substrates with polyubiquitin chains through a cascade of activating, conjugating and ligase enzymes, and feeds them into the 26S proteasome, which unfolds and cleaves them. Lysosomal degradation handles bulk cargo and organelles by [[autophagy]], and handles individual proteins bearing a specific pentapeptide motif through chaperone-mediated autophagy, in which HSC70 delivers substrates to the lysosomal receptor LAMP2A for direct translocation.

Three inducible stress responses supervise the network in different compartments: the cytosolic heat-shock response, governed by the transcription factor HSF1; the unfolded protein response of the endoplasmic reticulum, running through IRE1, PERK and ATF6; and the mitochondrial unfolded protein response, which couples proteostasis to [[mitochondrial-dysfunction|mitochondrial]] state.

## What collapses, and when

The most striking result on timing came from *Caenorhabditis elegans*, where the capacity to buffer misfolded proteins falls sharply in early adulthood rather than degrading gradually across life. Morimoto's group showed that this collapse coincides with reproductive maturity and precedes any visible aging phenotype, suggesting proteostasis capacity is actively downregulated once reproduction is under way rather than simply worn out.[^benzvi2009] Long-lived insulin/IGF-1 signalling mutants of the kind [[cynthia-kenyon|Cynthia Kenyon's]] laboratory characterized delay that collapse, and the FOXO-family transcription factor their longevity depends on drives expression of chaperones and other network components. A companion finding is that hundreds of proteins become detergent-insoluble in aging worms, well beyond the handful associated with named aggregation diseases.[^david2010]

In mammals the picture is less crisp but consistent in direction. HSF1-driven induction of chaperones in response to heat or oxidative stress weakens with age, and that response is the mechanism usually invoked for [[heat-and-cold-exposure|sauna bathing]], where the human mortality data are observational. Proteasome activity declines in several tissues, though not uniformly. Chaperone-mediated autophagy declines markedly with age in rodent liver, and restoring LAMP2A expression in old mice preserved proteolytic activity and hepatic function.[^zhang2008] That experiment is one of the cleaner demonstrations that the decline is not merely correlated with aging but contributes to organ dysfunction.

> [!note] Terminology
> "Collapse" describes the abrupt loss of buffering capacity seen in invertebrate models. In mammals the change is gradual and tissue-specific, and the word is used loosely. Neither usage implies that all proteostasis functions fail together.

## Aggregation and neurodegeneration

The clinical face of proteostasis failure is protein aggregation in the nervous system. Amyloid-β and tau in [[alzheimers-disease|Alzheimer's disease]], α-synuclein in Parkinson's disease and Lewy body dementia, TDP-43 in most amyotrophic lateral sclerosis and frontotemporal dementia, and polyglutamine-expanded huntingtin in Huntington's disease all form ordered assemblies that spread between cells in a templated, prion-like manner. Age is by a wide margin the largest risk factor for all of them, which is the argument for treating them as manifestations of an aging process rather than as independent diseases — the core claim of the [[geroscience-hypothesis|geroscience hypothesis]].

Whether the aggregates themselves are the toxic species is unsettled. Soluble oligomers correlate better with neuronal dysfunction than mature fibrils in several systems, and inclusion body formation may sequester damaging material rather than cause damage. The therapeutic record is sobering: decades of amyloid-clearing programmes produced two anti-amyloid antibodies, lecanemab and donanemab, approved in the United States in the 2020s. Both slow cognitive decline modestly, at the cost of amyloid-related imaging abnormalities including brain swelling and microhaemorrhage. The size and clinical meaning of that benefit remain contested among neurologists.

## Long-lived proteins

Not all proteins are replaced. Crystallins in the eye lens are synthesized before birth and never turned over, which is why they accumulate racemized and glycated residues across a lifetime and eventually scatter light as cataract. Nuclear pore complex scaffold proteins, some histones, and elastin and collagen in the extracellular matrix persist for years to decades.[^toyama2013] For this fraction of the proteome, no amount of chaperone capacity helps; the damage is chemical, spontaneous, and irreversible without replacement of the whole structure.

This category is a distinctive problem for aging biology because it is not a regulatory failure that can be corrected by restoring a signalling pathway. Advanced glycation end-products cross-link long-lived extracellular matrix proteins and stiffen arteries and skin; breaking those cross-links pharmacologically has been attempted and has not produced a durable clinical result.

## Interventions and evidence

Interventions divide into three groups. Raising folding capacity: HSF1 activators and heat-shock protein co-inducers, of which arimoclomol is the best-known example — it failed a phase 3 trial in ALS before being approved in the United States for a rare lysosomal storage disease. Raising degradation capacity: proteasome activation extends lifespan under proteotoxic stress in nematodes, and enhancing autophagy with [[rapamycin]], spermidine or [[caloric-restriction|dietary restriction]] improves outcomes in mouse models of aggregation disease. Reducing the load: antisense oligonucleotides and RNA interference that lower production of the aggregating protein, an approach now in clinical use for transthyretin amyloidosis and in trials for Huntington's disease.

Comparative biology offers a suggestive parallel. Naked mole-rats maintain higher proteasome activity and more accurate translation than mice, and their proteins are more resistant to unfolding — one of several traits invoked in explanations of [[negligible-senescence|negligible senescence]].[^perez2009nmr] Human embryonic and [[induced-pluripotent-stem-cells|induced pluripotent stem cells]] likewise sustain unusually high proteasome activity, which falls on differentiation. That observation connects proteostasis to [[stem-cell-exhaustion|stem cell decline]] and to [[epigenetic-reprogramming|reprogramming]], since restoring a more pluripotent-like state also restores proteostatic capacity in culture.

## Open problems

There is no validated way to measure proteostatic capacity in a living human. Chaperone transcript levels, proteasome activity in blood cells, and circulating aggregate species have all been proposed and none has been qualified as an [[aging-biomarkers|aging biomarker]]. Without such a measure, the hypothesis that boosting proteostasis would slow aging cannot be tested except through disease endpoints in populations already sick.

The direction of causality between proteostasis and the other hallmarks is also unresolved. Failing proteostasis produces damaged proteins that trigger [[cellular-senescence|senescence]] and [[inflammaging|inflammatory signalling]]; senescent cells and inflammation in turn impair chaperone induction. Nor is more degradation obviously better: excessive autophagy and excessive proteasome activity both cause pathology, and the network's set points appear to be tuned rather than maximized. Whether an intervention could raise capacity in old tissue without disturbing that tuning is untested in any mammal.

## See also

- [[hallmarks-of-aging]]
- [[autophagy]]
- [[cellular-senescence]]
- [[mitochondrial-dysfunction]]
- [[geroscience-hypothesis]]
- [[negligible-senescence]]

## References

[^balch2008]: `paper` Balch, W. E., Morimoto, R. I., Dillin, A., Kelly, J. W. "Adapting proteostasis for disease intervention." *Science*, 2008.
[^benzvi2009]: `paper` Ben-Zvi, A., Miller, E. A., Morimoto, R. I. "Collapse of proteostasis represents an early molecular event in Caenorhabditis elegans aging." *PNAS*, 2009. {The abrupt collapse is a nematode finding at reproductive maturity; mammalian decline is gradual and tissue by tissue.}
[^david2010]: `paper` David, D. C. et al. "Widespread protein aggregation as an inherent part of aging in C. elegans." *PLoS Biology*, 2010.
[^zhang2008]: `paper` Zhang, C., Cuervo, A. M. "Restoration of chaperone-mediated autophagy in aging liver improves cellular maintenance and hepatic function." *Nature Medicine*, 2008. {A mouse experiment restoring one lysosomal receptor in aged liver; nothing comparable has been attempted in humans.}
[^toyama2013]: `paper` Toyama, B. H. et al. "Identification of long-lived proteins reveals exceptional stability of essential cellular structures." *Cell*, 2013.
[^perez2009nmr]: `paper` Pérez, V. I. et al. "Protein stability and resistance to oxidative stress are determinants of longevity in the longest-living rodent, the naked mole-rat." *PNAS*, 2009. {A comparison of naked mole-rat and mouse tissue; the traits are associated with longevity rather than shown to cause it.}

==============================================================================
ARTICLE: psychedelic-therapy
TITLE:   Psychedelic therapy
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/psychedelic-therapy
SOURCE:  https://futurehumanwiki.com/raw/psychedelic-therapy
==============================================================================

---
title: "Psychedelic therapy"
slug: "psychedelic-therapy"
type: "intervention"
status: "contested"
horizon: "late 2020s"
trl: 7
categories: ["minds", "enhancement"]
tags: ["psilocybin", "mdma", "depression", "ptsd", "clinical trials", "blinding", "psychiatry"]
summary: "The administration of serotonergic psychedelics or MDMA alongside psychological support for psychiatric conditions, in trials that cannot be reliably blinded."
updated: "2026-07-27"
humanEvidence: "Randomized trials in human patients report large short-term reductions in depression and PTSD symptoms, but participants and therapists almost always guess the assignment correctly, so drug and expectancy effects remain confounded."
access: "Esketamine is approved for treatment-resistant depression; psilocybin and MDMA remain unapproved medicines, reachable through trials or Australia's authorised-prescriber pathway, and psilocybin is sold at licensed non-medical centres in two US states."
reversibility: "context"
issues: ["Phase 3 psilocybin results rest partly on company announcements and may be superseded", "Colorado licensing timeline needs a primary source"]
---

```infobox
{
  "caption": "Drug-plus-psychotherapy intervention",
  "rows": [
    { "label": "Form", "value": "Supervised dosing plus preparation and integration" },
    { "label": "Main compounds", "value": "Psilocybin, MDMA, LSD, ketamine" },
    { "label": "Classical target", "value": "5-HT2A receptor" },
    { "label": "Lead indications", "value": "Depression, PTSD, addiction" },
    { "label": "Approved for depression", "value": "Esketamine only" },
    { "label": "FDA action on MDMA", "value": "Declined, August 2024" },
    { "label": "Central problem", "value": "Functional unblinding" },
    { "label": "Readiness", "value": "TRL 7" }
  ]
}
```

**Psychedelic therapy** is the administration of a psychoactive drug, most often psilocybin or MDMA, in a small number of supervised sessions framed by preparatory meetings and follow-up discussion with trained personnel. What distinguishes it from ordinary psychopharmacology is the claim that one or two sessions, rather than daily medication, can produce durable change in depression, post-traumatic stress disorder, or addiction. Randomized trials in patient populations have reported large short-term effects. Whether those effects belong to the drugs is the question the field has not resolved, because almost everyone involved can tell who received what.

## How it works

Classical psychedelics act as agonists or partial agonists at the serotonin 5-HT2A receptor; in human volunteers, pretreatment with a 5-HT2A antagonist blocks most of the subjective effects, the clearest evidence that the experience depends on this receptor. Neuroimaging during the acute state finds increased diversity of spontaneous cortical signals and reduced segregation between large-scale networks, which has made these drugs a tool in the study of [[neural-correlates-of-consciousness]] apart from any therapeutic use. In rodents and cultured neurons, psychedelics and ketamine promote dendritic spine growth and synapse formation in cortex.[^ly2018] Nothing equivalent has been measured in a living human brain.

The rationale is that the acute state opens a window in which entrenched patterns of thought become revisable, and that the psychological support around the session determines what fills it. Preparation, a dosing session of several hours with two attendants, and later integration meetings all count as part of the intervention rather than as scaffolding around a pill. The pharmacology clears within a day; the psychological content does not, which is both the intended effect and the source of the main harms.

> [!note] Terminology
> The phrase covers three unrelated pharmacologies. Psilocybin and LSD are 5-HT2A agonists. MDMA is an entactogen, a monoamine releaser that heightens feelings of closeness and blunts fear responses without much perceptual change; calling it a psychedelic is a convenience of practice, not a claim about mechanism. Ketamine is a dissociative NMDA-receptor antagonist whose antidepressant effect is rapid and short-lived.

## Development history

```timeline
[
  { "year": "1938–1943", "title": "LSD synthesised, then discovered", "text": "Albert Hofmann makes LSD at Sandoz in 1938 and identifies its psychoactivity five years later; the company distributes it to psychiatric researchers." },
  { "year": "1958", "title": "Psilocybin isolated", "text": "Hofmann isolates the active compound from Psilocybe mushrooms; Sandoz markets it to researchers as Indocybin." },
  { "year": "1970", "title": "Schedule I in the United States", "text": "The Controlled Substances Act places LSD and psilocybin in the most restrictive category, and clinical work in the West effectively stops for two decades." },
  { "year": "1985", "title": "MDMA scheduled", "text": "The DEA places MDMA in Schedule I on an emergency basis, over objections from psychotherapists who had been using it; the placement is challenged in court and later made permanent." },
  { "year": "2006", "title": "The revival's founding study", "text": "Griffiths and colleagues at Johns Hopkins report that psilocybin reliably occasions experiences volunteers rate as personally meaningful, re-establishing the drug as a legitimate subject of study." },
  { "year": "2016", "title": "Distress in cancer patients", "text": "Randomized crossover trials at Johns Hopkins and NYU report substantial reductions in depression and anxiety among patients with life-threatening cancer." },
  { "year": "2019", "title": "Esketamine approved", "text": "The FDA approves esketamine nasal spray for treatment-resistant depression, restricted to certified clinics under a risk-management programme." },
  { "year": "2021", "title": "A comparator and a phase 3", "text": "Psilocybin fails to beat escitalopram on the primary outcome of a head-to-head depression trial, while MAPS reports a positive phase 3 of MDMA-assisted therapy for PTSD." },
  { "year": "2023", "title": "Australia opens a pathway", "text": "The Therapeutic Goods Administration permits authorised psychiatrists to prescribe MDMA for PTSD and psilocybin for treatment-resistant depression, the first such national framework." },
  { "year": "2024", "title": "The FDA declines", "text": "An advisory committee votes against MDMA-assisted therapy on both effectiveness and benefit–risk, and the agency asks Lykos Therapeutics, the renamed MAPS corporate arm, for an additional phase 3 trial." }
]
```

The first clinical wave, from the 1950s to about 1970, produced a large literature and little that survives scrutiny: open-label case series, token controls, outcomes chosen after the fact. Prohibition ended it before the methodology improved. The revival was built to avoid a repeat, which makes its central design flaw more striking, not less.

## Clinical evidence

The strongest results come from narrow, severely ill populations. In patients with life-threatening cancer, single psilocybin sessions with psychological support produced large reductions in depression and anxiety lasting months.[^griffiths2016] In treatment-resistant depression, a company-sponsored phase 2b trial of the synthetic formulation COMP360 found a dose-related reduction in depression scores at three weeks, with the advantage over the lowest comparator dose narrowing by week twelve; serious adverse events, including suicidal ideation and self-injurious behaviour, occurred and were not confined to one arm.[^goodwin2022] Compass Pathways has since taken COMP360 into a phase 3 programme in the same indication and announced topline results the company describes as positive. No psilocybin product had been approved by any major regulator as of mid-2026.

The one head-to-head test against a standard antidepressant is instructive. Psilocybin did not outperform escitalopram on the primary outcome; several secondary measures favoured it, but those were not corrected for multiple comparisons and the authors said so.[^carhart2021] For PTSD, MAPS reported a phase 3 trial in which MDMA-assisted therapy reduced clinician-rated symptom severity more than placebo delivered with the same psychological support.[^mitchell2021]

Ketamine is the exception that clarifies the rest. Its antidepressant effect appears within hours and is well replicated, and esketamine, marketed as Spravato, is approved in the United States and elsewhere for treatment-resistant depression, the only member of this loose family to clear a regulator. Even here expectancy is hard to exclude: masked by general anaesthesia during surgery, ketamine did not separate from placebo, and both groups improved markedly.[^lii2023]

## The blinding problem

A person given an active dose knows it within an hour, and so does the therapist sitting with them. Guess rates approach ceiling for both parties, which means the double-blind design that licenses causal inference in drug trials is not actually operating.[^muthu2021] Inert placebos are transparent; active comparators such as niacin or a very low dose of the same drug reduce the problem without removing it.

The confound compounds. Participants are recruited from populations enthusiastic about psychedelics, arrive with strong expectations, and are assessed on scales that depend on their own report or a clinician's judgement rather than on anything measurable. The therapists are unblinded and typically advocates. Sham-controlled trials of [[non-invasive-neuromodulation]] and [[deep-brain-stimulation]] face a milder version of the difficulty, but no sham resembles several hours of altered consciousness.

> [!caution] What the FDA actually said
> The 2024 rejection was not a safety refusal. The advisory committee's objections centred on functional unblinding, on the impossibility of separating the drug from an unstandardized psychotherapy the agency does not regulate, and on missing data about durability and abuse potential.[^lykos2024] In the same month the journal *Psychopharmacology* retracted three papers reporting earlier MDMA trial data, citing undisclosed unethical conduct at one site.

The FDA's 2023 draft guidance asks sponsors to measure expectancy and characterize the psychotherapy component, and stops short of a remedy, because none is known.[^fda2023]

## Regulation and access

Two regulatory tracks are routinely conflated. The medical track runs through drug regulators, and on it only esketamine has succeeded; the agency later approved its use without a concurrent oral antidepressant. Australia's Therapeutic Goods Administration opened a narrower door in 2023, allowing specifically authorised psychiatrists to prescribe MDMA for PTSD and psilocybin for treatment-resistant depression.

The second track is not medical. Oregon's Measure 109, passed in 2020, created a state-licensed psilocybin services programme, operating since 2023: adults may consume psilocybin with a trained facilitator at a licensed centre, with no diagnosis, no prescription, and no claim of treating anything. Colorado's Proposition 122 established a comparable healing-centre framework. Neither is an approval, neither generates controlled outcome data, and calling them legalized psychedelic medicine misstates what voters passed.

Cost constrains the medical track: the active ingredient is cheap, a full day of two clinicians' time is not, which places the intervention inside the problems catalogued in [[access-and-inequality]]. Scheduling from the first wave remains in force in most jurisdictions, an application of the [[precautionary-principle]] critics blame for decades of lost evidence.

## Risks

Serious harms exist and are concentrated in the people trials exclude. A personal or family history of psychotic or bipolar disorder is a standard exclusion criterion, so little is known about giving these drugs to the population most plausibly at risk of a precipitated psychotic episode. Blood pressure and heart rate rise transiently, which matters for MDMA in anyone with cardiac disease. Prolonged psychological difficulty after a session is reported in surveys of non-clinical users at rates that are not negligible.

A distinct hazard follows from the mechanism. A drug that heightens suggestibility, trust, and emotional openness leaves the person receiving it unusually dependent for several hours, which is why documented boundary violations at a trial site were read as structural rather than isolated. The [[bioethics-of-enhancement|ethical questions]] here concern the therapist relationship at least as much as the drug.

## Healthy people and enhancement

The evidence base is entirely clinical. Claims that psychedelics improve creativity, wellbeing, or moral sensibility in people without a diagnosis are not supported by controlled work. The healthy-volunteer literature is small, its samples are rarely representative, and its outcomes are overwhelmingly self-rated questionnaires administered to people who know what they took. Microdosing has been tested more rigorously than much of the field, in a self-blinding citizen-science study whose participants randomized their own capsules: the reported benefits appeared in the placebo condition too.[^szigeti2021] Subjective improvement outruns measured change, the pattern the literature on [[nootropics]] keeps finding and the one [[quantified-self|self-tracking]] cannot correct for.

Advocates within [[human-enhancement]] circles nonetheless treat these compounds as candidate tools, framing supervised access as an instance of [[morphological-freedom]] and prosocial effects as a route to [[moral-enhancement]]. Bioconservative critics answer in the terms [[leon-kass]] set out against pharmacological mood improvement: a chemically induced sense of meaning is not meaning. Ego dissolution draws philosophical interest for what it might show about [[personal-identity-and-continuity]], though a first-person report of self-loss is evidence about the report. The unregulated market, overlapping with [[biohacking]] and with mushroom products sold under the post-market rules described in [[dietary-supplements]], has no outcome data.

## Outlook

Three things would move the field. The first is the additional MDMA phase 3 the FDA requested, together with the psilocybin phase 3 programme, read for effect sizes rather than headlines. The second is a design that measures the drug's contribution instead of assuming it: prespecified expectancy assessment, blinded independent raters, dose–response comparisons among active arms. The third is pharmacological. Analogues engineered to keep the plasticity-promoting effects without the subjective experience have shown antidepressant-like effects in rodents; one that worked in humans would settle whether the experience is necessary and restore the blind at once.

> [!debate] Is the trip required
> One camp holds that the acute experience is the therapy, that its intensity predicts outcome, and that a non-hallucinogenic analogue would be a different drug. The other holds that the experience is a side effect of engaging a plasticity mechanism, and that the intensity–outcome correlation is what expectancy alone would produce. No human trial has separated them.

The strongest objection to the field is not that its drugs are dangerous or its trials fraudulent. It is that after two decades of revival, no study has distinguished the effect of the drug from the effect of believing one received it.

## See also

- [[nootropics]]
- [[non-invasive-neuromodulation]]
- [[deep-brain-stimulation]]
- [[neural-correlates-of-consciousness]]
- [[human-enhancement]]
- [[moral-enhancement]]
- [[bioethics-of-enhancement]]
- [[access-and-inequality]]

## References

[^ly2018]: `paper` Ly, C. et al. "Psychedelics Promote Structural and Functional Neural Plasticity." *Cell Reports*, 2018. {The spine-growth and synaptogenesis results are from rodent cortex and cultured neurons, not from people.}
[^griffiths2016]: `paper` Griffiths, R.R. et al. "Psilocybin produces substantial and sustained decreases in depression and anxiety in patients with life-threatening cancer: a randomized double-blind trial." *Journal of Psychopharmacology*, 2016.
[^goodwin2022]: `paper` Goodwin, G.M. et al. "Single-Dose Psilocybin for a Treatment-Resistant Episode of Major Depression." *New England Journal of Medicine*, 2022. {The comparator was a lower dose of the same drug rather than an inert placebo, which helps blinding and complicates interpretation.}
[^carhart2021]: `paper` Carhart-Harris, R. et al. "Trial of Psilocybin versus Escitalopram for Depression." *New England Journal of Medicine*, 2021. {Widely reported as a win for psilocybin; the primary outcome showed no significant difference between the two drugs.}
[^mitchell2021]: `paper` Mitchell, J.M. et al. "MDMA-assisted therapy for severe PTSD: a randomized, double-blind, placebo-controlled phase 3 study." *Nature Medicine*, 2021.
[^lii2023]: `paper` Lii, T.R. et al. "Randomized trial of ketamine masked by surgical anesthesia in patients with depression." *Nature Mental Health*, 2023. {Masking an infusion under general anaesthesia is the closest anyone has come to a genuinely blinded trial in this family of drugs.}
[^muthu2021]: `paper` Muthukumaraswamy, S.D., Forsyth, A. and Lumley, T. "Blinding and expectancy confounds in psychedelic randomised controlled trials." *Expert Review of Clinical Pharmacology*, 2021.
[^lykos2024]: `statement` Lykos Therapeutics. "Lykos Therapeutics Announces Complete Response Letter for Midomafetamine Capsules for PTSD." Company announcement, 2024. {The FDA does not publish complete response letters, so the request for another phase 3 trial is known from the sponsor's own account.}
[^fda2023]: `regulator` U.S. Food and Drug Administration. *Psychedelic Drugs: Considerations for Clinical Investigations*. Draft guidance for industry, 2023.
[^szigeti2021]: `paper` Szigeti, B. et al. "Self-blinding citizen science to explore psychedelic microdosing." *eLife*, 2021.

==============================================================================
ARTICLE: quantified-self
TITLE:   Quantified self
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/quantified-self
SOURCE:  https://futurehumanwiki.com/raw/quantified-self
==============================================================================

---
title: "Quantified self"
slug: "quantified-self"
type: "concept"
status: "established"
horizon: "present"
categories: ["enhancement", "society"]
tags: ["self-tracking", "measurement", "n-of-1", "privacy", "wearables", "citizen science"]
summary: "A movement and a method for learning about oneself through personal measurement, built on n-of-1 self-experiment and named by two Wired editors in 2007."
updated: "2026-07-27"
humanEvidence: "Self-monitoring raises measured activity in people over the short term; a randomised trial of a workplace wellness programme found more self-reported exercise but no change in clinical measures or spending at 18 months."
access: "Trackers ship in most phones and cost from tens to a few hundred dollars, but much tracking now arrives through an employer or insurer programme rather than a purchase."
reversibility: "context"
issues: ["Origins section cites no independent source for the meetup count or the final conference year."]
---

```infobox
{
  "caption": "Movement and method in self-tracking",
  "rows": [
    { "label": "Term coined", "value": "2007" },
    { "label": "Coined by", "value": "Gary Wolf, Kevin Kelly" },
    { "label": "Motto", "value": "Self-knowledge through numbers" },
    { "label": "Core method", "value": "n-of-1 self-experiment" },
    { "label": "Meetup format", "value": "Show-and-tell, three questions" },
    { "label": "Conference series", "value": "2011–2018" },
    { "label": "Successor term", "value": "Personal science" },
    { "label": "Adjacent practice", "value": "Biohacking", "link": "/wiki/biohacking" }
  ]
}
```

**Quantified self** is the practice of learning about oneself by measuring oneself, and the name of the movement that formed around that practice after two *Wired* editors, Gary Wolf and Kevin Kelly, coined the phrase in 2007. Its slogan was "self-knowledge through numbers", its characteristic artefact was a talk given at a meetup about a personal experiment, and its intellectual core was the n-of-1 trial. Nearly two decades on, self-tracking is a default feature of consumer devices and the movement that argued for it has largely dissolved.

```keyfacts
[
  { "value": "2007", "label": "Term coined", "note": "Gary Wolf and Kevin Kelly, at Wired" },
  { "value": "3", "label": "Questions in the meetup format", "note": "what did you do, how did you do it, what did you learn" },
  { "value": "1986", "label": "n-of-1 trial method formalised", "note": "Guyatt and colleagues, in the New England Journal of Medicine" }
]
```

## Origins

The label arrived at the moment the instruments did. Cheap accelerometers, phones with sensors, and the first consumer activity trackers made it possible to keep a continuous record of sleep, movement, weight or mood without a laboratory. Wolf and Kelly started the Quantified Self blog in 2007; the first show-and-tell meetup followed in 2008, held at Kelly's studio in Pacifica, California, and attended by about thirty people. Wolf's 2010 essay in *The New York Times Magazine* and a TED talk the same year carried the idea to a general audience.[^wolf2010]

The meetup format did more of the intellectual work than the name suggests. A speaker answered three questions in order: what did you do, how did you do it, and what did you learn. The rule excluded product pitches and forced any presenter to state a method and an outcome, including a null one. Subjects ranged from diet and sleep through [[heat-and-cold-exposure|cold exposure]] to mood and productivity. Meetups spread to dozens of cities, and a conference series ran from 2011 until 2018, when the last global meeting was held in Portland, Oregon.

```timeline
[
  { "year": "2007", "title": "The phrase and the hardware", "text": "Gary Wolf and Kevin Kelly begin the Quantified Self blog; the first consumer activity-tracker companies are founded around the same time." },
  { "year": "2008", "title": "First show-and-tell", "text": "About thirty people attend the first Quantified Self meetup, at Kevin Kelly's studio in Pacifica, California, establishing the format the movement kept." },
  { "year": "2010", "title": "Wider audience", "text": "Wolf's essay in The New York Times Magazine and a TED talk carry the label beyond the Bay Area technology community." },
  { "year": "2011", "title": "First conference", "text": "The first Quantified Self conference is held in Mountain View, California in May, followed by a European meeting in Amsterdam in November; the show-and-tell format is retained at scale." },
  { "year": "2015", "title": "Tracking becomes a default", "text": "The Apple Watch ships and Apple releases ResearchKit, putting continuous measurement and study enrolment into a mass-market device rather than a hobbyist one." },
  { "year": "2018", "title": "The costs surface", "text": "Strava's global activity heatmap is found to expose the layout of overseas military sites, and John Hancock announces it will sell only interactive life insurance, attaching a tracking-based wellness programme to new policies." },
  { "year": "2019", "title": "Null results at scale", "text": "Two randomised evaluations of workplace wellness programmes report little or no effect on clinical measures, medical spending or employment outcomes." },
  { "year": "2020", "title": "Personal science", "text": "Wolf and Martijn de Groot publish a framework recasting the practice as personal science, an attempt to define the method independently of the movement." },
  { "year": "2025", "title": "Data as an asset", "text": "23andMe enters Chapter 11 bankruptcy with its customer genetic database among the assets, and state attorneys general advise customers to delete their data." }
]
```

## The method

The movement's strongest claim is methodological, not technological. An n-of-1 trial treats a single person as the whole study population, alternating between conditions across multiple periods, ideally randomised, blinded and separated by washout intervals long enough for a previous treatment to clear. It is a recognised design in clinical research, formalised in the 1980s for chronic stable conditions where a patient's own response is the question of interest.[^guyatt1986] Evidence-based-medicine frameworks rank a well-conducted randomised n-of-1 trial at or near the top for one specific purpose: deciding whether a treatment helps this patient.

The premise behind it has independent support. A study of postprandial glucose in a large cohort found that responses to the same meal varied widely between individuals and were partly predictable from personal features, which undercuts the assumption that a population-average dietary recommendation fits everyone in the population.[^zeevi2015] A later and larger study reported the same between-person spread alongside substantial variation within one person eating the same food on different days, which limits how much any single measured response establishes.[^berry2020] That is the best scientific case the movement's founding intuition has, and it is the same argument that motivates [[continuous-glucose-monitoring]] outside diabetes and [[human-digital-twins|individualised physiological models]].

> [!note] A self-experiment is not an n-of-1 trial
> The clinical design gets its strength from randomisation, blinding, repeated crossovers and pre-specified endpoints. A self-experiment that changes one thing for a month and compares it with the month before shares the name and almost none of the machinery. That gap is where most disagreement about the movement's rigour sits.

## Where the method breaks

Four failure modes account for most self-tracking conclusions that do not survive a controlled study.

**Regression to the mean.** People start tracking when something is unusually bad: sleep at its worst, weight at its highest, mood at its lowest. The next measurement tends back toward the personal average whether or not anything was done, and an intervention begun at the trough inherits credit for the return.

**Expectancy.** The person running the experiment is the person hoping it works, and is also the instrument reading the outcome. Where the endpoint is self-reported — energy, focus, sleep quality — this is close to unmanageable. Placebo research has found reported symptom improvement even when participants were told they were receiving an inert pill.

**Seasonality.** Activity, weight, sleep duration, mood and vitamin D all vary across the year. An experiment run from February to April is confounded with spring.

**Multiplicity.** Someone logging forty variables and looking for relationships among them will find several at conventional thresholds by chance alone. Self-trackers rarely pre-register a hypothesis, and the exploratory search is usually reported as though it had been a test.

A fifth problem is peculiar to the practice: measurement changes the thing measured. Wearing a step counter increases steps. That reactivity is how tracking produces behaviour change in the short term, and also why a tracked baseline is not a baseline; whether the change survives beyond the short term is a separate question, and the randomised evidence on that, summarised under [[wearable-health-sensors]], is weak.

## The instruments

Consumer sensors are better at some quantities than at others, and nothing about a displayed number communicates its error bar. Wrist-worn optical heart-rate measurement is reasonably accurate at rest and during steady activity, while energy-expenditure estimates from the same devices have shown large errors against laboratory reference methods. In one laboratory comparison, none of the seven wrist devices tested estimated energy expenditure within an acceptable margin.[^shcherbina2017] Sleep trackers separate sleep from wake tolerably and stage it poorly against polysomnography, which matters for anyone drawing conclusions about deep sleep from a ring; the evidence on sleep and health is set out in [[sleep-and-longevity]]. Reporting sleep efficiency to the nearest percentage point claims a precision the sensor cannot support, a problem shared with consumer [[epigenetic-clock|methylation age tests]] and with the difficulty of validating [[aging-biomarkers|biomarkers of aging]] generally. Device-level detail is treated in [[wearable-health-sensors]].

Occasionally consumer scale does what a clinical study cannot. In a smartwatch study enrolling roughly 419,000 US participants, about 0.5% received an irregular-pulse notification, and around a third of those who then returned a usable ECG patch recording had atrial fibrillation confirmed.[^perez2019] The result cuts both ways: consumer hardware detected real disease, and most notifications went unconfirmed, though a patch mailed days after the alert cannot rule out an intermittent rhythm. The trial evidence is set out under [[wearable-health-sensors]].

## From choice to default

The movement's odd fate is that it won and then disappeared. Tracking is no longer something a person decides to take up. Phones count steps without being asked, watches record heart rate continuously, US regulators cleared the first over-the-counter glucose sensor in 2024 for adults who do not use insulin, with others following within the year, and [[consumer-blood-testing|direct-to-consumer blood panels]] and [[biological-age]] reports are sold as retail products. The hobbyist device makers of the late 2000s were acquired or liquidated, and their function moved into general-purpose hardware.

Default tracking is a different thing from voluntary self-experiment, and the difference is who holds the data and who set the goal. Deborah Lupton's sociology of the practice distinguishes tracking that is private, communal, pushed, imposed and exploited, the last two covering employer wellness programmes, insurance schemes, and the resale of behavioural data by parties the tracked person never dealt with.[^lupton2016] A step count logged to answer a personal question and the same count logged to qualify for a premium discount are not the same measurement.

> [!debate] Whether tracking improves outcomes
> Two large randomised evaluations of workplace wellness programmes converged on the same answer. In a trial at a US retailer, employees offered the programme reported more regular exercise and more active weight management at 18 months, while clinical measures, health spending and employment outcomes did not differ significantly from controls.[^song2019] The Illinois study raised screening rates but found no significant effect on medical spending, other health behaviours or productivity, and it added a result that complicates every observational claim here: employees who chose to enrol were already healthier and lower-spending before the programme began, so such schemes look effective in uncontrolled data largely through selection.[^jones2019] Advocates answer that the interventions tested were weak, which is a fair point about those trials and not an argument that a stronger one works.

## Data, ownership and privacy

Health data from a consumer device generally falls outside the medical privacy rules people assume protect it: in the United States, HIPAA binds covered entities such as clinicians and insurers, not fitness apps. The consequences have been demonstrated rather than hypothesised. Strava's published global heatmap of user activity was found in 2018 to reveal the layout and patrol routes of overseas military installations, an aggregate disclosure no individual user consented to or could have foreseen.[^hern2018] Location histories held by period-tracking and fitness apps became a live legal concern in the US after 2022. And when 23andMe entered Chapter 11 bankruptcy in 2025, its customer genetic database was treated as an asset in the proceedings, which shows that data-ownership terms survive only as long as the company does.

This is where the reversibility of self-tracking splits. Stopping is trivial: the device goes in a drawer. Disclosure is not, because a record already given to a platform, employer, insurer or broker cannot be recalled, and inference from it improves over time. The same asymmetry drives [[genetic-discrimination]] and, for a more intimate class of signal, [[mental-privacy]].

## Criticism

The sociological objection is that quantification does not merely describe a life but reshapes what counts in it, privileging whatever a sensor happens to capture and moralising health as personal diligence. Participants were disproportionately affluent, technically skilled and already healthy, which limits what their conclusions generalise to and ties the practice to [[access-and-inequality]]. Clinicians treating eating disorders have raised specific concerns about calorie- and step-counting features, which can supply a socially approved vocabulary for restriction.

A narrower criticism concerns the yield. No self-tracking result has entered the physiology literature as an accepted finding, in contrast to the do-it-yourself insulin-delivery work described under [[biohacking]], which produced a technology and then a randomised controlled trial. Defenders answer that population knowledge was never the point, and that a method for answering a question about one person is valuable even if it generalises to nobody — a claim also made for individualised use of [[dietary-supplements]] and [[nootropics|cognitive supplements]].

## Outlook

The method has outlived the movement. Wolf and Martijn de Groot's framework recast the practice as "personal science", defining it by the process of asking and answering a personal empirical question rather than by any device.[^wolfdegroot2020] Rigorous n-of-1 designs have meanwhile found institutional homes in rare-disease therapeutics, chronic pain and precision nutrition, with denser data than the paper-diary era allowed. Whether that infrastructure serves the person wearing the sensor or the parties buying the exhaust is now a governance question rather than a technical one, and it will be settled by data-protection law and employment practice rather than by anything the [[human-enhancement|enhancement]] debate contributes. What the founders wanted — a person answering a question about their own body, with numbers, and telling others how it went — turned out to be the easy part.

## See also

- [[wearable-health-sensors]]
- [[continuous-glucose-monitoring]]
- [[biohacking]]
- [[consumer-blood-testing]]
- [[biological-age]]
- [[mental-privacy]]
- [[access-and-inequality]]
- [[healthspan]]

## References

[^wolf2010]: `news` Wolf, G. "The Data-Driven Life." *The New York Times Magazine*, 2010. {Wolf co-coined the term, so the piece is a participant's account of the movement rather than an outside assessment of it.}
[^guyatt1986]: `paper` Guyatt, G. et al. "Determining Optimal Therapy: Randomized Trials in Individual Patients." *New England Journal of Medicine*, 1986.
[^zeevi2015]: `paper` Zeevi, D. et al. "Personalized Nutrition by Prediction of Glycemic Responses." *Cell*, 2015. {Establishes between-person variation in a physiological response, not that acting on the variation improves any clinical outcome.}
[^berry2020]: `paper` Berry, S.E. et al. "Human postprandial responses to food and potential for precision nutrition." *Nature Medicine*, 2020. {Reports within-person as well as between-person variation, so a single measured response is not a stable personal trait.}
[^shcherbina2017]: `paper` Shcherbina, A. et al. "Accuracy in Wrist-Worn, Sensor-Based Measurements of Heart Rate and Energy Expenditure in a Diverse Cohort." *Journal of Personalized Medicine*, 2017.
[^perez2019]: `paper` Perez, M.V. et al. "Large-Scale Assessment of a Smartwatch to Identify Atrial Fibrillation." *New England Journal of Medicine*, 2019. {A single-arm study without a control group, so it measures detection rather than any benefit from detecting.}
[^song2019]: `paper` Song, Z. & Baicker, K. "Effect of a Workplace Wellness Program on Employee Health and Economic Outcomes: A Randomized Clinical Trial." *JAMA*, 2019.
[^jones2019]: `paper` Jones, D., Molitor, D. & Reif, J. "What Do Workplace Wellness Programs Do? Evidence from the Illinois Workplace Wellness Study." *The Quarterly Journal of Economics*, 2019.
[^hern2018]: `news` Hern, A. "Fitness tracking app Strava gives away location of secret US army bases." *The Guardian*, 2018.
[^lupton2016]: `book` Lupton, D. *The Quantified Self: A Sociology of Self-Tracking.* Polity Press, 2016.
[^wolfdegroot2020]: `paper` Wolf, G.I. & De Groot, M. "A Conceptual Framework for Personal Science." *Frontiers in Computer Science*, 2020.

==============================================================================
ARTICLE: radiation-hardening-humans
TITLE:   Radiation tolerance in humans
PORTAL:  Space & Extreme Environments
URL:     https://futurehumanwiki.com/wiki/radiation-hardening-humans
SOURCE:  https://futurehumanwiki.com/raw/radiation-hardening-humans
==============================================================================

---
title: "Radiation tolerance in humans"
slug: "radiation-hardening-humans"
type: "concept"
status: "speculative"
horizon: "2040s"
categories: ["space", "genetics"]
tags: ["radiation", "spaceflight", "dna repair", "shielding", "gene editing", "cancer risk"]
summary: "The problem of keeping humans healthy under deep-space radiation, and the shielding, pharmacological, and genetic strategies proposed to raise their tolerance."
updated: "2026-07-27"
---

```infobox
{
  "caption": "Problem in space physiology and genetic engineering",
  "rows": [
    { "label": "Principal hazard", "value": "Galactic cosmic rays" },
    { "label": "Secondary hazard", "value": "Solar particle events" },
    { "label": "Deep-space dose rate", "value": "~1.8 mSv/day" },
    { "label": "Main endpoints", "value": "Cancer, CNS, cardiovascular" },
    { "label": "Established mitigation", "value": "Mass shielding, mission duration" },
    { "label": "Genetic hardening", "value": "Cell culture only" }
  ]
}
```

**Radiation tolerance in humans** refers to the set of strategies — engineering, pharmacological, and genetic — intended to let people survive radiation exposures that current biology handles poorly, principally the charged-particle environment beyond Earth's magnetosphere. It is the constraint that most clearly separates orbital spaceflight from interplanetary travel. Unlike bone and muscle loss, which are disuse phenomena that respond to exercise, radiation damage accumulates whether or not the crew does anything, and no validated countermeasure exists.

```keyfacts
[
  { "value": "~1.8 mSv/day", "label": "Dose equivalent in deep-space cruise", "note": "measured by the RAD instrument aboard Mars Science Laboratory" },
  { "value": "600 mSv", "label": "Career limit NASA proposed", "note": "reviewed by the National Academies in 2021" },
  { "value": "0", "label": "Humans given engineered radiation tolerance", "note": "all genetic work remains in cells or animals" }
]
```

## The exposure

Two components matter, and they pose opposite engineering problems.

Galactic cosmic rays are a continuous, near-isotropic flux of nuclei accelerated outside the solar system: mostly protons, about a tenth helium, and a small fraction of heavier nuclei up to iron. That last fraction is the difficulty. High-energy heavy ions — HZE particles — carry energies in the hundreds of MeV per nucleon and deposit dense, correlated tracks of ionisation as they pass through tissue, producing clustered DNA lesions that repair machinery handles far worse than the sparse damage from X-rays. Flux varies inversely with the solar cycle, roughly doubling between solar maximum and minimum.

Solar particle events are episodic bursts, overwhelmingly protons, associated with flares and coronal mass ejections. They are unpredictable, can last hours to days, and a large event outside the magnetosphere could deliver an acutely dangerous dose. The event of August 1972, falling between two Apollo missions, is the standard illustration of the risk.

Measurements from the Mars Science Laboratory's RAD instrument put the dose equivalent in interplanetary cruise at roughly 1.8 mSv per day,[^zeitlin2013] and at the surface of Gale crater at roughly a third of that, thanks to the thin atmosphere and the bulk of the planet blocking half the sky.[^hassler2014] A conventional round trip with a surface stay therefore approaches or exceeds a sievert of accumulated dose equivalent, above the career exposure limit NASA has proposed.[^nasem2021]

> [!note] Terminology
> "Dose" in grays measures energy deposited; "dose equivalent" in sieverts weights it by how biologically damaging the particle type is. The weighting factors for HZE particles are extrapolated from limited data and are themselves a major source of uncertainty in every risk estimate quoted here.

## What the damage does

Cancer risk dominates public discussion and mission rules, but it is one of four endpoint classes NASA formally tracks. Degenerative tissue effects include cataracts, which are documented in astronauts, and cardiovascular disease, where the evidence from the small Apollo cohort is genuinely contested. Acute radiation syndrome is a concern only for an unshielded large solar event.

The fourth class, central nervous system effects, is the most uncertain. Rodents exposed to simulated GCR show deficits in attention, recognition memory, and dendritic complexity. The relevance is hard to judge: most experiments use single ion species delivered acutely at dose rates orders of magnitude above spaceflight, and results at realistic chronic low dose rates have been inconsistent. Whether a Mars crew would arrive cognitively impaired is not known and cannot be settled with existing data.

Persistent damage also feeds mechanisms this wiki covers elsewhere. Clustered double-strand breaks drive [[cellular-senescence]], and irradiated tissue shows the chronic inflammatory signature described under [[inflammaging]]. Radiation is, in this sense, an accelerator of several [[hallmarks-of-aging|hallmarks]] at once, which is part of why [[space-medicine]] treats deep-space flight as an aging model.

## Shielding and its ceiling

```compare
{
  "columns": ["Passive mass", "Active magnetic", "Biological hardening"],
  "rows": [
    { "label": "Works against solar events", "values": ["Yes, effectively", "Yes, in principle", "Partially"] },
    { "label": "Works against GCR", "values": ["Poorly; secondaries offset gains", "Only with impractical field strengths", "Unknown"] },
    { "label": "Mass or power cost", "values": ["Very high", "Very high power, cryogenics", "Negligible"] },
    { "label": "Demonstrated in humans", "values": ["Yes", "No", "No"] }
  ]
}
```

Passive shielding works well for solar protons and poorly for cosmic rays. A high-energy iron nucleus striking aluminium produces a shower of secondary neutrons and lighter fragments, so adding structural mass can leave the dose roughly unchanged or, in some geometries, worse. Hydrogen-rich materials fragment incoming nuclei more efficiently per unit mass, which is why polyethylene, water, and stored propellant, food, and waste are preferred as shielding. Even so, meaningful GCR reduction requires depths measured in tens of centimetres of water-equivalent over the whole habitable volume, which is prohibitive for a transit vehicle and straightforward for a surface base buried under regolith.

A storm shelter — a small, heavily shielded volume the crew occupies during a solar event — is the standard architecture, and is genuinely effective for that hazard alone.

Active shielding by superconducting magnets has been studied repeatedly and repeatedly found impractical: deflecting particles at the relevant rigidities requires field strengths and coil masses that dominate the vehicle, plus cryogenic power and stray-field management inside the crew volume. It remains a concept.

## Pharmacological approaches

Radioprotectors given before exposure and mitigators given after are separate categories. Amifostine, a thiol scavenger, is approved as a cytoprotectant during radiotherapy, but is administered intravenously and causes hypotension and nausea, making chronic use over a multi-year mission implausible. Granulocyte colony-stimulating factors are approved for the haematopoietic form of acute radiation syndrome; they address a scenario a crew would rather avoid than treat.

Antioxidants, dietary strategies, and [[dietary-supplements|nutraceutical]] candidates dominate the literature by volume and have not produced a countermeasure with demonstrated efficacy against chronic HZE exposure in any species. The underlying difficulty is that free-radical scavenging addresses the indirect component of damage while clustered lesions from a heavy ion track are largely direct.

More speculative pharmacology overlaps with geroscience. If radiation exposure works partly by driving senescent-cell accumulation, then [[senolytics|senolytic]] clearance is a candidate mitigator, and preclinical work in irradiated mice supports the mechanism. Nothing has been tested in flight.

## Genetic approaches

Proposals to edit humans for radiation tolerance are a serious research topic and, as of 2026, entirely preclinical.

The most-discussed candidate is Dsup, a damage-suppressor protein from the tardigrade *Ramazzottius varieornatus*. Expressed in cultured human cells, it binds chromatin and reduced X-ray-induced DNA damage by a substantial margin in the original report.[^hashimoto2016] The result is real, replicated, and much narrower than its coverage implied — see [[tardigrade-genes]] for what did and did not transfer.

A second family of proposals borrows tumour-suppressor redundancy from long-lived large animals. Elephants carry roughly twenty copies of *TP53* and show unusually low cancer incidence for their body mass, with irradiated elephant cells apoptosing more readily than human cells.[^abegglen2015] Adding *TP53* copies to humans is a poor idea on current evidence: mice engineered for hyperactive p53 showed reduced tumour incidence together with premature aging phenotypes, exactly the tradeoff the damage-response literature would predict.[^tyner2002]

Other suggestions — importing repair systems from *Deinococcus radiodurans*, engineering melanin-based shielding, boosting base-excision repair capacity — face a common problem. *Deinococcus* resistance rests substantially on manganese-antioxidant chemistry that protects proteins, not on a transferable repair enzyme, so there is no single gene to move.

Delivery is the other unsolved half. A somatic approach would need to reach every dividing tissue, which is beyond current [[aav-vectors|AAV]] and [[lipid-nanoparticles|lipid nanoparticle]] capability and would need to persist for years; see [[somatic-gene-therapy]]. A germline approach would make the change heritable and inherits every objection catalogued under [[germline-editing]] and [[crispr-off-target-effects]], with the added feature that the trait is useless on Earth and possibly harmful.

> [!debate] Engineer the crew or shorten the mission
> Faster propulsion reduces integrated dose linearly and requires no biology. Critics of biological hardening argue that any resource spent on editing humans would buy more risk reduction spent on propulsion, shielding mass, or mission architecture. Proponents reply that transit time has a floor set by orbital mechanics and that permanent settlement, unlike a mission, has no exposure endpoint at all.

## Open problems

The risk models themselves are the weakest link. Quality factors for HZE particles rest on sparse experimental data; NASA's own uncertainty analyses put wide confidence intervals on any predicted excess cancer risk, and the intervals are wide enough that a Mars mission may be within or outside limits depending on model choices. Better human data are impossible to obtain ethically and impossible to obtain in advance.

Selection rather than modification is the nearer-term possibility, and the more uncomfortable one. Screening candidates for DNA repair capacity or for variants associated with radiosensitivity is technically feasible now and raises the issues treated in [[genetic-discrimination]]. Whether a space agency should select crew on genotype is a live question that no agency has answered publicly.

For permanent off-world populations, the framing changes from mitigation to adaptation, which is the subject of [[pantropy]] and, over generations, of [[generation-ship-biology]]. Nothing in the current evidence base indicates that this is achievable, and the honest position is that radiation remains the least-solved problem in human spaceflight.

## See also

- [[space-medicine]]
- [[tardigrade-genes]]
- [[pantropy]]
- [[microgravity-adaptation]]
- [[germline-editing]]
- [[generation-ship-biology]]
- [[human-hibernation]]
- [[human-enhancement]]

## References

[^zeitlin2013]: `paper` Zeitlin, C. et al. "Measurements of Energetic Particle Radiation in Transit to Mars on the Mars Science Laboratory." *Science*, 2013.
[^hassler2014]: `paper` Hassler, D.M. et al. "Mars' Surface Radiation Environment Measured with the Mars Science Laboratory's Curiosity Rover." *Science*, 2014.
[^nasem2021]: `report` National Academies of Sciences, Engineering, and Medicine. *Space Radiation and Astronaut Health: Managing and Communicating Cancer Risks*. 2021.
[^hashimoto2016]: `paper` Hashimoto, T. et al. "Extremotolerant tardigrade genome and improved radiotolerance of human cultured cells by tardigrade-unique protein." *Nature Communications*, 2016. {Dsup was tested in cultured human cells against X-rays, not against the heavy ions that dominate deep-space dose.}
[^abegglen2015]: `paper` Abegglen, L.M. et al. "Potential Mechanisms for Cancer Resistance in Elephants and Comparative Cellular Response to DNA Damage in Humans." *JAMA*, 2015. {Compares elephant and human cells in culture alongside zoo mortality records; no extra TP53 copies have been put into a person.}
[^tyner2002]: `paper` Tyner, S.D. et al. "p53 mutant mice that display early ageing-associated phenotypes." *Nature*, 2002. {The mice carried a mutant hyperactive p53 allele rather than extra intact copies, which weakens the analogy to elephant TP53 duplication.}

==============================================================================
ARTICLE: rapamycin
TITLE:   Rapamycin
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/rapamycin
SOURCE:  https://futurehumanwiki.com/raw/rapamycin
==============================================================================

---
title: "Rapamycin"
slug: "rapamycin"
type: "intervention"
status: "experimental"
horizon: "2030s"
trl: 5
categories: ["longevity"]
tags: ["mtor", "aging", "drug repurposing", "autophagy", "immunosenescence", "clinical trials"]
summary: "A macrolide immunosuppressant that inhibits the mTOR growth pathway and extends lifespan across several laboratory species, with no confirmed effect on human aging."
updated: "2026-07-28"
humanEvidence: "No human trial has tested lifespan or healthspan; low-dose rapalogs improved influenza vaccine response in older adults, a phase 3 respiratory-illness trial failed, and every lifespan result is from mice and other laboratory species."
access: "Approved as sirolimus for transplant rejection and available as a generic; longevity use is off-label prescribing by a minority of physicians, at doses and schedules no regulator has evaluated."
reversibility: "reversible"
issues: ["PEARL trial results are described without a citation", "The 2019 RTB101 phase 3 failure is a load-bearing claim and carries no source"]
---

```infobox
{
  "caption": "Small-molecule drug",
  "rows": [
    { "label": "Drug class", "value": "Macrolide, mTOR inhibitor" },
    { "label": "Source", "value": "Streptomyces hygroscopicus" },
    { "label": "Isolated", "value": "Early 1970s, Ayerst Research" },
    { "label": "Molecular target", "value": "mTORC1 (via FKBP12)" },
    { "label": "Approved as", "value": "Sirolimus, transplant rejection" },
    { "label": "Approved for aging", "value": "No" },
    { "label": "Readiness", "value": "TRL 5" }
  ]
}
```

**Rapamycin** is a bacterial macrolide, marketed as sirolimus, that inhibits the mechanistic target
of rapamycin (mTOR) — the kinase complex through which cells decide whether conditions favour growth
or conservation. It is one of the few compounds shown to extend lifespan in yeast, nematodes, flies,
and mice, and the most reproducible positive result in the US National Institute on Aging's
[[interventions-testing-program|Interventions Testing Program]]. Whether it does anything for human aging is unknown: the human trials run so far
have tested immune and functional surrogates in small populations, and the one large randomized trial
of a related compound failed its primary endpoint.

## How it works

mTOR sits at the centre of nutrient sensing. As part of complex 1 (mTORC1) it integrates amino-acid
availability, growth-factor signalling, and cellular energy state, and when those inputs are
favourable it drives protein synthesis through S6 kinase and 4E-BP1, promotes ribosome biogenesis,
and suppresses [[autophagy]]. Rapamycin does not bind mTOR directly. It first binds the small protein
FKBP12, and the resulting complex docks onto mTORC1 and allosterically restricts its activity.

Inhibiting mTORC1 therefore shifts a cell from building to recycling. Autophagic flux rises, damaged
organelles and aggregated proteins are degraded, and the burden on the
[[proteostasis|protein quality-control system]] falls. This is the same axis that
[[caloric-restriction]] engages nutritionally, and the overlap is the main reason mTOR inhibition is
treated as a pharmacological approximation of dietary restriction rather than an independent
mechanism.

The complication is mTORC2. Acute rapamycin spares it, but chronic exposure disrupts its assembly in
some tissues, and mTORC2 inhibition is associated with insulin resistance and glucose
intolerance.[^lamming2012] Much of the current interest in intermittent dosing, and in
mTORC1-selective inhibitors built to avoid touching complex 2 at all, follows from this single
pharmacological fact.

> [!key] Why mTOR keeps appearing
> Nearly every intervention that reliably extends lifespan in a model organism (dietary restriction,
> reduced growth-hormone signalling, several genetic mutants including the long-lived
> [[cynthia-kenyon|daf-2 nematodes]]) converges on reduced mTORC1 output and
> increased autophagy. Rapamycin is the most direct available way to hit that node with a drug.

## Development history

```timeline
[
  { "year": "1964", "title": "Soil sample collected", "text": "A Canadian medical expedition to Rapa Nui (Easter Island) collects soil that later yields Streptomyces hygroscopicus." },
  { "year": "1972", "title": "Compound isolated", "text": "Researchers at Ayerst in Montreal isolate an antifungal macrolide from the sample and name it rapamycin after the island." },
  { "year": "1991", "title": "TOR genes identified", "text": "Yeast mutants resistant to rapamycin reveal the TOR genes, opening the pathway to molecular study; the mammalian homologue follows in 1994." },
  { "year": "1999", "title": "Approved as sirolimus", "text": "The drug is approved in the United States to prevent rejection in kidney transplantation, establishing decades of human safety data at immunosuppressive doses." },
  { "year": "2009", "title": "Lifespan extension in mice", "text": "The Interventions Testing Program reports that rapamycin started at 600 days of age extends median lifespan in genetically heterogeneous mice of both sexes." },
  { "year": "2014", "title": "Immune benefit in older adults", "text": "A rapalog given at low dose for six weeks improves antibody response to influenza vaccination in adults over 65." },
  { "year": "2019", "title": "Phase 3 failure", "text": "resTORbio's RTB101 does not reduce clinically symptomatic respiratory illness in older adults, ending the most advanced attempt to license an mTOR inhibitor for an aging-related indication." },
  { "year": "2024", "title": "First dedicated longevity RCT reported", "text": "The PEARL trial of intermittent rapamycin in healthy adults reports tolerability and scattered secondary signals, without a clinical benefit." }
]
```

## Evidence in animals

The mouse result is unusually robust. Rapamycin extended median and maximal lifespan in the ITP even
when treatment began at an age equivalent to human late middle age,[^harrison2009] and the effect has
been replicated across ITP cohorts and sites — a standard the field rarely meets. Higher doses
produce larger effects, with a sex difference that has never been fully explained: females generally
show a larger response.[^miller2014]

Lifespan is not the same as [[healthspan]]. Treated mice show delayed decline in several tissues,
including reduced accumulation of [[cellular-senescence|senescent cells]] and better maintenance of
immune and cognitive function, but other age-related pathologies proceed unchanged, and rapamycin
does not produce the uniform rejuvenation across the [[hallmarks-of-aging]] that popular accounts
imply. Work in companion dogs, where the Dog Aging Project's randomized trial tests cardiac and
cognitive outcomes over years, is the most informative non-laboratory test underway; an earlier
ten-week pilot reported improved echocardiographic measures, and the programme's funding was
disrupted in 2024. A separate commercial effort, [[loyal]], pursues canine lifespan drugs through
the veterinary regulator on other mechanisms. In common marmosets, daily dosing has been reported as well tolerated without an
established lifespan effect.

## Human evidence

No trial has tested whether rapamycin changes how long people live or how long they stay healthy.
What exists falls into three groups.

The first is decades of transplant experience, which establishes the safety profile at
immunosuppressive doses: stomatitis and mouth ulcers, raised blood lipids, impaired glucose
handling, delayed wound healing, oedema, and increased infection risk. These are the doses geroscience
does not propose to use.

The second is immune-function work at low or intermittent doses. Rapalog treatment improved influenza
vaccine responses in older adults[^mannick2014] and, in a follow-up combination study, was associated
with fewer reported respiratory infections. This produced the field's most cited paradox: a drug
classed as an immunosuppressant appearing to improve some aspects of immune function in the old,
plausibly by reducing the accumulation of exhausted and senescent lymphocytes that also drives
[[inflammaging]]. The subsequent phase 3 of RTB101 failed, which is the strongest single piece of
negative human evidence about the strategy.

The third is the PEARL trial, a decentralized, placebo-controlled study of weekly rapamycin in
healthy adults over 48 weeks. It reported no serious adverse events and a small number of favourable
secondary measures, with most prespecified outcomes unchanged. It was not designed or powered to
detect a change in aging.

> [!caution] Off-label use runs ahead of evidence
> Rapamycin is prescribed off-label for longevity by a minority of physicians and taken by a
> self-selected community of users. Surveys of that population report a tolerable side-effect profile
> at intermittent doses, but self-reported outcomes from people who chose the drug cannot establish
> benefit, and no dose, schedule, or monitoring standard has been validated for healthy adults.

## Limitations

The dose–schedule question is unresolved. Animal lifespan data come from continuous exposure at
concentrations that would be immunosuppressive in humans; the human longevity community uses weekly
pulses chosen largely by inference from mTORC2 pharmacology. Nobody knows whether the pulsed schedule
retains the effect that the continuous schedule produced in mice.

The endpoint question is worse. Because no regulator accepts a measure of
[[aging-biomarkers|biological aging]] as a surrogate, a trial would have to be powered on disease
events over years. A movement in an [[epigenetic-clock]] reading or in a
[[biological-age|composite age estimate]] is not evidence of clinical benefit. Rapamycin is also
generic, so no company has a commercial reason to fund such a trial — the same structural problem
that has stalled [[metformin|the TAME trial]], which [[nir-barzilai]] has spent more than a decade
trying to finance, and the reason the [[longevity-dividend]] remains an argument for research rather
than a description of an available drug.

## Outlook

Two lines could break the impasse. Bi-steric and otherwise mTORC1-selective inhibitors, designed to
leave complex 2 alone, would remove the metabolic penalty that limits chronic dosing and would test
whether the metabolic side effects and the geroprotective effect are separable. Narrow indications
with measurable endpoints, among them vaccine response, periodontal disease,
[[reproductive-longevity|ovarian reserve]], and age-related cardiac decline, offer trials short
enough to run and specific enough to interpret.

Until one of those produces a positive controlled result, rapamycin occupies an awkward position: the
best animal evidence in [[geroscience-hypothesis|geroscience]], a plausible mechanism converging with
[[senolytics]] and dietary restriction on shared damage pathways, and nothing in humans that would
justify recommending it to a healthy person over the interventions with actual outcome data, starting
with [[exercise-and-aging|exercise]].

## See also

- [[autophagy]]
- [[caloric-restriction]]
- [[metformin]]
- [[senolytics]]
- [[geroscience-hypothesis]]
- [[hallmarks-of-aging]]
- [[healthspan]]
- [[longevity-dividend]]
- [[nad-precursors]]

## References

[^lamming2012]: `paper` Lamming, D.W. et al. "Rapamycin-induced insulin resistance is mediated by mTORC2 loss and uncoupled from longevity." *Science*, 2012.
[^harrison2009]: `paper` Harrison, D.E. et al. "Rapamycin fed late in life extends lifespan in genetically heterogeneous mice." *Nature*, 2009. {The Interventions Testing Program ran this at three sites in genetically heterogeneous mice; dosing was continuous in food, not the intermittent schedule people take off-label.}
[^miller2014]: `paper` Miller, R.A. et al. "Rapamycin-mediated lifespan increase in mice is dose and sex dependent and metabolically distinct from dietary restriction." *Aging Cell*, 2014.
[^mannick2014]: `paper` Mannick, J.B. et al. "mTOR inhibition improves immune function in the elderly." *Science Translational Medicine*, 2014. {The drug tested was the rapalog everolimus rather than rapamycin, and the endpoint was antibody titre after influenza vaccination rather than any infection outcome.}

==============================================================================
ARTICLE: ray-kurzweil
TITLE:   Ray Kurzweil
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/ray-kurzweil
SOURCE:  https://futurehumanwiki.com/raw/ray-kurzweil
==============================================================================

---
title: "Ray Kurzweil"
slug: "ray-kurzweil"
type: "person"
status: "contested"
horizon: "present"
categories: ["people", "minds"]
tags: ["singularity", "forecasting", "artificial intelligence", "futurism", "life extension"]
summary: "American inventor and futurist who formulated the law of accelerating returns and forecasts human-level machine intelligence by 2029 and a technological singularity by 2045."
updated: "2026-07-27"
issues: ["Hofstadter's often-quoted verdict is paraphrased without a source", "The self-assessment scorecard is described without citing Kurzweil's own review"]
---

```infobox
{
  "caption": "Inventor and futurist",
  "rows": [
    { "label": "Born", "value": "12 February 1948, New York City" },
    { "label": "Nationality", "value": "American" },
    { "label": "Education", "value": "MIT (BS, 1970)" },
    { "label": "Known for", "value": "Law of accelerating returns", "link": "/wiki/accelerating-change" },
    { "label": "Field", "value": "Pattern recognition; forecasting" },
    { "label": "Affiliation", "value": "Google (from 2012)" },
    { "label": "Key book", "value": "The Singularity Is Near (2005)" }
  ]
}
```

**Ray Kurzweil** is an American inventor, computer scientist and futurist whose reading of
technological history — that information technologies improve exponentially rather than linearly,
and that the exponent itself is stable across substrates — underwrites his forecast of a
[[technological-singularity]] around 2045. He built a substantial engineering career in pattern
recognition before becoming the most widely read forecaster of machine intelligence, and he is a
prominent proponent of radical life extension.

## Career

Kurzweil built a pattern-recognition program that matched high-school students to colleges while
still in high school, and studied at MIT under Marvin Minsky. His inventions cluster around reading
and hearing. In the mid-1970s he combined the first omni-font optical character recognition with a
flatbed scanner and a speech synthesizer to produce the Kurzweil Reading Machine, a print-to-speech
device for blind readers developed with the National Federation of the Blind. He later founded
Kurzweil Music Systems, whose K250 synthesizer reproduced sampled acoustic instruments, and
Kurzweil Applied Intelligence, an early large-vocabulary speech-recognition company. He received
the National Medal of Technology in 1999 and entry to the National Inventors Hall of Fame in 2002.

In 2012 he joined Google as a director of engineering working on natural-language understanding,
and he has remained affiliated with the company since. His forecasting work runs in parallel:
*The Age of Intelligent Machines* (1990), *The Age of Spiritual Machines* (1999), *The Singularity
Is Near* (2005), *How to Create a Mind* (2012) and *The Singularity Is Nearer* (2024).

## The law of accelerating returns

Kurzweil's central claim is that evolutionary and technological processes that use their own
outputs as inputs improve at an exponential rather than linear rate, and that the rate of
exponential improvement itself increases.[^lar2001] He supports this with time-series plots of
computing price-performance running back through electromechanical relays, vacuum tubes and
transistors, arguing that Moore's law is the fifth in a sequence of substrate-specific curves
rather than a one-off. The generalized claim is that when one paradigm saturates, another takes
over, so the composite curve keeps its slope. See [[accelerating-change]] for the broader debate
over this style of argument.

```keyfacts
[
  { "value": "2029", "label": "Predicted date for human-level AI", "note": "a date Kurzweil has held since 1999" },
  { "value": "2045", "label": "Predicted singularity", "note": "when he expects nonbiological intelligence to dominate" },
  { "value": "5", "label": "Computing paradigms in his composite curve", "note": "relays through integrated circuits" }
]
```

From the curve he derives specific dates. Sufficient compute to emulate a human brain arrives at
consumer prices in the late 2020s; artificial intelligence passes a properly administered Turing
test by 2029; and by 2045 machine intelligence so exceeds the biological kind that projection past
that point fails. He also treats the merger of human and machine cognition as the expected outcome
rather than displacement, an argument developed in [[human-ai-merger]].

## Longevity and the three bridges

Kurzweil's interest in aging is personal: his father died of heart disease in his fifties, and
Kurzweil was diagnosed with type 2 diabetes as a young man. With the physician Terry Grossman he
wrote *Fantastic Voyage* (2004) and *Transcend* (2009), setting out a "three bridges" argument.
Bridge one is aggressive present-day management of health through diet, exercise and
[[dietary-supplements|supplements]] —
Kurzweil has reported taking a very large number of pills daily, a figure he has said fell over
time. Bridge two is biotechnology capable of reprogramming the body's information processes, the
domain of [[gene-therapy-for-aging]] and [[epigenetic-reprogramming]]. Bridge three is
[[medical-nanorobots]] operating inside cells. The argument is that a person need only survive on
each bridge long enough to reach the next, which is the same structure as
[[longevity-escape-velocity]]. He has also said he expects to be cryopreserved if he dies before
the bridges arrive; see [[cryonics]].

> [!caution] Bridge one has thin support
> The supplement regimen at the base of the argument is not backed by clinical evidence for
> life extension in humans. Trials of [[nad-precursors]], [[metformin]] and antioxidant
> supplementation have not demonstrated an effect on human lifespan, and
> [[exercise-and-aging|exercise]] remains the only well-evidenced geroprotector.

## Predictions and their scorecard

Kurzweil has published self-assessments of his own forecasts, scoring most of his 1999 predictions
for 2009 as correct or "essentially correct". Independent reviewers have been harsher, noting that
many predictions were phrased loosely enough to admit generous grading, and that his most confident
near-term calls — ubiquitous speech interfaces, self-driving cars, translating telephones — arrived
years to decades later than stated, or in narrower form.

The strongest structural criticism is that curve-fitting is not a mechanism. Paul Allen and Mark
Greaves argued that software understanding of biology and cognition faces a "complexity brake" that
no hardware curve resolves.[^allen2011] Theodore Modis, whose own work Kurzweil cites, has
disputed the data selection behind the composite exponential.[^modis2006] Douglas Hofstadter's
often-quoted verdict is that the good and the indefensible ideas are mixed together so thoroughly
that separating them is the reader's problem.

Large language models have complicated both sides. Systems released since 2022 achieved
conversational fluency and coding ability on roughly the schedule Kurzweil's compute curve implied,
which his supporters treat as confirmation. Critics observe that the same systems fail at
persistent memory, reliable reasoning and grounded world models in ways the brain-emulation framing
did not anticipate, and that the connection between his compute extrapolation and any specific
capability was always asserted rather than derived. See
[[artificial-general-intelligence]] for the definitional fight this exposes.

## Reception and legacy

Kurzweil's practical legacy is assistive technology: reading machines and speech systems that
changed what blind and print-disabled people could do, decades before the forecasting became his
public identity. His intellectual legacy is more contested. He gave [[singularitarianism]] its
popular form and its dates, and *The Singularity Is Near* did more than any other book to move the
idea from science fiction and the [[extropianism|extropian]] mailing lists into mainstream
discussion. Nearly every subsequent treatment of machine superintelligence, including work by
[[nick-bostrom]] that reaches very different conclusions about risk, positions itself relative to
him.

The unresolved question is whether his method has any predictive content beyond the hardware curve
it started from. Compute has behaved close to the way he said it would. Whether that fact licenses
the claim about 2045 — or about anything downstream of hardware at all — is exactly what his
critics deny, and no result since 2022 has settled it.

## See also

- [[technological-singularity]]
- [[accelerating-change]]
- [[singularitarianism]]
- [[artificial-general-intelligence]]
- [[human-ai-merger]]
- [[mind-uploading]]
- [[nick-bostrom]]
- [[eric-drexler]]

## References

[^lar2001]: `statement` Kurzweil, R. "The Law of Accelerating Returns." Essay, 2001. {Kurzweil's own statement of the argument; the historical series in it are his selection and his normalisation, which is precisely what critics dispute.}
[^allen2011]: `news` Allen, P. and Greaves, M. "The Singularity Isn't Near." *MIT Technology Review*, 2011. {An opinion essay by two software figures in a technology magazine, not a research result; its claim is about software understanding rather than about the hardware curve.}
[^modis2006]: `paper` Modis, T. "The Singularity Myth." *Technological Forecasting & Social Change*, 2006.

==============================================================================
ARTICLE: reproductive-longevity
TITLE:   Reproductive longevity
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/reproductive-longevity
SOURCE:  https://futurehumanwiki.com/raw/reproductive-longevity
==============================================================================

---
title: "Reproductive longevity"
slug: "reproductive-longevity"
type: "concept"
status: "emerging"
horizon: "2030s"
categories: ["reproduction", "longevity"]
tags: ["aging", "ovary", "menopause", "fertility", "oocyte", "healthspan"]
summary: "The functional lifespan of the reproductive system, and efforts to extend it in the ovary, which loses function decades before most other organs."
updated: "2026-07-27"
humanEvidence: "Oocyte and ovarian tissue cryopreservation are established practice in women, but no intervention has been shown to delay human menopause; the rapamycin work reports surrogate ovarian-reserve measures in a small group."
issues: ["The Columbia rapamycin ovarian trial is described from preliminary reports with no citation", "Anti-Mullerian hormone therapeutic development is attributed to an unnamed company"]
---

```infobox
{
  "caption": "Concept in reproductive biology and geroscience",
  "rows": [
    { "label": "Fastest-aging system", "value": "Ovary" },
    { "label": "Median menopause", "value": "About 51 years" },
    { "label": "Oocyte pool", "value": "Fixed before birth" },
    { "label": "Main quality failure", "value": "Chromosome missegregation" },
    { "label": "Established preservation", "value": "Oocyte and ovarian tissue freezing" },
    { "label": "Proven delay of menopause", "value": "None" }
  ]
}
```

**Reproductive longevity** is the duration over which the reproductive system remains functional, and, as a research programme, the attempt to extend it. In women the relevant organ is the ovary, which loses most of its function around the age of 51 while the heart, liver, and brain are still decades from failure. That asymmetry makes the ovary the clearest example in human biology of an organ ageing on its own schedule, and it is why the field has become a testbed for the wider claims of the [[geroscience-hypothesis]]. The uterus is not the constraint: it can carry a pregnancy well after menopause when donated oocytes are used, and [[uterus-transplantation]] has produced live births in women who had no working one of their own.

```keyfacts
[
  { "value": "~51", "label": "Median age at natural menopause", "note": "in high-income populations; varies by genetics, smoking, and parity" },
  { "value": "Before birth", "label": "When the oocyte pool is fixed", "note": "peak follicle number occurs in mid-gestation and declines from then on" },
  { "value": "~400", "label": "Oocytes ovulated in a lifetime", "note": "out of hundreds of thousands present at puberty; the rest are lost to atresia" }
]
```

## Two clocks in one organ

The ovary declines along two axes that are often conflated.

**Quantity.** Germ cells reach their maximum number in mid-gestation, at several million, and fall continuously thereafter through atresia, a programmed loss largely independent of ovulation. By puberty a few hundred thousand primordial follicles remain; by menopause, a few hundred. Clinically this pool is estimated by anti-Müllerian hormone in blood and by antral follicle count on ultrasound, neither of which reliably predicts when an individual will reach menopause.

**Quality.** The proportion of eggs that segregate their chromosomes correctly falls steeply from the mid-thirties. The mechanistic explanation with the strongest support concerns cohesin, the protein complex that holds sister chromatids together. It is loaded onto chromosomes in fetal oocytes and is not appreciably replenished afterwards, so a woman's oocytes have been holding their chromosomes with the same molecules for decades by the time they are ovulated. Cohesion deteriorates, the meiotic spindle checkpoint is permissive in oocytes, and missegregation follows. Declining mitochondrial function in the oocyte, connected to the mechanisms in [[mitochondrial-dysfunction]], is a contributing but probably secondary factor.

The consequence is the age curve familiar from fertility clinics: the fraction of embryos that are euploid falls from a clear majority in the early thirties to a minority by the early forties, which is the dominant reason live-birth rates per cycle decline and the reason [[embryo-selection]] is used most in exactly the group where there are fewest embryos to select from.

> [!key] Why this is a geroscience problem, not only a fertility problem
> Earlier menopause is associated with higher risk of cardiovascular disease, osteoporosis, and all-cause mortality, and surgical removal of the ovaries before natural menopause carries similar associations. In mice, transplanting young ovaries into old ovariectomised animals extended lifespan, suggesting the ovary is not merely an early casualty of ageing but a contributor to it.[^cargill2003] Whether the same causal arrow operates in humans is unresolved.

## Why the ovary ages first

There is no consensus mechanism, but three explanations recur.

The first is the fixed pool: unlike blood, gut, or skin, the ovary has no renewing stem cell population that is generally accepted, so losses are permanent. Claims of oogonial stem cells in adult human ovaries have been made and remain disputed, with several laboratories unable to reproduce the key findings.

The second is DNA damage. Genome-wide association studies of age at natural menopause identify hundreds of loci enriched for genes in DNA damage response and repair pathways, and manipulating one of them, the checkpoint kinase *CHEK2*, extended reproductive lifespan in mice.[^ruth2021] Human carriers of loss-of-function variants in the same gene reach menopause later. This is one of the cleaner links from a genetic association to a candidate intervention anywhere in [[hallmarks-of-aging]] research.

The third is evolutionary. Menopause is rare among mammals, documented in humans and in a handful of toothed whale species. The grandmother hypothesis holds that a post-reproductive phase was selected because assisting descendants outweighed continued childbearing; a competing account emphasises reproductive conflict between generations of females sharing a group. Both imply that the ovary's schedule is not simply a failure but a trait under selection, which complicates the assumption that extending it is straightforwardly beneficial.

## What can be done now

Existing practice preserves gametes rather than extending the organ.

**Oocyte cryopreservation.** Vitrification made egg freezing viable, and professional bodies removed the experimental label from it in the 2010s. Outcomes depend heavily on the age at which eggs are frozen and on how many are stored; freezing at 38 buys much less than freezing at 30, and clinics vary widely in how clearly they communicate that.

**Ovarian tissue cryopreservation.** Strips of ovarian cortex are removed, frozen, and later grafted back, restoring both fertility and endocrine function for a period. Developed for patients facing gonadotoxic cancer treatment, it now has a substantial record of live births and is no longer classed as experimental by the main professional societies. Companies have marketed it to healthy women as a way to delay menopause, a use for which there is no controlled evidence and which the same societies have criticised.

**Hormone therapy.** Replacing oestrogen treats vasomotor symptoms and preserves bone density. It does not preserve the ovary, and the trial evidence on cardiovascular and mortality outcomes remains contested, with the timing of initiation relative to menopause the main variable in dispute. It is a treatment for consequences, not a geroprotector.

## Interventions under test

The candidate that has drawn most attention is [[rapamycin]]. Inhibiting mTOR slows the activation of primordial follicles in mice, preserving the reserve. A small trial at Columbia University has tested low-dose weekly rapamycin in women in their late thirties and early forties with ovarian reserve and follicular decline as endpoints; preliminary reports have described a slower decline in the treated group. The study is small, short, and uses surrogate measures, and no result of this kind has yet been shown to translate into a later menopause or a live birth advantage.

Anti-Müllerian hormone is being developed from the other direction. It acts as a brake on primordial follicle recruitment, and sustained administration protects the reserve in animal models; at least one company is developing it as a therapeutic. Senolytic and NAD-raising approaches, discussed in [[senolytics]] and [[nad-precursors]], have shown effects on ovarian markers in rodents with no human ovarian data to speak of.

The area acquired dedicated funding in the late 2010s through a philanthropically supported consortium hosted at the [[buck-institute]], which treats ovarian ageing as a research programme in its own right rather than a subfield of fertility medicine. That framing is itself the argument: reproductive ageing had previously been studied by clinicians trying to achieve pregnancies rather than by biologists trying to understand why one organ fails first.

> [!caution] The endpoint problem
> Every intervention above is measured against surrogates: anti-Müllerian hormone, antral follicle count, oocyte yield. None of these has been validated as predicting the outcomes people care about, namely a healthy birth or a later menopause. This is the same obstacle described in [[aging-biomarkers]] and the same correlation-versus-causation trap that limits the [[epigenetic-clock]], and it is arguably more acute here because a definitive trial would have to run for a decade.

The bypass route is different in kind. [[in-vitro-gametogenesis]] would make eggs from somatic cells, decoupling the age of the gamete from the age of the ovary entirely. It would not address the endocrine consequences of ovarian failure, and it carries its own unresolved question about whether a gamete made from a fifty-year-old's cells inherits that person's accumulated mutations.

## Male reproductive aging

The male system declines more gradually and differently. Sperm production continues into old age, but semen parameters deteriorate slowly and the rate of de novo mutations transmitted to offspring rises with paternal age, by roughly one to two additional mutations per year of the father's age at conception.[^kong2012] This contributes to a measurable increase in the risk of some neurodevelopmental conditions with older fathers, though the absolute risks remain small. There is no male equivalent of menopause and no clear analogue of the ovarian reserve, which is why the field is overwhelmingly focused on the ovary.

## Open questions

Whether ovarian ageing causes systemic ageing or merely precedes it is the central unresolved question, and it determines whether extending reproductive lifespan is a fertility intervention or a [[healthspan]] intervention. The mouse transplant data point one way; the association between later menopause and higher rates of hormone-sensitive cancers points to a trade-off rather than a free gain.

A second question is whether the ovary is a good general model for organ-specific ageing. Its advantages are obvious: a clear functional endpoint, a measurable decline, a timescale short enough for trials, and a large motivated population. Its disadvantages are equally clear. The fixed-pool architecture is unusual, and an organ that runs down a non-renewable stock may say little about renewing tissues, whose failure looks more like [[stem-cell-exhaustion]] or the accumulation described in [[cellular-senescence]]. If it does generalise, reproductive longevity becomes the first place a geroprotective drug could be shown to work in humans within a decade rather than a lifetime.

## See also

- [[in-vitro-gametogenesis]]
- [[embryo-selection]]
- [[rapamycin]]
- [[aging-biomarkers]]
- [[hallmarks-of-aging]]
- [[geroscience-hypothesis]]
- [[healthspan]]
- [[mitochondrial-dysfunction]]

## References

[^cargill2003]: `paper` Cargill, S. L. et al. "Age of ovary determines remaining life expectancy in old ovariectomized mice." *Aging Cell*, 2003.
[^ruth2021]: `paper` Ruth, K. S. et al. "Genetic insights into biological mechanisms governing human ovarian ageing." *Nature*, 2021.
[^kong2012]: `paper` Kong, A. et al. "Rate of de novo mutations and the importance of father's age to disease risk." *Nature*, 2012.

==============================================================================
ARTICLE: respirocytes
TITLE:   Respirocytes
PORTAL:  Nanomedicine
URL:     https://futurehumanwiki.com/wiki/respirocytes
SOURCE:  https://futurehumanwiki.com/raw/respirocytes
==============================================================================

---
title: "Respirocytes"
slug: "respirocytes"
type: "concept"
status: "speculative"
horizon: "indefinite"
categories: ["nanomedicine"]
tags: ["nanomedicine", "nanorobotics", "blood", "oxygen", "design study", "enhancement"]
summary: "Robert Freitas's 1998 design study for an artificial red blood cell built as a pressurised micron-scale gas tank, widely cited and never built."
updated: "2026-07-28"
humanEvidence: "No respirocyte has been built, so none has been tested in any person or animal; the published performance figures are outputs of a 1998 physical model rather than measurements."
issues: ["Drexler's Nanosystems is discussed but does not appear in the reference list"]
---

```infobox
{
  "caption": "Nanomedical design study",
  "rows": [
    { "label": "Proposed by", "value": "Robert A. Freitas Jr." },
    { "label": "First described", "value": "1998" },
    { "label": "Proposed diameter", "value": "About 1 micrometre" },
    { "label": "Proposed material", "value": "Diamondoid pressure vessel" },
    { "label": "Function", "value": "Oxygen and carbon dioxide transport" },
    { "label": "Power source", "value": "Serum glucose (proposed)" },
    { "label": "Prototypes built", "value": "None" },
    { "label": "Status", "value": "Theoretical design only" }
  ]
}
```

**Respirocytes** are a proposed artificial replacement for the red blood cell: micron-scale pressure vessels that would load oxygen in the lungs, release it in tissue, and carry carbon dioxide back, storing both gases at pressures far beyond anything haemoglobin achieves. The design was published by [[robert-freitas|Robert Freitas]] in 1998 and has been reproduced in popular accounts of nanomedicine ever since.[^freitas1998] It is a calculation, not a device; nothing resembling a respirocyte has been fabricated, and the manufacturing method it assumes does not exist.

## Overview

A red blood cell carries oxygen by reversible binding to haemoglobin. Freitas proposed carrying it as compressed gas instead. A respirocyte, in his description, is a roughly spherical vessel about a micrometre across with separate high-pressure chambers for oxygen and carbon dioxide, an outer hull of diamondoid carbon, molecular sorting rotors in the wall that pump specific gas molecules in or out, an onboard sensor and computer to decide when to load and unload, and a glucose engine drawing fuel from blood plasma.

Because gas under roughly a thousand atmospheres is far denser than gas bound to protein, the device would carry much more oxygen per unit volume than the cell it replaces. Freitas put the advantage at 236 times that of a natural red cell and suggested that an injection of a concentrated suspension containing trillions of devices could let a person hold their breath for hours or sprint at full effort for several minutes without breathing.[^freitas1998] Those figures are the source of the design's fame, and they are outputs of the model rather than measurements.

> [!note] What a design study is
> Freitas calls this work *exploratory engineering*: an attempt to show that a device is consistent with physical law and to bound its performance if built. It deliberately does not address how to build it. Treating the resulting numbers as a technology forecast misreads the genre, a confusion that recurs throughout the [[medical-nanorobots]] literature.

## Origins

The respirocyte follows directly from [[eric-drexler]]'s programme for atomically precise manufacturing. Drexler's *Nanosystems* had argued that diamondoid components — stiff, chemically inert lattices of carbon — could be analysed with classical mechanical engineering, and that positional chemistry could in principle build them. Freitas took that toolkit and applied it to medicine, first in the respirocyte paper and then across the volumes of *Nanomedicine*.

The respirocyte was the opening move because oxygen transport is unusually tractable: a single well-defined cargo, a simple duty cycle, and a natural benchmark to beat. Later designs in the same series extended the approach to immune defence (the microbivore, an artificial phagocyte), haemostasis (the clottocyte), and chromosome replacement (the chromallocyte).

## How it would work

The proposed cycle is straightforward. In pulmonary capillaries, sorting rotors in the hull bind oxygen molecules from plasma and pump them into the oxygen chamber while a second set vents carbon dioxide. In systemic capillaries the process reverses. Onboard sensors track partial pressures and the internal computer decides when to release, with the whole population addressable by external ultrasound so a clinician could switch devices on or off.

Two features distinguish it from haemoglobin. It stores gas rather than binding it, so its capacity is set by tank pressure rather than by binding chemistry. And it releases on command rather than in response to local pH and oxygen tension. Haemoglobin's cooperative binding and the Bohr effect give natural red cells automatic regulation — they unload where tissue is acidic and oxygen-poor — which a pressure vessel must reproduce with active control.

## The assumptions it rests on

The design presupposes a fabrication route. Diamondoid mechanosynthesis, the positional construction of stiff carbon lattices, has never been demonstrated; the dispute over whether it is chemically possible is covered in [[molecular-assembler]]. Without it there is no path to the hull, the rotors, or the internal machinery.

It presupposes a scale of production with no precedent. Replacing even a modest fraction of an adult's roughly 25 trillion red cells requires manufacturing devices in comparable numbers, each functional and sterile.

It presupposes biocompatibility that has not been tested because the surfaces do not exist. A rigid micron-scale particle in blood is, to the innate immune system, about the size and shape of a bacterium. Freitas devoted a volume of *Nanomedicine* to the question and argued that suitably passivated diamondoid surfaces would provoke little response, but the analysis cannot be checked.[^freitas2003] Real particles in blood acquire a protein corona within seconds, and that layer, not the engineered surface, largely determines clearance — the recurring lesson of [[targeted-drug-delivery]].

It presupposes safe failure. A rigid sphere is not a deformable disc; red cells fold to pass through capillaries narrower than themselves. Trillions of incompressible particles in the microvasculature raise an obstruction risk that the design addresses by assumption rather than by test. A ruptured thousand-atmosphere vessel is a further hazard with no natural analogue.

## Criticism

The most substantive objection is physiological rather than nanotechnical. Oxygen delivery in a healthy person is not limited by the blood's carrying capacity. At rest, tissues extract only about a quarter of the oxygen arriving in arterial blood; the limits on delivery are cardiac output, capillary transit time, and the diffusion distance from capillary to mitochondrion. A carrier with 236 times the volumetric capacity does not multiply whole-body oxygen delivery by 236, because the bottleneck sits downstream of the carrier. It would matter most where carrying capacity genuinely binds — severe anaemia, carbon monoxide poisoning, submersion — and much less as a general enhancement.

Carbon dioxide transport is also mischaracterised by the simple picture. Most CO₂ leaves tissue as bicarbonate dissolved in plasma rather than bound to haemoglobin, so a device that handles gaseous CO₂ addresses a minority of the traffic.

Hyperoxia is a hazard in its own right. Raising tissue oxygen tension above normal causes vasoconstriction and increases oxidative damage, so a device that could deliver much more oxygen than physiology calls for would need to be prevented from doing so most of the time.

The comparison case is instructive. Attempts to build a functional oxygen carrier from ordinary chemistry have a long record of failure: cell-free haemoglobin-based carriers scavenge nitric oxide and cause vasoconstriction, and a pooled analysis of their trials found increased rates of death and myocardial infarction.[^natanson2008] Perfluorocarbon emulsions performed poorly for related reasons. The history in [[artificial-blood]] suggests that the hard part of replacing a red cell is not storage density but everything the red cell does incidentally.

> [!debate] The disagreement
> Proponents treat the respirocyte as an engineering target awaiting a fabrication technology. Most biomedical engineers treat it as an illustration of what atomically precise manufacturing would permit, with no bearing on what can be attempted now. Both readings accept that the device cannot currently be built; they differ on whether the design constitutes progress toward building it.

## Status and influence

As of 2026 the respirocyte remains what it was in 1998: a paper design, occasionally updated in review articles, never fabricated in whole or in part. The functioning nanoscale therapeutics that reached patients in the intervening decades came from a different direction entirely — liposomes, albumin-bound drugs, [[lipid-nanoparticles]], and the addressable structures of [[dna-nanotechnology]] — none of which uses rigid machinery or stored pressure.

Its influence has been rhetorical rather than technical. The respirocyte gave [[transhumanism]] a concrete image of a body improved by engineering rather than by medicine, and it appears in discussions of [[human-enhancement]] and [[morphological-freedom]] as the standard example of a purely elective internal modification. It also anticipates a governance problem: an oxygen carrier that could be switched on before an event would be undetectable by the metabolic markers used in [[gene-doping]] screening, which makes it a recurring hypothetical in debates over [[enhancement-in-sport]]. The applications most often cited as legitimate — extending survival in trauma, supporting crews in [[space-medicine]] scenarios, buying time in induced [[human-hibernation]] — all depend on a device that no one knows how to make.

## See also

- [[medical-nanorobots]]
- [[molecular-assembler]]
- [[artificial-blood]]
- [[dna-nanotechnology]]
- [[eric-drexler]]
- [[human-enhancement]]
- [[grey-goo]]
- [[microrobots-in-medicine]]

## References

[^freitas1998]: `paper` Freitas, R.A. "Exploratory Design in Medical Nanotechnology: A Mechanical Artificial Red Cell." *Artificial Cells, Blood Substitutes, and Immobilization Biotechnology*, 1998. {A design paper: it derives performance from physical models and reports no fabricated device, experiment, or measurement.}
[^freitas2003]: `book` Freitas, R.A. *Nanomedicine, Volume IIA: Biocompatibility.* Landes Bioscience, 2003.
[^natanson2008]: `paper` Natanson, C. et al. "Cell-free hemoglobin-based blood substitutes and risk of myocardial infarction and death: a meta-analysis." *JAMA*, 2008. {A pooled analysis across several different haemoglobin-based products and indications rather than a trial of any one carrier.}

==============================================================================
ARTICLE: retinal-implant
TITLE:   Retinal implants and visual prostheses
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/retinal-implant
SOURCE:  https://futurehumanwiki.com/raw/retinal-implant
==============================================================================

---
title: "Retinal implants and visual prostheses"
slug: "retinal-implant"
type: "technology"
status: "experimental"
horizon: "late 2020s"
trl: 6
categories: ["cybernetics"]
tags: ["vision", "sensory restoration", "implants", "blindness", "neuroprosthetics", "phosphenes"]
summary: "Implanted devices that restore rudimentary vision to blind people by electrically stimulating surviving retinal neurons or, more invasively, the visual cortex."
updated: "2026-07-27"
humanEvidence: "Devices have been implanted in blind people for over a decade; the strongest result is a 2025 European trial in which most PRIMA recipients identified letters and read words under electronic magnification."
access: "Both commercially approved devices have been withdrawn and existing recipients have lost manufacturer support; access now runs through clinical trials, chiefly of the PRIMA subretinal system, at device-plus-surgery costs among the highest in ophthalmology."
reversibility: "difficult"
issues: ["The 2025 PRIMA pivotal trial result is the strongest claim here and carries no citation", "Phosphene-count estimates from simulation studies are stated without a source"]
---

```infobox
{
  "caption": "Visual neuroprosthesis",
  "rows": [
    { "label": "Type", "value": "Sensory neuroprosthesis" },
    { "label": "Approaches", "value": "Epiretinal, subretinal, cortical" },
    { "label": "First CE-marked device", "value": "Argus II, 2011" },
    { "label": "US authorization", "value": "2013, humanitarian exemption" },
    { "label": "Electrode counts", "value": "60–400 (retinal)" },
    { "label": "Best demonstrated acuity", "value": "Well below 20/200" },
    { "label": "Commercial status", "value": "Two leading devices withdrawn" },
    { "label": "Readiness", "value": "TRL 6" }
  ]
}
```

**Retinal implants and visual prostheses** are devices that produce visual sensation in blind people
by stimulating the surviving parts of the visual pathway with electric current. They target
conditions in which the photoreceptors have died but the downstream retina and optic nerve remain
largely intact — chiefly retinitis pigmentosa and the geographic atrophy form of age-related macular
degeneration. Every device fielded so far produces vision built from phosphenes, discrete spots of
light with no colour and little detail, and no system has restored acuity anywhere near the legal
threshold for blindness.

## How it works

A camera, usually mounted on spectacles, captures a scene. A processor reduces the image to a
low-resolution activation map and transmits it to an implanted array, which converts each pixel into
a current pulse delivered to nearby retinal neurons. Because the retina is a layered circuit rather
than a screen, where the electrodes sit changes what they excite.

**Epiretinal** arrays sit on the inner retinal surface and stimulate retinal ganglion cells, the
output neurons whose axons form the optic nerve. This bypasses all intervening retinal processing,
so the signal that reaches the brain is not the code the retina would normally send; axons of
passage from distant regions are also stimulated, producing elongated or displaced phosphenes.
Argus II was the leading epiretinal device.

**Subretinal** arrays sit under the retina in the space the dead photoreceptors vacated and
stimulate bipolar cells, letting the remaining retinal circuitry perform some of its normal
processing. Photovoltaic designs go further: each pixel is a photodiode that converts projected
light into local current, so the implant needs no cables and the eye's own movements scan the image
across the array, preserving the natural link between gaze and percept.

**Cortical** devices skip the eye entirely and stimulate primary visual cortex, which makes them the
only option for people whose optic nerve is destroyed by glaucoma or trauma. They are also the most
invasive, they require a craniotomy, and the retinotopic map they must exploit is folded into a
sulcus, so much of it is difficult to reach with a surface array.

```compare
{
  "columns": ["Epiretinal", "Subretinal photovoltaic", "Cortical"],
  "rows": [
    { "label": "Cells stimulated", "values": ["Ganglion cells", "Bipolar cells", "Cortical neurons"] },
    { "label": "Retinal processing retained", "values": ["Almost none", "Partial", "None"] },
    { "label": "Eye movements usable", "values": ["No", "Yes", "No"] },
    { "label": "Surgery", "values": ["Vitrectomy, tack", "Subretinal insertion", "Craniotomy"] },
    { "label": "Applies to optic nerve damage", "values": ["No", "No", "Yes"] }
  ]
}
```

## Development history

Electrical stimulation of the occipital cortex was known to produce phosphenes from work in the
mid-twentieth century, and Giles Brindley and William Lewin implanted an 80-electrode array over
visual cortex in a blind volunteer in 1968, demonstrating that patterned percepts were possible and
that the engineering was far from ready. Retinal approaches, which require far less surgery,
dominated from the 1990s.

```timeline
[
  { "year": "1968", "title": "First cortical visual prosthesis", "text": "Brindley and Lewin implant an array of surface electrodes over the occipital cortex of a blind volunteer, who reports discrete phosphenes at predictable locations." },
  { "year": "2011", "title": "Argus II CE-marked", "text": "Second Sight's 60-electrode epiretinal implant is approved in Europe for retinitis pigmentosa, followed by a US humanitarian device exemption in 2013." },
  { "year": "2013", "title": "Subretinal photodiode arrays", "text": "Retina Implant AG's Alpha IMS, with roughly 1,500 light-sensitive pixels, receives European approval; the company later ceases operations." },
  { "year": "2019–2022", "title": "Commercial collapse", "text": "Second Sight halts Argus II production and nearly winds down; Retina Implant AG closes. Implanted patients lose software support and repair paths." },
  { "year": "2020–2021", "title": "Cortical stimulation revisited", "text": "A Utah array in the visual cortex of a blind volunteer supports letter identification, and a 1,024-channel array in monkey V1 evokes recognizable shapes." },
  { "year": "2024–2025", "title": "Photovoltaic results and new ownership", "text": "Pixium Vision enters receivership and its PRIMA subretinal system is acquired by Science Corporation; trial results reported in 2025 describe letter and word reading with electronic zoom in most implanted participants." }
]
```

## Current state

As of 2026 the field is in an unusual position: its two commercially approved devices have been
withdrawn, and its most promising results come from a system whose developer went insolvent.

The PRIMA subretinal photovoltaic chip is a wireless array of a few hundred pixels implanted under
the macula, driven by near-infrared images projected from augmented-reality glasses. Feasibility
work in geographic atrophy showed that patients could perceive prosthetic vision in the atrophic
central field while retaining their natural peripheral vision, an unusual property of the
approach.[^palanker2020] Results from the pivotal European trial, reported in 2025, described most
implanted participants regaining the ability to identify letters and read words when the projected
image was electronically magnified. This is the strongest functional result any visual prosthesis
has produced, and it should be read carefully: reading with a zoom function at a few pixels per
letter is not restored sight.

Cortical work has been rebuilt on modern intracortical hardware. A 96-channel [[utah-array]]
implanted in the visual cortex of a blind volunteer produced stable, reproducible phosphenes and
supported identification of simple letter shapes.[^fernandez2021] A high-channel-count array in
monkey V1 evoked percepts of shapes and letters assembled from many simultaneous
phosphenes.[^chen2020] Both establish feasibility; neither approaches a usable device.

## Why resolution is the binding constraint

The intuition that a prosthesis with *n* electrodes yields an *n*-pixel image is wrong in a way that
governs the whole field. Current spreads, so adjacent electrodes excite overlapping populations and
phosphenes merge. Epiretinal stimulation activates axons passing over the array from other retinal
regions, so a phosphene may appear far from the stimulated site and be drawn out into a streak.
Percepts fade under sustained stimulation, so images must be refreshed. And the phosphene map is
idiosyncratic to each patient and must be characterized individually.

Simulation studies of prosthetic vision in sighted observers suggest that face recognition and
fluent reading require on the order of hundreds to thousands of well-separated, independently
controllable phosphenes. Fielded retinal devices have delivered tens. The gap is not merely
manufacturing: packing electrodes more densely without proportionally increasing current confinement
does not increase the number of distinguishable percepts, the same saturation that limits the
[[cochlear-implant]] to roughly eight effective channels.

> [!caution] What "restored sight" means in this field
> Reported successes are usually object localization, motion detection, and letter identification
> under favourable conditions. Users typically retain their cane and continue to rely on other
> senses. Press coverage that describes patients as "seeing again" overstates every result published
> to date.

## Risks and the abandoned-patient problem

Surgical risks include retinal detachment, hypotony, conjunctival erosion over the implant, and
endophthalmitis; cortical devices add the risks of craniotomy and of a percutaneous connector where
one is used. Long-term electrode-tissue stability is limited by the same foreign-body response that
degrades cortical recordings in a [[brain-computer-interface]].

The distinctive risk is commercial. When Second Sight stopped supporting Argus II, several hundred
implanted patients were left with devices that could not be repaired or upgraded, and some lost
function when hardware failed. The episode was widely reported and has become the reference case for
the argument that manufacturers of implanted neurotechnology incur obligations that outlast their
business models.[^spectrum2022] Proposed remedies include escrowed device documentation, mandated
open-sourcing on discontinuation, and explicit disclosure of company-failure risk in consent
documents. None is required by regulators as of 2026, and the same exposure applies to every
implant discussed under [[neuroprosthetics]] and to the emerging commercial implants covered by
[[neurorights]] debates.

Cost compounds the exposure. Device, vitreoretinal surgery, and the months of rehabilitation needed
to learn to interpret phosphenes place these systems among the most expensive interventions per
patient in ophthalmology, for a functional gain that is real but small — the trade-off examined
under [[access-and-inequality]]. Parts of the blind community also reject the framing of blindness
as a defect awaiting correction, an argument developed at length under
[[disability-rights-and-enhancement]] and less prominent here than in deafness only because
prosthetic vision has never worked well enough to force the question.

## Alternatives and outlook

Prostheses compete with biological approaches that did not exist when the field began.
[[somatic-gene-therapy]] delivered by [[aav-vectors]] is approved for one inherited retinal
dystrophy caused by *RPE65* mutations, which slows or partially reverses loss in a small patient
population; it treats a cause rather than substituting for the sense. [[optogenetics]] offers a middle path — making surviving retinal
ganglion cells directly light-sensitive, then driving them with goggles that convert scenes into
patterned light. Partial visual function was reported in a single patient with retinitis pigmentosa
in 2021, and other opsin-based programmes have since been tested in early trials.[^sahel2021]
Photoreceptor and pigment-epithelium replacement derived from
[[induced-pluripotent-stem-cells]] has been attempted in small numbers of patients for macular
degeneration, and retinal organoids grown by the methods described under
[[tissue-engineering]] are a source of transplantable cells rather than a route to a working eye.

None of these will help someone whose retina is entirely gone or whose optic nerve is severed, which
is why cortical prostheses continue despite their difficulty. The honest near-term expectation is a
subretinal device that gives central-field pattern vision to people with geographic atrophy, a large
and growing population, alongside continued failure to deliver anything resembling normal sight. The
unresolved scientific question is whether the visual system can learn to interpret an artificial
code at all, or whether prosthetic vision is permanently limited to what a naive read-out of
phosphenes can convey — a question that [[neural-decoding]] research on the write side of the
interface has barely begun to address.

## See also

- [[cochlear-implant]]
- [[neuroprosthetics]]
- [[optogenetics]]
- [[brain-computer-interface]]
- [[sensory-augmentation]]
- [[utah-array]]
- [[somatic-gene-therapy]]
- [[disability-rights-and-enhancement]]

## References

[^palanker2020]: `paper` Palanker, D., Le Mer, Y., Mohand-Said, S., Muqit, M., Sahel, J.-A. "Photovoltaic restoration of central vision in atrophic age-related macular degeneration." *Ophthalmology*, 2020.

[^fernandez2021]: `paper` Fernández, E. et al. "Visual percepts evoked with an intracortical 96-channel microelectrode array inserted in human occipital cortex." *Journal of Clinical Investigation*, 2021. {A single blind volunteer, implanted for a limited period under a study protocol rather than as a therapy, with percepts identified after training.}

[^chen2020]: `paper` Chen, X., Wang, F., Fernandez, E., Roelfsema, P. R. "Shape perception via a high-channel-count neuroprosthesis in monkey visual cortex." *Science*, 2020.

[^spectrum2022]: `news` Strickland, E., Harris, M. "Their bionic eyes are now obsolete and unsupported." *IEEE Spectrum*, 2022. {Journalism built on interviews with implanted patients, and the main public record of what happened to them after Second Sight withdrew support.}

[^sahel2021]: `paper` Sahel, J.-A. et al. "Partial recovery of visual function in a blind patient after optogenetic therapy." *Nature Medicine*, 2021. {One patient, and the reported visual improvement occurred only with the stimulating goggles in place and after months of training.}

==============================================================================
ARTICLE: retro-biosciences
TITLE:   Retro Biosciences
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/retro-biosciences
SOURCE:  https://futurehumanwiki.com/raw/retro-biosciences
==============================================================================

---
title: "Retro Biosciences"
slug: "retro-biosciences"
type: "organization"
status: "emerging"
horizon: "2030s"
categories: ["organizations", "longevity"]
tags: ["aging", "autophagy", "reprogramming", "biotech", "protein engineering", "healthspan"]
summary: "A longevity biotechnology company funded by Sam Altman with the stated goal of adding ten years of healthy life, working on autophagy, reprogramming and plasma-inspired therapies."
updated: "2026-07-27"
issues: ["The 2025 report of a larger funding round is stated without a citation"]
---

```infobox
{
  "caption": "Private biotechnology company",
  "rows": [
    { "label": "Founded", "value": "2021" },
    { "label": "Chief executive", "value": "Joe Betts-LaCroix" },
    { "label": "Principal funder", "value": "Sam Altman" },
    { "label": "Initial capital", "value": "$180 million" },
    { "label": "Headquarters", "value": "Redwood City, California" },
    { "label": "Stated goal", "value": "Add 10 years of healthy life" },
    { "label": "Programmes", "value": "Autophagy, reprogramming, plasma" }
  ]
}
```

**Retro Biosciences** is a privately held longevity company founded in 2021 and funded initially with $180 million from Sam Altman, with the stated objective of adding ten years to healthy human lifespan. It runs three programmes — small molecules that enhance [[autophagy]], cellular [[partial-reprogramming|reprogramming]], and therapies derived from the [[parabiosis-and-young-blood|young-plasma]] literature — and is more specific in public than its better-funded peers about programmes, goals and timelines. An early collaboration with OpenAI on protein design has drawn most of the attention to it.

## Overview

The company's framing is deliberately narrower than the field's usual rhetoric. Rather than treating aging as a single process to be reversed, it selected three mechanisms with existing preclinical support and organized around getting each into humans. Joe Betts-LaCroix, a hardware engineer turned entrepreneur who previously co-founded a computing company and had been involved in longevity advocacy, is chief executive.

Retro is smaller than [[altos-labs]] by an order of magnitude in capital, and its public posture is correspondingly more concrete: it names its programmes, states a numerical goal, and has discussed clinical timelines, none of which its larger competitor does.

## History

The company was formed in 2021 and operated quietly until March 2023, when MIT Technology Review reported that Sam Altman, then chief executive of OpenAI, had personally invested $180 million — the whole of the company's funding at that point.[^regalado2023] Altman has described the investment as motivated by an interest in extending healthy life rather than by a commercial thesis, and holds no operational role.

In January 2025 the same publication reported a collaboration between Retro and OpenAI on a model called GPT-4b micro, trained to propose redesigned versions of proteins in the manner described under [[ai-protein-design]].[^regalado2025] The reported application was engineering variants of the [[yamanaka-factors]] — specifically SOX2 and KLF4 — to improve reprogramming efficiency, with Retro claiming a large increase in the expression of markers associated with pluripotency induction. The result was announced through press coverage rather than a peer-reviewed paper, and independent replication has not been published.

```timeline
[
  { "year": "2021", "title": "Founded", "text": "Retro Biosciences is established in the Bay Area with three programme areas and a stated ten-year healthspan goal." },
  { "year": "2023", "title": "Funding disclosed", "text": "MIT Technology Review reports Sam Altman's $180 million personal investment, at the time the company's entire capital." },
  { "year": "2025", "title": "Protein design collaboration", "text": "Retro and OpenAI report a purpose-built model proposing improved Yamanaka factor variants, announced via press rather than publication." },
  { "year": "2025–2026", "title": "Toward the clinic", "text": "The company moves its lead autophagy-enhancing compound toward first-in-human testing and reports seeking substantially larger funding." }
]
```

## Research programme

### Autophagy enhancement

The most developed programme targets autophagy, the lysosomal degradation pathway whose molecular machinery Yoshinori Ohsumi worked out in yeast and for which he received the 2016 Nobel Prize in Physiology or Medicine.[^ohsumi2016] Autophagic flux declines with age in most tissues studied, and genetic or pharmacological enhancement extends lifespan in flies, worms and mice; nearly every intervention that reliably extends lifespan in model organisms, including [[caloric-restriction]] and [[rapamycin]], appears to require an intact autophagy pathway, which is why it recurs across the [[hallmarks-of-aging]] as a shared downstream node. Retro's approach is a small molecule intended to raise flux without the immunosuppression that accompanies mTOR inhibition, with an initial indication in neurodegeneration, where failing [[proteostasis]] is a well-established feature of disease.

The gap between the model-organism evidence and human benefit is wide and the company acknowledges it. No autophagy-enhancing drug has demonstrated a clinical benefit in an aging-related indication in humans.

### Cellular reprogramming

The reprogramming programme pursues the same goal as [[altos-labs]] and [[newlimit]]: restoring youthful cell function without loss of cell identity. It differs from both in route. Retro works on the reprogramming factors themselves, engineering variants of the canonical set, which is where the OpenAI collaboration was applied, while NewLimit treats the canonical factors as the wrong tool and screens for combinations that never approach pluripotency. Neither route removes the delivery problem.

### Plasma-derived therapeutics

The third programme derives from heterochronic parabiosis, in which young and old mice share a circulation and the old animal shows improvements in several tissues. The interpretation of those experiments is contested — whether the effect comes from youthful factors added or from aged factors diluted — and the plasma-transfusion clinics that grew up around them drew a public warning from the US Food and Drug Administration in 2019. Retro's stated approach is to identify specific circulating factors rather than to transfuse plasma, which sidesteps the strongest objection to the commercial versions.

> [!caution] Mice, not people
> Every mechanism Retro works on rests on rodent evidence. Autophagy enhancement, partial reprogramming and heterochronic blood exchange have all produced functional improvements in mice. None has been shown to extend healthy life in humans, and the company's ten-year target is a goal, not a projection supported by data.

## Funding and structure

Altman's $180 million was, at the time it was reported, one of the largest single-individual investments in a longevity company. Reports in 2025 indicated the company was raising a substantially larger round, and Retro has said it intends to fund clinical development rather than remain a discovery organization.

The funding structure raises a governance question the company has addressed only briefly. Altman was simultaneously the chief executive of OpenAI and the sole funder of a company that entered a collaboration with OpenAI, an arrangement that would attract scrutiny in a public company. Retro's position is that the collaboration was conducted on ordinary commercial terms.

## Reception

The scientific reception has been more favourable than that given to most well-publicized longevity ventures, largely because the company works on established mechanisms rather than novel claims and has been explicit about the preclinical state of the evidence. Specialists who are skeptical of the field's rhetoric have noted that autophagy enhancement is a reasonable target with a plausible path to an approvable indication.

Criticism has focused on three points. The ten-years-of-healthy-life goal has no operational definition; [[healthspan]] is not a standardized measure, no qualified [[aging-biomarkers|surrogate endpoint]] exists, and no trial design has been proposed that would demonstrate the claim within the lifetime of the company. The GPT-4b micro result was communicated through journalism rather than publication, which is the pattern the field has been criticized for since the resveratrol episode associated with [[david-sinclair]]. And a company organized around a specific numerical promise faces an incentive to define success downward, which the absence of a preregistered endpoint makes easy.

## Outlook

Retro's near-term significance is that it may be first among the well-funded reprogramming and geroprotection companies to generate human data. A phase-one safety result for an autophagy enhancer, followed by a phase-two efficacy signal in a neurodegenerative indication, would be the first time any of this generation of longevity companies has produced evidence a regulator recognizes. A null result would be equally informative, and considerably more likely on base rates for central nervous system drugs.

The AI-assisted protein engineering strand is the more consequential unknown. If purpose-trained models can reliably propose transcription-factor variants with improved function, that changes the cost structure of protein engineering across the field, well beyond [[epigenetic-reprogramming]]. The claim as reported is large enough that the absence of a peer-reviewed description more than a year later is itself a data point.

## See also

- [[autophagy]]
- [[altos-labs]]
- [[newlimit]]
- [[partial-reprogramming]]
- [[parabiosis-and-young-blood]]
- [[healthspan]]
- [[geroscience-hypothesis]]
- [[rapamycin]]

## References

[^regalado2023]: `news` Regalado, A. "Sam Altman invested $180 million into a company trying to delay death." *MIT Technology Review*, March 2023. {Trade-press reporting of a private financing; the $180 million figure is as reported by the magazine, not an audited disclosure.}
[^regalado2025]: `news` Regalado, A. "OpenAI has created an AI model for longevity science." *MIT Technology Review*, January 2025.
[^ohsumi2016]: `statement` The Nobel Assembly at Karolinska Institutet. "The Nobel Prize in Physiology or Medicine 2016: Yoshinori Ohsumi." Press release, 2016. {A prize citation for the yeast genetics of autophagy; it is not evidence about autophagy enhancement as a therapy.}

==============================================================================
ARTICLE: right-to-die-and-right-to-live
TITLE:   Right to die and the duty to live
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/right-to-die-and-right-to-live
SOURCE:  https://futurehumanwiki.com/raw/right-to-die-and-right-to-live
==============================================================================

---
title: "Right to die and the duty to live"
slug: "right-to-die-and-right-to-live"
type: "concept"
status: "contested"
horizon: "present"
categories: ["society", "longevity"]
tags: ["bioethics", "law", "assisted dying", "autonomy", "aging", "mortality"]
summary: "How the law and ethics of assisted dying interact with radical life extension, and whether an indefinitely extendable life could still be voluntarily ended."
updated: "2026-07-27"
issues: ["The 1984 governor's duty-to-die remark is quoted without naming him or citing a source", "The early-1990s California cryonics case is described without naming it or citing the ruling"]
---

```infobox
{
  "caption": "Problem in bioethics and law",
  "rows": [
    { "label": "First modern statute", "value": "Oregon, in force 1997" },
    { "label": "Euthanasia legalised", "value": "Netherlands and Belgium, 2002" },
    { "label": "Canadian regime", "value": "MAID, from 2016" },
    { "label": "Core legal principle", "value": "Refusal of treatment" },
    { "label": "Contested extension", "value": "Non-terminal eligibility" },
    { "label": "Longevity implication", "value": "Untested" }
  ]
}
```

**Right to die and the duty to live** names the intersection of two bodies of argument that developed separately and are beginning to collide: the law and ethics of assisted dying, built around people whose deaths are near and unwanted, and the prospect of interventions that make continued life a standing option rather than a fixed allotment. If aging becomes treatable, refusing treatment becomes a choice, and every question about whether death may be chosen acquires a mirror image about whether life may be declined.

## Two doctrines, separately built

Modern end-of-life law rests on a distinction that is stable in most jurisdictions and philosophically contested. A competent person may refuse any medical treatment, including life-sustaining treatment, and this right is close to absolute. Whether anyone may assist them to die actively is a separate question that most legal systems answer differently.

The US Supreme Court set out both halves in the 1990s: a 1990 case established a constitutionally protected liberty interest in refusing unwanted medical treatment, and two 1997 decisions held that there is no corresponding constitutional right to assistance in suicide, leaving the question to the states. Oregon's statute, in force from 1997, permits a physician to prescribe lethal medication to a terminally ill adult who self-administers it; roughly a dozen US jurisdictions now have comparable laws.

The Netherlands and Belgium legalised physician-administered euthanasia in 2002 under conditions of unbearable suffering without prospect of improvement, which need not be terminal. Switzerland has permitted assisted suicide since long before the modern debate, through a penal-code provision that criminalises assistance only when the motive is selfish — an arrangement that made it the destination for cross-border cases. Canada's regime began with a 2015 constitutional ruling, was legislated in 2016 for those whose death was reasonably foreseeable, and was extended in 2021 to people with grievous and irremediable conditions who are not dying; eligibility on the basis of mental illness alone has been postponed repeatedly. The United Kingdom's Terminally Ill Adults (End of Life) Bill passed the House of Commons in 2025 and was under consideration in the Lords; this is a fast-moving area and the position may have changed since.

> [!note] What "right to die" usually means
> In almost all existing law it means a right against interference — to refuse treatment, or to be assisted without the assistant being prosecuted. It is not a right to be provided with death on request, and no jurisdiction treats it as one.

## The suffering criterion and its instability

Every permissive regime rests on a criterion: terminal illness, unbearable suffering, or a grievous and irremediable condition. Each has proved harder to hold than expected.

Terminal illness requires a prognosis, and prognosis is unreliable at the margins. Unbearable suffering is subjective by construction, which is why the Dutch regime relies on physician assessment rather than an objective test. And any criterion defined by the absence of a remedy is unstable in a period of medical advance: a condition that is irremediable in 2026 may not be in 2036, which means the same person's eligibility changes without their condition changing.

This is the first point of contact with longevity research. If the frailty and multimorbidity that drive many requests are treatable — the claim behind the [[geroscience-hypothesis]] — then a person requesting assistance because their body is failing is requesting it for a remediable condition. Whether that should affect eligibility is a question no jurisdiction has faced. It becomes acute only if interventions extend [[healthspan]] rather than lifespan alone; a therapy that adds years of dependency would strengthen the case for assisted dying rather than weaken it.

## The duty-to-die worry

Concerns that a permission would become an expectation predate the technology. A US state governor's 1984 remark that the terminally ill elderly have "a duty to die and get out of the way" prompted decades of argument, and the philosopher John Hardwig later defended a limited version — that a person may have obligations to family that bear on how long they seek to live.[^hardwig1997] Daniel Callahan had earlier argued for age-based limits on publicly funded life-extending care, on the ground that a health system cannot both pursue indefinite extension and meet other needs.[^callahan1987]

The empirical worry is about pressure rather than duty. Disability-rights organisations and, in 2021, several UN human-rights experts raised concerns that Canada's extension to non-terminal conditions could make assisted death a response to inadequate housing, care, or disability support — that where the state funds death more reliably than it funds living, the choice is not free. Reported Canadian cases in which applicants cited unavailable services rather than intractable symptoms sharpened the objection. [[disability-rights-and-enhancement]] treats the broader argument about which lives medicine implicitly ranks.

## The duty-to-live worry

The symmetrical concern has had almost no attention, because the technology that would create it does not exist. It arises if life extension becomes routine.

Where continued life requires ongoing intervention — periodic [[senolytics]], repeated [[epigenetic-reprogramming]], or whatever a working geroprotective regimen turns out to be — declining that intervention becomes an act rather than an omission. The distinction between letting die and killing, which carries enormous legal weight, becomes hard to locate. Someone who stops taking a rejuvenation therapy at 130 has not refused an extraordinary measure in the sense the doctrine contemplates; they have declined the ordinary maintenance that everyone around them undertakes.

Institutional pressures compound this. Pension systems, insurers, and employers all have financial interests in when people die, and those interests would run in the opposite direction from the ones that worry critics of assisted dying today. A society in which retirement is indefinitely deferred because working life is indefinitely extended, examined in [[longevity-dividend]], is one in which stopping is a decision with economic consequences for others.

> [!debate] Is an indefinite life voluntary
> Proponents argue that life extension adds an option and removes none: anyone may still decline. Critics reply that options with strong social defaults are not neutral, and that the pressure not to burden others, which currently pushes toward accepting death, would push toward accepting treatment with equal force and less scrutiny.

## Whether an unending life would be worth having

The philosophical literature on this question is older than the biology. Bernard Williams argued from the Makropulos case that an unending life would eventually exhaust the categorical desires that give a person reason to go on, leaving tedium rather than fulfilment.[^williams1973] John Martin Fischer's reply is that repeatable pleasures do not exhaust, that a sufficiently long life allows new projects, and that Williams's argument assumes a fixed character which the immortal person need not have.[^fischer1994]

Leon Kass makes a stronger claim: that mortality is not merely compatible with meaning but constitutive of it, because urgency, seriousness, and the structure of a life course depend on its finitude. It is the most substantive bioconservative argument and is treated in [[leon-kass]] and [[bioconservatism]]. Nick Bostrom's *Fable of the Dragon-Tyrant* is the standard reply, arguing that such reasoning is a rationalisation of an unavoidable evil that persists after the evil becomes avoidable.[^bostrom2005] Neither position is empirically testable, and the oldest people who have ever lived, discussed in [[maximum-human-lifespan]], have not lived long enough to bear on it.

## Cryonics and the timing of death

[[cryonics]] creates a live legal version of the problem. Preservation quality depends on beginning quickly after circulation stops, and legal death is required before any procedure can begin — so the process starts at the worst possible moment for the outcome its subscribers want. A California case in the early 1990s tested this directly: a man with a terminal brain tumour sought a court order permitting cryopreservation before his brain deteriorated, and was refused. Organisations such as [[alcor]] therefore operate standby procedures designed to compress the interval, without ever crossing the line the law draws.

Jurisdictions that permit assisted dying create a route that circumvents this, and a small number of cases have reportedly combined the two. It is the only context in which the right to die and the pursuit of indefinite life are not opposed but operationally connected, and neither the assisted-dying literature nor the [[brain-preservation]] literature has addressed what standards should apply.

## Open problems

No legal system has a category for a person who declines an available and effective life-extension therapy, and it is not obvious whether existing refusal-of-treatment doctrine would cover them or whether the case would be treated as suicide. Nor has any jurisdiction considered how eligibility criteria framed around irremediable conditions should be interpreted when the irremediability is a fact about current medicine rather than about the patient. Both questions are currently hypothetical. They stop being hypothetical the first time a geroprotective intervention works.

## See also

- [[cryonics]]
- [[longevity-escape-velocity]]
- [[leon-kass]]
- [[bioconservatism]]
- [[compression-of-morbidity]]
- [[disability-rights-and-enhancement]]
- [[overpopulation-and-longevity]]
- [[personal-identity-and-continuity]]

## References

[^hardwig1997]: `paper` Hardwig, J. "Is There a Duty to Die?" *Hastings Center Report*, 1997.
[^callahan1987]: `book` Callahan, D. *Setting Limits: Medical Goals in an Aging Society*. Simon and Schuster, 1987.
[^williams1973]: `book` Williams, B. "The Makropulos Case: Reflections on the Tedium of Immortality." In *Problems of the Self*, Cambridge University Press, 1973.
[^fischer1994]: `paper` Fischer, J.M. "Why Immortality Is Not So Bad." *International Journal of Philosophical Studies*, 1994.
[^bostrom2005]: `paper` Bostrom, N. "The Fable of the Dragon-Tyrant." *Journal of Medical Ethics*, 2005.

==============================================================================
ARTICLE: robert-freitas
TITLE:   Robert Freitas
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/robert-freitas
SOURCE:  https://futurehumanwiki.com/raw/robert-freitas
==============================================================================

---
title: "Robert Freitas"
slug: "robert-freitas"
type: "person"
status: "contested"
horizon: "present"
categories: ["people", "nanomedicine"]
tags: ["nanotechnology", "nanomedicine", "nanorobots", "molecular manufacturing", "self-replication", "cryonics"]
summary: "American researcher whose Nanomedicine volumes set out detailed numerical designs for medical nanorobots, among them the respirocyte, none of which has been fabricated."
updated: "2026-07-28"
---

```infobox
{
  "caption": "Engineer and theorist",
  "rows": [
    { "label": "Born", "value": "1952, Camden, Maine" },
    { "label": "Nationality", "value": "American" },
    { "label": "Education", "value": "Harvey Mudd College; Santa Clara University" },
    { "label": "Known for", "value": "Respirocyte design study", "link": "/wiki/respirocytes" },
    { "label": "Key work", "value": "Nanomedicine (1999, 2003)" },
    { "label": "Field", "value": "Molecular nanotechnology; nanomedicine" },
    { "label": "Affiliation", "value": "Institute for Molecular Manufacturing" },
    { "label": "Award", "value": "Feynman Prize in Nanotechnology, 2009" }
  ]
}
```

**Robert Freitas** is an American researcher who has spent three decades producing detailed engineering
analyses of machines that do not exist. His four-volume *Nanomedicine*, of which two volumes have
appeared, works through the physics of micron-scale devices intended to operate inside the human body:
power, heat, navigation, sensing, communication, and what happens when a rigid manufactured object meets
blood. The best known of the designs is the [[respirocytes|respirocyte]], an artificial red cell. None
of the devices has been fabricated, and Freitas has been consistent that his work is not a prediction
that any of them will be.

## Overview

Freitas works in a genre he calls exploratory engineering: analysis intended to establish that a
proposed device violates no physical law and to bound its performance if built, deliberately setting
aside how anyone would build it. The output is a specification and a set of numbers rather than an
experiment. Judged as engineering analysis the volumes are unusually thorough; judged as evidence that
[[medical-nanorobots]] are coming, they establish nothing, because a design consistent with physics is
not the same as a design that can be made. Most of the argument about his work is an argument about
which of those two things it is being read as.

## Career

Freitas took degrees in physics and psychology at Harvey Mudd College and a law degree at Santa Clara
University. His early work was in exobiology rather than medicine: a long treatise on the scientific
study of extraterrestrial life and civilization, written in the 1970s, and a proposed scale ranking
information-processing systems by their rate of computation per unit mass. With William Gilbreath he co-edited
the report of the 1980 NASA/ASEE summer study on advanced automation for space missions, whose most
discussed section is a design for a self-replicating lunar factory.[^nasa1982]

He began the *Nanomedicine* series in the mid-1990s at the Institute for Molecular Manufacturing in
Palo Alto, where he is a senior research fellow, and spent the early 2000s as a research scientist at
Zyvex, a Texas company then attempting instrument-based approaches to atomically precise fabrication.

## The Nanomedicine volumes

Volume I sets out basic capabilities: how a micron-scale device would draw power from blood glucose,
shed waste heat without cooking the tissue around it, know where it is, and talk to its
neighbours.[^freitas1999] Volume IIA is devoted entirely to biocompatibility — what the immune system,
the complement cascade and the protein corona that coats anything injected into blood would do to a rigid
diamondoid object, and what surface chemistry might reduce it.[^freitas2003] Volumes IIB and III have
been announced for many years and have not appeared.

The device studies are the part of the work that circulates. The 1998 respirocyte paper describes a
pressurised gas vessel about a micrometre across that would carry far more oxygen per unit volume than
haemoglobin does, competing with the oxygen carriers whose repeated clinical failures are described
under [[artificial-blood]].[^freitas1998] The microbivore is an artificial phagocyte that would trap a
bacterium, digest it to amino acids and sugars, and discharge the residue, on a cycle timed in
seconds.[^freitas2005] The clottocyte would stop bleeding faster than platelets; the chromallocyte would
enter a cell and replace its chromosomes with corrected copies.

> [!note] What the numbers are
> Figures such as a respirocyte's oxygen-carrying advantage over a red cell are outputs of a physical
> model of a specified object. They are neither measurements nor forecasts, and they carry no
> information about when or whether the object could be made. The distinction is stated plainly in the
> papers themselves and lost in most secondary accounts of them.

## Self-replication and global ecophagy

Freitas has also worked on the theory of machines that copy themselves, a thread running from the 1980
NASA study to the book-length review he wrote with Ralph Merkle in 2004, which catalogues the physical
and mathematical literature on kinematic self-replication and proposes a classification for
it.[^ksrm2004]

That interest produced the most-cited thing he has written outside the medical designs. His 2000
analysis of global ecophagy asked what would physically limit a runaway population of biomass-consuming
replicators — the [[grey-goo]] scenario — and concluded that waste heat binds before materials do:
converting the biosphere at the rates the scenario requires would release enough energy to destroy the
machinery doing it, and would produce a thermal signature detectable long beforehand. He treated this as
grounds for monitoring and for design restrictions on replicating systems rather than for
dismissal.[^freitas2000] It is a rare case of a technology's proponent producing the quantitative risk
analysis, and it is cited in the [[existential-risk]] literature by people who reject the rest of the
programme.

## Diamond mechanosynthesis

Everything in the volumes assumes a manufacturing capability that does not exist. The devices are
specified in stiff diamondoid components with atomically precise surfaces, which only the
[[molecular-assembler|positional chemistry]] proposed by [[eric-drexler]] could produce. Freitas's
response has been to work on the smallest tractable piece of that problem: mechanosynthesis of diamond,
in which a positionally controlled tip places carbon atoms on a diamond surface one or two at a time.

With Merkle he has published computational studies of candidate tooltips and reaction sequences,
evaluated with quantum-chemistry methods, and the two have coordinated a loose collaboration aimed at a
first experimental demonstration. A patent covering a tool for positional diamond mechanosynthesis
issued in 2010, and the 2009 Feynman Prize for theory recognized the tooltip work. No experimental
demonstration of controlled diamond mechanosynthesis has been reported, which is the same gap that has
stood open since Richard Smalley's objections to Drexler in the early 2000s.

## Cryonics

Freitas has argued at length that cell-by-cell repair of cryopreserved tissue is a well-posed
engineering problem, publishing a book-length treatment through the Alcor Life Extension Foundation in
2022 that works through failure modes in preserved tissue and the operations a repair device would have
to perform.[^cryostasis2022] It is the fullest statement of the biological revival route described under
[[cryonics]], and it inherits that route's dependence on machinery nobody has built. That dependence is
part of why others in the field prefer the informational route, in which the preserved brain is scanned
rather than repaired: the programme pursued under [[brain-preservation]] and
[[whole-brain-emulation]]. [[alcor|Alcor]]'s own literature endorses neither route.

> [!debate] Whether the analysis can be checked
> Supporters hold that a quantitative design that survives physical scrutiny is a stronger claim than
> an assertion that revival is impossible, and that the burden has not been met on the other side.
> Critics hold that an analysis of surfaces, tissues and failure modes that have never been produced
> cannot be checked at all, and that unfalsifiable engineering is not engineering.

## Reception and legacy

Chemists and clinical researchers have largely not engaged with the programme. The medical successes at
the nanoscale arrived by other routes entirely: [[lipid-nanoparticles]] and antibody conjugates from
formulation chemistry and immunology, [[dna-nanotechnology|DNA origami]] devices with no onboard power or
computation, [[microrobots-in-medicine|magnetically steered particles]] that are steered from outside
rather than autonomous. Freitas is cited far more in futurist, transhumanist and cryonics writing than
in the clinical [[targeted-drug-delivery]] literature, and a substantial part of his output
appeared in books from a small press, in institute reports, or in journals associated with the movement
rather than in mainstream biomedical venues.

The case for the work is that it is specific enough to be wrong. It names dimensions, materials, power
budgets and cycle times, which is more than most speculation about medicine's future offers, and the
biocompatibility volume takes the strongest objections seriously instead of waving at them. The case
against is that specificity about an object nobody can make is not the same as evidence, and that
thirty years of increasingly detailed designs with no fabrication attempt is the signature of a research
programme that has substituted analysis for experiment. Both readings survive the record as it stands,
which is why the argument has not moved much since *Nanomedicine* Volume I appeared.

## See also

- [[respirocytes]]
- [[medical-nanorobots]]
- [[molecular-assembler]]
- [[eric-drexler]]
- [[grey-goo]]
- [[cryonics]]
- [[dna-nanotechnology]]
- [[artificial-blood]]

## References

[^nasa1982]: `report` Freitas, R. A. and Gilbreath, W. P. (eds.) *Advanced Automation for Space Missions: Proceedings of the 1980 NASA/ASEE Summer Study*. NASA Conference Publication 2255, 1982.
[^freitas1998]: `paper` Freitas, R. A. "Exploratory Design in Medical Nanotechnology: A Mechanical Artificial Red Cell." *Artificial Cells, Blood Substitutes, and Immobilization Biotechnology*, 1998. {A design paper: performance is derived from physical models, and no device, experiment or measurement is reported.}
[^freitas1999]: `book` Freitas, R. A. *Nanomedicine, Volume I: Basic Capabilities*. Landes Bioscience, 1999.
[^freitas2003]: `book` Freitas, R. A. *Nanomedicine, Volume IIA: Biocompatibility*. Landes Bioscience, 2003. {Analyses the immune response to surfaces that have never been synthesised, so its conclusions cannot be tested against anything.}
[^freitas2005]: `paper` Freitas, R. A. "Microbivores: Artificial Mechanical Phagocytes using Digest and Discharge Protocol." *Journal of Evolution and Technology*, 2005. {The journal is published by the Institute for Ethics and Emerging Technologies, a body advocating for the technologies the paper describes.}
[^freitas2000]: `report` Freitas, R. A. "Some Limits to Global Ecophagy by Biovorous Nanoreplicators, with Public Policy Recommendations." Institute for Molecular Manufacturing, 2000. {Written from inside the molecular manufacturing community; it assumes assemblers are feasible and asks only what would limit them.}
[^ksrm2004]: `book` Freitas, R. A. and Merkle, R. C. *Kinematic Self-Replicating Machines*. Landes Bioscience, 2004.
[^cryostasis2022]: `book` Freitas, R. A. *Cryostasis Revival: The Recovery of Cryonics Patients through Nanomedicine*. Alcor Life Extension Foundation, 2022. {Published by a cryonics provider, so it is a proponent's technical case rather than an independent assessment.}

==============================================================================
ARTICLE: senolytics
TITLE:   Senolytics
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/senolytics
SOURCE:  https://futurehumanwiki.com/raw/senolytics
==============================================================================

---
title: "Senolytics"
slug: "senolytics"
type: "intervention"
status: "experimental"
horizon: "2030s"
trl: 5
categories: ["longevity"]
tags: ["cellular senescence", "aging", "sasp", "clinical trials", "drug repurposing", "apoptosis"]
summary: "Drugs that selectively kill senescent cells, an approach that reduces age-related pathology in mice but has so far produced modest or null results in humans."
updated: "2026-07-27"
humanEvidence: "Human data are small open-label pilots of nine to fourteen patients, one showing reduced senescent-cell markers in fat and skin, plus two randomized failures; every lifespan result is from mice."
access: "No senolytic is approved for any aging indication; dasatinib is an approved cancer drug available only on prescription, and fisetin and quercetin are sold over the counter as supplements at doses taken from mouse studies."
reversibility: "partly-reversible"
issues: ["The Alzheimer's pilot and the cardiac-glycoside senolytic reports carry no citations"]
---

```infobox
{
  "caption": "Class of experimental drugs",
  "rows": [
    { "label": "Target", "value": "Senescent cells" },
    { "label": "Mechanism", "value": "Blocking anti-apoptotic pathways" },
    { "label": "Term coined", "value": "2015" },
    { "label": "Lead candidates", "value": "Dasatinib + quercetin, fisetin, navitoclax" },
    { "label": "Clinical stage", "value": "Phase 1–2" },
    { "label": "Approved for aging", "value": "No" },
    { "label": "Readiness", "value": "TRL 5" }
  ]
}
```

**Senolytics** are drugs that kill [[cellular-senescence|senescent cells]] selectively while leaving
dividing and quiescent cells intact. The class was proposed in 2015 on the reasoning that senescent
cells survive despite extensive damage because they upregulate a small number of anti-apoptotic
pathways, and that transiently blocking those pathways would tip them into cell death.[^zhu2015] In
mice, clearing senescent cells improves function in a long list of tissues and has been reported to
extend lifespan. In humans the completed trials have been small, mostly uncontrolled, and their
results have been modest or negative.

```keyfacts
[
  { "value": "2015", "label": "Year the drug class was named", "note": "Zhu et al., Aging Cell" },
  { "value": "~25%", "label": "Median lifespan gain in mice", "note": "genetic clearance of p16-positive cells, Baker et al. 2016" },
  { "value": "0", "label": "Senolytics approved for any indication", "note": "as of mid-2026" }
]
```

## How they work

Senescent cells have withdrawn stably from the cell cycle but remain metabolically active. They
enlarge, resist apoptosis, and secrete a mixture of cytokines, chemokines, proteases, and growth
factors known as the senescence-associated secretory phenotype (SASP), characterized in the 2000s by
groups including those at the [[buck-institute]]. The SASP is what makes senescent cells plausible
drivers of aging rather than inert debris, and its paracrine spread links the process to
[[inflammaging]] and to [[stem-cell-exhaustion]]. Cells reach the state by several routes —
[[telomeres-and-telomerase|telomere attrition]] in replicative senescence, DNA damage, oncogene
activation, severe [[mitochondrial-dysfunction|mitochondrial stress]] — which is one reason the
resulting populations are not biochemically uniform. [[cellular-senescence]] treats the cell state
itself, including its triggers, markers, and useful roles; this article treats the drugs. Senescence
appears in its own right as one of the [[hallmarks-of-aging]].

Resistance to apoptosis is the drugging opportunity. Senescent cells depend on what the field calls
senescent cell anti-apoptotic pathways: members of the BCL-2 family, PI3K–AKT signalling, p53–p21
regulation, HIF-1α, and heat-shock protein 90. Crucially, different senescent cell types depend on
different pathways. Senescent human preadipocytes are killed by dasatinib, an approved kinase
inhibitor; senescent endothelial cells are killed by the flavonoid quercetin. Neither alone covers
both, which is why the founding paper proposed the combination.

The pharmacology has an unusual feature. Because senescent cells do not divide, a drug does not need
to maintain steady-state exposure — it needs only to be present long enough to trigger apoptosis, and
the cleared population takes weeks to months to rebuild. This licenses "hit-and-run" intermittent
dosing, typically a few days every few weeks, which in principle limits cumulative toxicity from
drugs that would be intolerable continuously.

> [!note] Terminology
> A *senolytic* kills the cell. A *senomorphic* (or senostatic) leaves it alive but suppresses the
> SASP; [[rapamycin]], [[metformin]], and JAK inhibitors have all been used this way. The two
> strategies have different risk profiles and are frequently conflated in press coverage.

## Development history

```timeline
[
  { "year": "1961", "title": "Replicative senescence described", "text": "Hayflick and Moorhead report that normal human fibroblasts stop dividing after a finite number of passages, a phenomenon later tied to telomere shortening." },
  { "year": "2011", "title": "Genetic clearance in progeroid mice", "text": "Baker and colleagues use the INK-ATTAC transgene to kill p16-positive cells on demand, delaying cataract, sarcopenia, and fat loss in a fast-aging mouse strain." },
  { "year": "2015", "title": "First senolytic drugs identified", "text": "Zhu et al. use a transcriptomic screen for anti-apoptotic dependencies to nominate dasatinib and quercetin, and coin the term senolytic." },
  { "year": "2016", "title": "Lifespan extension in normally aged mice", "text": "Clearing p16-positive cells from genetically engineered mice raises median lifespan by roughly a quarter and delays several age-related pathologies." },
  { "year": "2018", "title": "Senescent cells shown to be sufficient", "text": "Transplanting a small number of senescent cells into young mice causes persistent physical dysfunction, which senolytic treatment partially reverses." },
  { "year": "2019", "title": "First human pilots reported", "text": "Open-label studies in idiopathic pulmonary fibrosis and diabetic kidney disease report feasibility and, in the kidney study, a measurable drop in senescent-cell burden in adipose tissue." },
  { "year": "2020", "title": "First randomized failure", "text": "Unity Biotechnology's UBX0101, injected into osteoarthritic knees, performs no better than placebo in a phase 2 trial." },
  { "year": "2020", "title": "Senolytic CAR T cells", "text": "Amor and colleagues engineer T cells against the surface protein uPAR, showing that senescent cells can be cleared immunologically rather than pharmacologically." },
  { "year": "2024", "title": "Randomized bone trial misses endpoint", "text": "A controlled trial of intermittent dasatinib plus quercetin in postmenopausal women does not meet its primary bone-turnover endpoint." }
]
```

The idea that accumulated "death-resistant cells" constitute a distinct, separately treatable form of
age-related damage predates the drugs. It appears as one of the seven categories in the damage-repair
framework that [[aubrey-de-grey]] and the [[sens-research-foundation]] promoted from the early 2000s,
at a time when the mainstream view still treated senescence mainly as a cell-culture artefact. The
2011 and 2016 mouse clearance experiments moved the question from argument to experiment, which is
why they are cited far outside biogerontology.

## Drug candidates

**Dasatinib plus quercetin (D+Q)** remains the most-studied combination and the one used in nearly
every human pilot. Dasatinib is a tyrosine kinase inhibitor approved for chronic myeloid leukaemia,
with a real oncology toxicity profile including cytopenias, bleeding risk, and pleural effusion.
Quercetin is a widely sold plant flavonoid with poor and variable oral bioavailability, which
complicates any inference about dose.

**Fisetin**, another flavonoid, showed senolytic activity and healthspan effects in aged mice and has
attracted disproportionate consumer interest because it is sold as a
[[dietary-supplements|supplement]]. Its human evidence
base is thin, and the bioavailability problem is worse than quercetin's.

**Navitoclax (ABT-263)** inhibits BCL-2, BCL-xL, and BCL-W, and clears senescent cells efficiently in
mice, including in aged haematopoietic stem cells.[^chang2016] Its dose-limiting thrombocytopenia
follows directly from the mechanism: platelets depend on BCL-xL for survival. Efforts to separate the
effects include BCL-xL-selective degraders built as PROTACs, which are designed to spare platelets
because platelets lack the machinery to execute targeted protein degradation.

**Unity Biotechnology** pursued locally injected senolytics to avoid systemic exposure. UBX0101, an
MDM2–p53 interaction inhibitor for knee osteoarthritis, failed its randomized phase 2. The company's
subsequent lead, UBX1325 (foselutoclax), a BCL-xL inhibitor injected into the eye for diabetic
macular edema, has produced phase 2 data the company describes as favourable; an independent
confirmatory trial had not been published as of 2026.

Beyond small molecules, cardiac glycosides such as ouabain and digoxin have been reported to have
broad senolytic activity, and immunological approaches (senolytic CAR T cells directed at
uPAR,[^amor2020] and vaccines raised against senescence-associated surface antigens) offer the
prospect of a single treatment with lasting effect rather than repeated dosing. Researchers have also
explored [[gene-therapy-for-aging|gene-delivered constructs]] that place a killing gene under the
control of a senescence-responsive promoter, which would reproduce in a treatable animal what the
INK-ATTAC transgene achieved by design.

## Human evidence

The human record is short and must be read carefully. The idiopathic pulmonary fibrosis pilot
enrolled fourteen patients with no control group and reported improvements in walking-based physical
function measures, which in an open-label study of a symptomatic population cannot be separated from
expectation and practice effects.[^justice2019] The diabetic kidney disease study, with nine
participants, is more informative on mechanism than on benefit: adipose and skin biopsies showed
reduced senescent-cell markers days after dosing, establishing target engagement in at least one
tissue.[^hickson2019] A pilot in Alzheimer's disease established that the drugs are tolerated in that
population and that dasatinib reaches cerebrospinal fluid, without addressing efficacy.

The most rigorous test to date is a randomized trial of intermittent D+Q in postmenopausal women with
bone-turnover endpoints, which did not meet its primary endpoint; the authors reported signals in
prespecified subgroups, which is hypothesis-generating rather than confirmatory.[^farr2024] Combined
with the UBX0101 failure, the pattern is a field whose animal data are strong and whose human data
are, so far, unpersuasive.

> [!caution] The missing measurement
> No validated assay reports senescent-cell burden in a living human across tissues. Without one,
> trials cannot confirm that a drug engaged its target in the organ of interest, dose selection is
> guesswork, and a null result cannot be distinguished from an underdosed one. This is the same
> surrogate-endpoint gap that constrains the whole of [[aging-biomarkers|geroscience measurement]].

## Limitations and unresolved biology

Senescence is not simply damage: it suppresses tumour formation, participates in wound healing and
tissue remodelling, and has a programmed role in embryonic development, all set out under
[[cellular-senescence]]. Indiscriminate clearance is therefore not obviously safe. In mice, deleting
p16-high cells damaged liver sinusoidal endothelial cells and perivascular cells, producing fibrosis
and impaired clearance from the blood.[^grosse2020]

The marker problem runs deeper than assay development. There is no single molecular signature of
senescence, and the cells that carry the usual markers differ by tissue, inducing stimulus, and time
since induction, so a drug that kills one senescent population may leave another untouched — the
[[cellular-senescence|heterogeneity problem]] in its therapeutic form. The NIH-funded Cellular
Senescence Network was established in part to build the tissue atlases it requires.

Mouse-to-human translation carries the field's usual discount. Many striking results come from
progeroid strains, from irradiated or transplanted models, or from young animals given senescent
cells experimentally. Laboratory mice are short-lived, inbred, and housed in conditions that
exaggerate some pathologies and suppress others, and the relationship between "delays a pathology in
a mouse" and "changes an outcome in an eighty-year-old" is not established for any senolytic.

A related trap is the use of composite molecular readouts as evidence of benefit. Senolytic
treatment can shift an [[epigenetic-clock]] estimate or a panel of inflammatory markers without any
demonstrated change in how a person functions, and a shift in [[biological-age|estimated biological
age]] is a change in a prediction, not in an outcome. Regulators have not accepted any such readout
as a surrogate endpoint, which means senolytic trials must still be powered on events or on measured
function, and therefore remain long and expensive.

## Risks

The systemic senolytics under study are not benign compounds. Dasatinib carries oncology-grade
toxicity; navitoclax-class agents cause thrombocytopenia; quercetin and fisetin interact with drug
metabolism at high doses. Repeated clearance over years has no safety precedent, and the theoretical
concern that removing senescence removes a tumour-suppressive brake has not been tested on a
human timescale.

A separate risk is commercial. Fisetin and quercetin are sold as supplements with senolytic
marketing claims that far outrun the evidence, at doses chosen from mouse studies with no human
pharmacokinetic basis. This is the recurring pattern in longevity medicine, shared with
[[nad-precursors]] and with off-label mTOR-inhibitor use: a plausible mechanism, an unregulated
product, and no outcome data. Against that background it is worth noting the comparator. The
best-supported intervention for the functional endpoints senolytic trials target, among them
mobility, frailty, and cardiorespiratory capacity, is [[exercise-and-aging|structured exercise]],
and dietary approaches such as [[caloric-restriction]] have a far longer animal record than any
senolytic. A drug that cannot add to those has little claim on healthy people.

## Outlook

Three developments would change the picture. The first is a validated, tissue-resolved measure of
senescent-cell burden, which would convert trials from guesswork into pharmacology. The second is a
randomized trial with a functional endpoint such as mobility, frailty, or visual acuity, run in a
population with a high senescent-cell load, powered properly and controlled. The third is targeting: prodrugs
activated by senescence-associated enzymes, antibody-directed delivery of the kind developed for
[[targeted-drug-delivery|tumour-directed payloads]], or engineered immune clearance would let the
field escape the systemic toxicity ceiling that limits current compounds.

Senolytics remain among the most advanced tests of the [[geroscience-hypothesis]] — the claim that
hitting a shared upstream driver of aging beats treating diseases one at a time. Competitions such as
[[xprize-healthspan]], which reward measured restoration of muscle, cognitive, and immune function in
older adults, are structured to reward exactly that kind of demonstration, and senolytics are among
the candidate interventions entrants have pursued. Success at that level would be evidence for
[[compression-of-morbidity]], not for the far stronger claims about
[[longevity-escape-velocity|indefinite lifespan extension]] that the class is sometimes recruited to
support.

If the approach works in humans it would most likely first prove itself in a narrow indication with a
measurable organ endpoint rather than as a general anti-aging drug. If it does not, the informative
question will be whether the failure lies in the biology, in the compounds, or in the continuing
inability to measure what the compounds are supposed to be doing.

## See also

- [[cellular-senescence]]
- [[hallmarks-of-aging]]
- [[inflammaging]]
- [[rapamycin]]
- [[aging-biomarkers]]
- [[geroscience-hypothesis]]
- [[healthspan]]
- [[xprize-healthspan]]

## References

[^zhu2015]: `paper` Zhu, Y. et al. "The Achilles' heel of senescent cells: from transcriptome to senolytic drugs." *Aging Cell*, 2015. {The anti-apoptotic dependencies were identified in cultured cells; dasatinib and quercetin were existing compounds repurposed against them, not molecules designed for the target.}
[^chang2016]: `paper` Chang, J. et al. "Clearance of senescent cells by ABT263 rejuvenates aged hematopoietic stem cells in mice." *Nature Medicine*, 2016.
[^justice2019]: `paper` Justice, J.N. et al. "Senolytics in idiopathic pulmonary fibrosis: results from a first-in-human, open-label, pilot study." *EBioMedicine*, 2019.
[^hickson2019]: `paper` Hickson, L.J. et al. "Senolytics decrease senescent cells in humans: preliminary report from a clinical trial of dasatinib plus quercetin in individuals with diabetic kidney disease." *EBioMedicine*, 2019.
[^amor2020]: `paper` Amor, C. et al. "Senolytic CAR T cells reverse senescence-associated pathologies." *Nature*, 2020.
[^grosse2020]: `paper` Grosse, L. et al. "Defined p16High senescent cell types are indispensable for mouse healthspan." *Cell Metabolism*, 2020. {Clearance here was genetic and sustained rather than a short drug course, so the result bounds the hazard of indiscriminate clearance without measuring what an intermittent senolytic would do.}
[^farr2024]: `paper` Farr, J.N. et al. "Effects of intermittent senolytic therapy on bone metabolism in postmenopausal women: a randomized controlled trial." *Nature Medicine*, 2024.

==============================================================================
ARTICLE: sens-research-foundation
TITLE:   SENS Research Foundation
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/sens-research-foundation
SOURCE:  https://futurehumanwiki.com/raw/sens-research-foundation
==============================================================================

---
title: "SENS Research Foundation"
slug: "sens-research-foundation"
type: "organization"
status: "established"
horizon: "present"
categories: ["organizations", "longevity"]
tags: ["aging", "damage repair", "nonprofit", "rejuvenation", "research funding", "sens"]
summary: "A California nonprofit founded in 2009 to fund rejuvenation research organized around seven categories of accumulated cellular damage, renamed Lifespan Research Institute in 2024."
updated: "2026-07-27"
issues: ["The Thiel and Buterin donations are named without a source", "The 2021 investigation's reported conclusion needs a citation"]
---

```infobox
{
  "caption": "Research nonprofit",
  "rows": [
    { "label": "Founded", "value": "2009" },
    { "label": "Co-founder", "value": "Aubrey de Grey", "link": "/wiki/aubrey-de-grey" },
    { "label": "Parent organization", "value": "Methuselah Foundation", "link": "/wiki/methuselah-foundation" },
    { "label": "Headquarters", "value": "Mountain View, California" },
    { "label": "Framework", "value": "Seven damage categories" },
    { "label": "Renamed", "value": "2024, as Lifespan Research Institute" },
    { "label": "Funding", "value": "Philanthropic" }
  ]
}
```

**SENS Research Foundation** is a California nonprofit established in 2009 to fund research on Strategies for Engineered Negligible Senescence, a framework that treats aging as the accumulation of seven classes of cellular and molecular damage, each of which is to be repaired directly rather than prevented. Spun out of the [[methuselah-foundation]] by [[aubrey-de-grey]] and colleagues, it was for a decade the main institutional home of the damage-repair approach to aging. It removed de Grey in 2021 after allegations of misconduct, and rebranded as the Lifespan Research Institute in 2024.

## Overview

The SENS framework's distinguishing claim is engineering rather than biological. Its proponents argue that understanding the metabolic processes that generate damage is not necessary in order to remove the damage, in the same way that a mechanic need not understand combustion chemistry to replace a corroded part. This is a deliberate contrast with the mainstream [[geroscience-hypothesis]], which seeks to slow damage accumulation by modulating pathways such as mTOR and nutrient sensing with drugs like [[rapamycin]], and with the descriptive [[hallmarks-of-aging]] framework, which SENS advocates regard as a taxonomy that stops short of telling anyone what to build.

The framework enumerates seven damage categories, and asserts that the list is complete in the sense that no new class of damage has been discovered since the 1980s:

- cell loss and tissue atrophy
- death-resistant cells, principally [[cellular-senescence|senescent cells]]
- nuclear mutations and epimutations, which matter chiefly through cancer
- [[mitochondrial-dysfunction|mitochondrial DNA mutations]]
- intracellular aggregates that lysosomes cannot degrade
- extracellular aggregates such as amyloid
- extracellular protein crosslinks, principally glucosepane in collagen

> [!debate] Repair or slow down
> The damage-repair position holds that intervention should target the accumulated products of aging, which are enumerable, rather than the metabolism that produces them, which is not fully understood. Most biogerontologists regard the seven-category list as an assertion rather than a finding, and note that the boundary between damage and normal biological variation is not sharp. The disagreement is about research strategy, not about whether damage accumulates.

## History

De Grey developed the framework in the early 2000s, initially at Cambridge, and set out its full form in 2007 in *Ending Aging*, written with Michael Rae.[^degrey2007] The Methuselah Foundation, which de Grey had co-founded with David Gobel in the early 2000s, funded early SENS work; in 2009 SENS Foundation was incorporated separately with de Grey, Michael Kope, Jeff Hall, Sarah Marr and Kevin Perrott among its founders.

The organization ran an intramural laboratory in Mountain View, funded extramural projects at universities, and operated an education programme that placed undergraduates in aging laboratories. Its early funding came heavily from individual donors in the technology sector, including a substantial early gift from Peter Thiel and, later, cryptocurrency donations from Vitalik Buterin.

```timeline
[
  { "year": "2000s", "title": "The framework", "text": "Aubrey de Grey formulates the seven-category damage-repair model and argues it in the biogerontology literature and in Ending Aging." },
  { "year": "2005", "title": "The MIT challenge", "text": "MIT Technology Review offers a prize for a refutation of SENS; judges conclude that the framework has not been disproven but is not scientifically demonstrated either." },
  { "year": "2009", "title": "SENS Foundation founded", "text": "The programme is spun out of the Methuselah Foundation as an independent nonprofit with its own laboratory in Mountain View." },
  { "year": "2010s", "title": "Spin-outs", "text": "Funded projects yield companies including Cyclarity Therapeutics, working on cyclodextrins for arterial cholesterol, and Revel Pharmaceuticals, working on crosslink-breaking enzymes." },
  { "year": "2021", "title": "Governance crisis", "text": "Allegations of misconduct lead to de Grey being placed on leave and then removed; he denies wrongdoing and later founds the LEV Foundation." },
  { "year": "2024", "title": "Renamed", "text": "The organization becomes the Lifespan Research Institute, retaining the damage-repair research agenda under new leadership." }
]
```

## Research programme

The foundation's approach was to fund the categories that no one else was working on. Cell loss is addressed by mainstream regenerative medicine and [[induced-pluripotent-stem-cells|stem cell]] research; cancer attracts enormous investment; senescent cells acquired a large commercial following after 2015 with the rise of [[senolytics]]. The organization concentrated its own money on the neglected remainder.

Two lines produced spin-out companies. Work on lysosomal aggregates in atherosclerosis — 7-ketocholesterol, which macrophages cannot degrade and which contributes to foam-cell formation — led to Cyclarity Therapeutics, founded by Matthew O'Connor, which developed an engineered cyclodextrin intended to extract the molecule from plaque and began first-in-human testing in Australia in the mid-2020s. Work on glucosepane, the dominant crosslink in aged human collagen, led to Revel Pharmaceuticals, founded with the Yale chemist David Spiegel, which pursued enzymes capable of cleaving it.

A third line, mitochondrial [[gene-therapy-for-aging|gene therapy]] by allotopic expression — inserting backup copies of the thirteen mitochondrially encoded genes into the nucleus so that a cell with damaged mitochondrial DNA can still make the proteins — was pursued in the foundation's own laboratory and produced published results for a subset of the genes. It has not reached human testing, and the approach is distinct from the [[mitochondrial-replacement-therapy|germline replacement]] techniques used to prevent inherited mitochondrial disease.

## The 2005 challenge

In 2005 *MIT Technology Review* offered a prize to any molecular biologist who could demonstrate that SENS was "so wrong that it is unworthy of learned debate". Submissions were judged by a panel including Rodney Brooks, Anita Goel, Nathan Myhrvold, Vikram Sheel Kumar and J. Craig Venter. No prize was awarded. The judges' conclusion, which both sides have quoted selectively ever since, was that the critics had not proven SENS false but that de Grey had not proven it plausible either, and that it belonged in the category of ideas that are unsupported rather than refuted.[^pontin2006] Two decades later that verdict has aged well: several SENS categories have become mainstream research areas, and none has produced a demonstrated human rejuvenation therapy.

## Funding

The organization has always been philanthropically funded, with annual budgets in the single-digit millions of dollars — small relative to [[calico]] or [[altos-labs]] by three orders of magnitude, and small relative to national aging-research budgets. Its donor base was concentrated, which made it vulnerable to the loss of individual supporters and shaped the governance problems that followed.

Cryptocurrency donations were significant during the 2017 and 2021 market peaks and volatile in the way that implies. Nominal values reported at the time of donation frequently exceeded what was realized.

## The 2021 crisis and rebranding

In August 2021 two women made public allegations concerning de Grey's conduct. The foundation placed him on administrative leave and commissioned an independent investigation, then terminated his employment. The investigation's reported conclusion was that it had not substantiated a finding of sexual harassment but that de Grey had improperly interfered with the investigation itself. De Grey denied wrongdoing throughout and subsequently reached a settlement with the organization; he founded the Longevity Escape Velocity Foundation in 2022, named for the [[longevity-escape-velocity]] threshold he had popularized, to run large mouse studies combining several interventions at once.

The episode cost the organization donors, staff and its most recognizable public figure at once. Under new leadership it narrowed its programme, and in 2024 renamed itself the Lifespan Research Institute, a change that separated the organization's identity from both the SENS acronym and its founder.

## Reception and legacy

Mainstream biogerontology has never accepted the SENS framework as a theory, and a 2005 rebuttal in *EMBO Reports* signed by a group of prominent aging researchers argued that it presents speculative proposals as an engineering roadmap and that none of the seven interventions had a demonstrated route to the clinic.[^warner2005] That criticism was substantially correct at the time and remains largely correct.

The foundation's actual influence runs through people and topics rather than through the framework. It funded senescent-cell work before senolytics were fashionable, kept crosslink chemistry and lysosomal aggregate research alive when no one else would pay for them, and trained a cohort of researchers now distributed across the field and the [[buck-institute]] and comparable centres. It also normalized the idea that aging is a legitimate target for intervention rather than an inevitability, which made the funding environment that produced Altos, [[retro-biosciences]] and [[xprize-healthspan]] easier to create.

Whether the seven-category framework is the right decomposition remains unsettled. It is at least falsifiable in principle: if a repair therapy for one category produced a measurable rejuvenation in a mammal without touching the others, the framework's modular premise would gain support it currently lacks.

## See also

- [[aubrey-de-grey]]
- [[methuselah-foundation]]
- [[longevity-escape-velocity]]
- [[senolytics]]
- [[hallmarks-of-aging]]
- [[geroscience-hypothesis]]
- [[buck-institute]]
- [[negligible-senescence]]

## References

[^degrey2007]: `book` de Grey, A. and Rae, M. *Ending Aging: The Rejuvenation Breakthroughs That Could Reverse Human Aging in Our Lifetime*. St. Martin's Press, 2007.
[^pontin2006]: `news` Pontin, J. "Is Defeating Aging Only A Dream?" *MIT Technology Review*, 2006. {The magazine both ran the challenge and published this account of its outcome, so it is a party to the episode as well as a source on it.}
[^warner2005]: `paper` Warner, H., Anderson, J., Austad, S. et al. "Science fact and the SENS agenda." *EMBO Reports*, 2005. {A signed critique rather than a research report; it argues from the state of the evidence in 2005 and presents no new experiment.}

==============================================================================
ARTICLE: sensory-augmentation
TITLE:   Sensory augmentation
PORTAL:  Human Enhancement
URL:     https://futurehumanwiki.com/wiki/sensory-augmentation
SOURCE:  https://futurehumanwiki.com/raw/sensory-augmentation
==============================================================================

---
title: "Sensory augmentation"
slug: "sensory-augmentation"
type: "technology"
status: "experimental"
horizon: "late 2020s"
trl: 6
categories: ["enhancement", "cybernetics"]
tags: ["senses", "enhancement", "neuroplasticity", "implants", "perception", "sensory substitution"]
summary: "Devices and biological modifications that route information the body cannot normally detect into a channel the brain can learn to read."
updated: "2026-07-27"
humanEvidence: "In people only worn devices and subdermal magnets have been used; the Osnabrück magnetic-north belt produced measurable navigation and cortical changes after weeks of wear, and the biological routes to a new channel exist only in monkeys and mice."
access: "Tactile and auditory substitution devices are cleared and sold as assistive equipment in some jurisdictions; augmentation devices such as magnetic-north wearables and subdermal fingertip magnets are unregulated consumer items."
reversibility: "reversible"
issues: ["The 2015 US clearance of a tongue display is stated without naming the device or a source"]
---

```infobox
{
  "caption": "Class of perceptual technology",
  "rows": [
    { "label": "First demonstration", "value": "1969, tactile-vision substitution" },
    { "label": "Main approaches", "value": "Substitution, addition, biological" },
    { "label": "Typical output channels", "value": "Skin, tongue, hearing" },
    { "label": "Cleared assistive devices", "value": "Tactile and auditory substitution" },
    { "label": "Readiness", "value": "Assistive marketed; augmentation experimental" },
    { "label": "Key limit", "value": "Channel bandwidth, not brain plasticity" }
  ]
}
```

**Sensory augmentation** is the use of devices or biological modification to deliver information a human body cannot normally detect — magnetic north, infrared, seismic activity, air quality — through a sensory channel that can carry it. It is distinguished from sensory substitution, which routes information from a lost sense through a surviving one, mainly by intent: the hardware and the neuroscience are largely the same, and most augmentation work descends directly from substitution research done for blind and deaf users.

## How it works

Every approach follows the same three steps: capture a signal with a sensor, compress it into a low-bandwidth code, and deliver that code through skin, hearing, or another intact channel. The brain is expected to do the rest.

The critical constraint is bandwidth. The optic nerve carries roughly a million axons; a tactile array on the back or a vibrating wristband delivers a few dozen distinguishable channels at best. Every substitution device therefore throws away nearly all of the information it captures, and the design problem is choosing which fraction to keep. This is why devices for navigation and obstacle avoidance work reasonably well and devices intended to convey scenes do not.

The second requirement is that the signal be coupled to action. Perceptual learning depends on the user changing the input by moving; a passive stream of vibration is experienced as vibration, while a stream that changes systematically when the head turns can come to be experienced as a property of the world. Researchers call this distal attribution, and it is the difference between feeling a buzz on the wrist and feeling that north is over there.

```timeline
[
  { "year": "1969", "title": "Tactile vision substitution", "text": "Paul Bach-y-Rita's group delivers a camera image as a pattern of vibration on the back and reports that blind users learn to identify objects and judge distance." },
  { "year": "1992", "title": "The vOICe", "text": "Peter Meijer describes a system converting images to soundscapes, later distributed as free software and still in use." },
  { "year": "c. 2004", "title": "The eyeborg", "text": "Neil Harbisson, who has achromatopsia, begins wearing a head-mounted device that converts colour to pitch, later reporting perception of frequencies outside the visible range." },
  { "year": "2009", "title": "A new colour channel in adult primates", "text": "Gene therapy adding a long-wavelength opsin gives adult dichromatic monkeys trichromatic colour vision, showing an adult brain can exploit a channel it never developed with." },
  { "year": "2015", "title": "Regulatory clearance", "text": "A tongue-based tactile display for the blind receives United States marketing authorisation, making sensory substitution a cleared medical device category." },
  { "year": "2016–2017", "title": "Consumer augmentation", "text": "Magnetic-north implants and sound-to-vibration wristbands reach small commercial markets, moving augmentation outside the laboratory." }
]
```

## Substitution, addition, and biological routes

**Substitution devices** convert one modality into another for people missing a sense. The tactile-vision systems of Bach-y-Rita's group established the template, using a matrix of vibrating points against the skin.[^bachyrita1969] The vOICe converts images to soundscapes.[^meijer1992] Tongue displays exploit the dense mechanoreceptor population of the tongue surface to deliver higher spatial resolution than skin allows. None of these restores anything resembling vision; they restore specific capabilities such as locating a doorway or reading large letters.

**Addition devices** give an intact person a channel they never had. The feelSpace belt, developed at Osnabrück, vibrates at whichever point on the waist faces magnetic north; after weeks of continuous use, participants reported changes in how they represented space, alongside measurable differences in navigation behaviour and in cortical responses.[^konig2016] The North Sense, a small chest-mounted device introduced by a cyborg-culture company in the mid-2010s, does the same job with a single vibrating point. Subdermal neodymium magnets in the fingertip, widespread in the [[biohacking]] community since the mid-2000s, transduce nearby alternating magnetic fields into vibration felt directly by mechanoreceptors — the only common augmentation with no electronics at all.

Neil Harbisson's antenna, which converts colour and eventually near-infrared and ultraviolet to bone-conducted pitch, is the most-publicised case, the one most closely tied to the cyborg-activist wing of [[transhumanism]], and the most instructive, because Harbisson reports the transition from conscious translation to automatic perception and to colour-associated dreaming. Such reports are single-subject and unblinded; they are evidence about what is possible to experience, not measurements of it.

**Biological routes** modify the receptor layer instead of bypassing it. Adding a long-wavelength opsin gene to the retinas of adult dichromatic monkeys, by the [[aav-vectors|adeno-associated viral]] route standard in [[somatic-gene-therapy]], produced behavioural trichromacy, which established that a mature visual system can extract information from a new receptor class without developmental rewiring.[^mancuso2009] Injectable upconversion nanoparticles bound to photoreceptors gave mice a near-infrared channel alongside normal vision.[^ma2019] Both are demonstrated in animals only. The nearest human precedent is the use of [[optogenetics]] to confer light sensitivity on surviving retinal cells in a patient with retinitis pigmentosa, which restored crude localisation rather than adding a channel. No comparable human augmentation experiment has been conducted, and neither result implies that a similar approach in humans would be safe or effective.

## Evidence on plasticity

The strongest neuroscientific result from this field is that the cortex is organised more by task than by input channel. In congenitally blind users of visual-to-auditory substitution, regions of the ventral visual stream normally associated with object shape respond to soundscapes conveying shape.[^amedi2007] Similar task-specific rather than modality-specific responses have been reported for reading, motion and spatial layout. This is why sensory substitution works at all, and why blind users typically outperform sighted users on the same devices.

> [!debate] Is it a new sense or a new skill
> Deroy and Auvray argue that sensory substitution is better described as a learned perceptual skill, closer to reading than to seeing, and that it does not create a new modality with its own phenomenal character.[^deroy2012] Defenders point to distal attribution and to automaticity as evidence that something more than inference is occurring. The dispute cannot be settled by task performance, since both accounts predict the same behaviour, and it is the clearest case in enhancement research where the question is about experience rather than capability.

## Current state

As of 2026 the clinical end of the field is modest but real: tactile and auditory substitution devices have regulatory clearance in some jurisdictions, and sound-to-vibration wristbands are sold to deaf users and for tinnitus. As a category of [[human-enhancement]] they are unusually cheap and unusually low-risk. They occupy the space below [[cochlear-implant|cochlear implants]] and [[retinal-implant|retinal prostheses]], which intervene directly on the nerve or retina and deliver correspondingly more information at correspondingly greater surgical cost.

The augmentation end is a subculture with a small commercial fringe. Devices are cheap, low-risk relative to implanted electrodes, and unregulated. What has not appeared is evidence that any added sense confers a measurable practical advantage over consulting an instrument. The Osnabrück belt work is the closest thing to a controlled demonstration of changed spatial cognition, and it involved weeks of continuous wear.

## Limitations and risks

Bandwidth is the binding constraint, and it is not obviously improvable: the skin's spatial and temporal resolution is fixed, and adding vibrators past a certain density produces confusion rather than detail. Training burden is the second constraint. Devices that require weeks of daily use before becoming automatic have high abandonment rates, the same pattern seen in [[myoelectric-prosthetics]].

Implanted augmentation carries the risks of any foreign body: infection at the insertion site, migration, and failure of the encapsulating coating. Fingertip magnets can fracture or lose their coating, causing local tissue damage, and complicate magnetic resonance imaging. Devices installed outside medical settings, which most are, have no adverse-event reporting, so the actual complication rate in this population is unknown. The legal position of such installations, and of the practitioners who perform them, is examined in [[morphological-freedom]].

Attentional cost is under-studied. A continuous stream of tactile information competes for the same resources as everything else the skin reports, and whether an added channel is free once automatic, or permanently taxing, has not been measured over long periods.

## Outlook

Three developments would change the picture. Higher-resolution non-invasive displays — dense electrotactile arrays, or ultrasound-based stimulation — would relax the bandwidth constraint. Direct cortical stimulation, already used experimentally to deliver touch feedback in [[neuroprosthetics|bidirectional neuroprostheses]], would bypass the peripheral channel entirely, at the cost of surgery; the [[brain-computer-interface]] literature treats write-in as considerably harder than the read-out problem described in [[neural-decoding]]. And a biological route in humans, along the lines of the primate opsin experiment, would create a genuinely new receptor population rather than repurposing an existing one, which is the only approach that would settle the question of whether new qualia are possible.

None of these is close. The more likely near-term outcome is that augmentation stays where it is: a small set of cheap devices with strong subjective reports, thin objective evidence, and an unresolved argument about whether their users are perceiving or inferring.

## See also

- [[cochlear-implant]]
- [[retinal-implant]]
- [[neuroprosthetics]]
- [[biohacking]]
- [[human-enhancement]]
- [[brain-computer-interface]]
- [[optogenetics]]
- [[morphological-freedom]]

## References

[^bachyrita1969]: `paper` Bach-y-Rita, P., Collins, C. C., Saunders, F. A., White, B. and Scadden, L. "Vision Substitution by Tactile Image Projection." *Nature*, 1969.
[^meijer1992]: `paper` Meijer, P. B. L. "An Experimental System for Auditory Image Representations." *IEEE Transactions on Biomedical Engineering*, 1992.
[^mancuso2009]: `paper` Mancuso, K. et al. "Gene therapy for red–green colour blindness in adult primates." *Nature*, 2009. {Adult dichromatic squirrel monkeys assessed on a colour-discrimination task; the result establishes behavioural capability, not what the animals experienced.}
[^ma2019]: `paper` Ma, Y. et al. "Mammalian Near-Infrared Image Vision through Injectable and Self-Powered Retinal Nanoantennae." *Cell*, 2019.
[^konig2016]: `paper` König, S. U. et al. "Learning New Sensorimotor Contingencies: Effects of Long-Term Use of Sensory Augmentation on the Brain and Conscious Perception." *PLOS ONE*, 2016. {A small study that cannot be blinded, since participants know they are wearing the belt, so the reported changes in spatial experience carry an expectation component.}
[^amedi2007]: `paper` Amedi, A. et al. "Shape conveyed by visual-to-auditory sensory substitution activates the lateral occipital complex." *Nature Neuroscience*, 2007.
[^deroy2012]: `paper` Deroy, O. and Auvray, M. "Reading the World through the Skin and Ears: A New Perspective on Sensory Substitution." *Frontiers in Psychology*, 2012. {A conceptual analysis that reinterprets existing substitution results and reports no new data of its own.}

==============================================================================
ARTICLE: shinya-yamanaka
TITLE:   Shinya Yamanaka
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/shinya-yamanaka
SOURCE:  https://futurehumanwiki.com/raw/shinya-yamanaka
==============================================================================

---
title: "Shinya Yamanaka"
slug: "shinya-yamanaka"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "genetics"]
tags: ["stem cells", "reprogramming", "ipsc", "nobel prize", "regenerative medicine"]
summary: "Japanese physician-scientist who showed that four transcription factors can reprogram adult cells to a pluripotent state, sharing the 2012 Nobel Prize for the discovery."
updated: "2026-07-27"
issues: ["The mid-2020s Parkinson's iPSC trial results are described without a citation"]
---

```infobox
{
  "caption": "Physician-scientist",
  "rows": [
    { "label": "Born", "value": "4 September 1962, Osaka" },
    { "label": "Nationality", "value": "Japanese" },
    { "label": "Education", "value": "MD Kobe (1987); PhD Osaka City (1993)" },
    { "label": "Known for", "value": "Induced pluripotent stem cells", "link": "/wiki/induced-pluripotent-stem-cells" },
    { "label": "Field", "value": "Stem cell biology" },
    { "label": "Affiliation", "value": "Kyoto University (CiRA); Gladstone Institutes" },
    { "label": "Award", "value": "Nobel Prize in Physiology or Medicine, 2012" }
  ]
}
```

**Shinya Yamanaka** is a Japanese physician-scientist who demonstrated in 2006 that introducing four
transcription factors into an adult mouse fibroblast can return it to an embryonic-like pluripotent
state, and repeated the result in human cells the following year. The four factors — Oct3/4, Sox2,
Klf4 and c-Myc — are now known as the [[yamanaka-factors]], and the cells they produce as
[[induced-pluripotent-stem-cells]]. He shared the 2012 Nobel Prize in Physiology or Medicine with
John Gurdon.

## Career

Yamanaka trained as an orthopedic surgeon at Kobe University, an experience he has described as
unsuccessful — he was slow in the operating theatre and found the limits of surgical repair
frustrating — and moved into research, completing a doctorate in pharmacology at Osaka City
University in 1993. A postdoctoral position at the Gladstone Institutes in San Francisco taught him
American laboratory culture and, in his own account, the value of pursuing unexpected results: a
transgenic mouse experiment aimed at cholesterol metabolism instead produced tumours, which redirected
his attention to gene regulation.

Returning to Japan, he held positions at Osaka City University and the Nara Institute of Science and
Technology before moving to Kyoto University in 2004. In 2010 he became the founding director of the
Center for iPS Cell Research and Application (CiRA) at Kyoto, which he led until 2022; he remains a
professor there, heads the associated CiRA Foundation, and retains a senior investigator position at
Gladstone.

## The discovery

Gurdon had shown in the early 1960s that transferring a differentiated frog nucleus into an
enucleated egg could produce a whole animal, establishing that differentiation does not delete
genetic information.[^gurdon1962] The same principle later produced Dolly and the whole field of
[[human-cloning|somatic cell nuclear transfer]]. What remained unknown was whether the reprogramming
activity resident in the egg could be reduced to a defined set of factors.

Yamanaka's group assembled twenty-four candidate genes associated with embryonic stem cell identity,
introduced them together by retrovirus into mouse fibroblasts carrying a selectable pluripotency
marker, and then removed genes one at a time to find the minimal set. Four sufficed.[^ty2006] The
resulting cells formed teratomas containing all three germ layers and, in later work, contributed to
chimeric mice and the germline. In 2007 the group reported the same result in human dermal
fibroblasts, published within days of an independent report from James Thomson's laboratory using a
partly different factor combination.[^ty2007]

```timeline
[
  { "year": "1962", "title": "Nuclear transfer in frogs", "text": "John Gurdon shows that a differentiated nucleus can support development, establishing that differentiation is reversible in principle." },
  { "year": "2006", "title": "Mouse iPS cells", "text": "Takahashi and Yamanaka narrow twenty-four candidate factors to four that reprogram mouse fibroblasts to pluripotency." },
  { "year": "2007", "title": "Human iPS cells", "text": "Yamanaka's group and, independently, James Thomson's group report reprogrammed human cells." },
  { "year": "2012", "title": "Nobel Prize", "text": "Yamanaka shares the Nobel Prize in Physiology or Medicine with Gurdon." },
  { "year": "2014", "title": "First clinical transplant", "text": "A Japanese team led by Masayo Takahashi transplants iPSC-derived retinal pigment epithelium into a patient with macular degeneration." }
]
```

The finding did two things at once. It removed the political obstacle that had constrained human
embryonic stem cell research in several countries, since iPSCs require no embryo. And it established
that cell identity is maintained by a transcription-factor network that can be overridden — the
premise on which [[partial-reprogramming]] and [[epigenetic-reprogramming]] as rejuvenation
strategies later rested.

## Clinical translation

CiRA's translational programme has moved deliberately. The first clinical use of iPSC-derived
tissue was a retinal pigment epithelium graft for age-related macular degeneration in 2014, using
cells derived from the patient's own skin.[^mandai2017] That autologous route proved too slow and
too expensive to scale, and Yamanaka pushed instead for a bank of iPSC lines from donors homozygous
at the major HLA loci, so that a small number of lines could be immunologically compatible with a
large fraction of the Japanese population. Later work has used [[crispr-cas9]] to edit HLA genes
directly, aiming at hypoimmunogenic universal lines.

Kyoto teams have since run early-phase trials of iPSC-derived dopaminergic progenitors for
Parkinson's disease, alongside independent programmes elsewhere; results reported in the mid-2020s
have concerned safety and graft survival, with motor improvement in some participants but no
controlled comparison that would establish clinical benefit. Cardiac muscle
sheets, platelets and corneal epithelium have also entered early trials. Twenty years after the
mouse paper, no iPSC-derived therapy is in routine clinical use, and the whole organs that would
address the [[organ-shortage]] remain the province of [[tissue-engineering]] and
[[lab-grown-organs]] rather than of reprogramming alone.

> [!note] Two uses of the same cells
> iPSCs matter clinically as a source of transplantable tissue, and they matter scientifically as
> patient-specific disease models and drug-screening substrates. The second use — including
> [[organoids]] grown from patient cells — has produced far more published results than the first,
> and requires no regulatory approval.

## Caution about timelines

Yamanaka is unusual among the figures covered on this wiki for consistently arguing that his own
field will take longer than its advocates say. He has repeatedly cautioned that iPSC therapies face
manufacturing cost, tumorigenicity and immune-rejection problems that are not solved by the
reprogramming step itself, and that residual undifferentiated cells in a graft are a genuine cancer
risk rather than a formality. On rejuvenation specifically, he has been more reserved than the
commercial reprogramming sector, noting that the factors that restore pluripotency are the same
factors that erase cell identity, and that the therapeutic window between the two is narrow and
poorly characterized in vivo. A reset [[epigenetic-clock]] reading in a reprogrammed cell is a
change in a measurement, not a demonstration that the tissue functions better.

He has also declined to profit personally from the discovery in the way his position would allow,
directing patent income to CiRA and raising research funds by running marathons — a detail that is
frequently mentioned because it is so far from the norm in the sector.

## Reception and legacy

The 2006 paper is one of the most cited in modern biology, and iPSCs are now standard laboratory
material. Reprogramming also supplied the conceptual foundation for the reprogramming-based
longevity companies: [[altos-labs]] pursues partial rather than complete factor expression precisely
to avoid the pluripotency Yamanaka was trying to reach, and [[newlimit]] screens for other factor
combinations that never approach it. He agreed to serve as a senior scientific adviser to Altos
while remaining at Kyoto University, an arrangement that sits alongside rather than in place of the
reservations described above.

The open problem he identified early has not moved much. Reprogramming is a continuum, and nobody
can yet specify how far along it a cell can be pushed to gain a younger epigenetic profile while
retaining its function and not becoming a tumour. Until that dose–response relationship is
characterized in vivo, the distance between his discovery and its most ambitious application
remains a matter of assertion.

## See also

- [[induced-pluripotent-stem-cells]]
- [[yamanaka-factors]]
- [[partial-reprogramming]]
- [[epigenetic-reprogramming]]
- [[organoids]]
- [[human-cloning]]
- [[david-sinclair]]
- [[lab-grown-organs]]

## References

[^gurdon1962]: `paper` Gurdon, J. B. "The developmental capacity of nuclei taken from intestinal epithelium cells of feeding tadpoles." *Journal of Embryology and Experimental Morphology*, 1962.
[^ty2006]: `paper` Takahashi, K. and Yamanaka, S. "Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors." *Cell*, 2006. {Selection in this first paper was on the Fbx15 marker, and those lines were not germline-competent; that came with Nanog-selected lines the following year.}
[^ty2007]: `paper` Takahashi, K. et al. "Induction of pluripotent stem cells from adult human fibroblasts by defined factors." *Cell*, 2007.
[^mandai2017]: `paper` Mandai, M. et al. "Autologous Induced Stem-Cell-Derived Retinal Cells for Macular Degeneration." *New England Journal of Medicine*, 2017. {One patient received the autologous graft; the second planned recipient was not treated after genomic changes were found in the prepared cells.}

==============================================================================
ARTICLE: singularitarianism
TITLE:   Singularitarianism
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/singularitarianism
SOURCE:  https://futurehumanwiki.com/raw/singularitarianism
==============================================================================

---
title: "Singularitarianism"
slug: "singularitarianism"
type: "concept"
status: "contested"
horizon: "indefinite"
categories: ["people", "minds"]
tags: ["singularity", "artificial intelligence", "movements", "transhumanism", "ai safety", "forecasting"]
summary: "The belief that a technological singularity is likely and that it should be actively prepared for, together with the movements organized around that belief."
updated: "2026-07-27"
issues: ["The Gray and Lanier criticisms are named without citations.", "Yudkowsky's c. 2000 essays are described but not cited."]
---

```infobox
{
  "caption": "Intellectual movement",
  "rows": [
    { "label": "Term coined", "value": "c. 1991" },
    { "label": "Central claim", "value": "Superhuman AI is achievable and near" },
    { "label": "Principal texts", "value": "Vinge 1993; Kurzweil 2005" },
    { "label": "Key institutions", "value": "MIRI, Singularity University" },
    { "label": "Main internal split", "value": "Safety versus acceleration" },
    { "label": "Related to", "value": "Transhumanism", "link": "/wiki/transhumanism" },
    { "label": "Scientific standing", "value": "Minority position" }
  ]
}
```

**Singularitarianism** is the position that a [[technological-singularity]] is likely within a foreseeable period and that this warrants a deliberate response now, along with the loose set of movements built around that position. It is distinct from the singularity hypothesis itself: one can hold the hypothesis without any accompanying commitment about what to do. What makes a singularitarian is the second half — that the transition is important enough, and near enough, to organize one's work or one's institution around.

## Overview

The label was applied within the [[extropianism|extropian]] community around 1991 and used through the 1990s as a self-description. By the mid-2000s it had two very different referents. [[ray-kurzweil]] used it in *The Singularity Is Near* to describe someone who understands the coming transition and plans their life accordingly, with an emphasis on personal survival to reach it. Eliezer Yudkowsky, in a set of essays written around 2000, defined it as a commitment to bring about a beneficial singularity and treated the safety of the transition as the whole point.

These readings diverged into two institutional traditions that still barely speak to one another, and the word is now used more often by critics than by adherents.

## Origins

Three texts do most of the work. I. J. Good's 1965 paper describes the mechanism.[^good1965] Vernor Vinge's 1993 essay names the event and argues that it ends the era of predictable futures.[^vinge1993] Kurzweil's *The Age of Spiritual Machines* in 1999 and *The Singularity Is Near* in 2005 supply the trend-extrapolation version and the 2045 date that most people associate with the term.[^kurzweil2005]

The movement's early social base was the extropians mailing list and the [[cryonics]] community, where the singularity slotted neatly into an existing structure of belief: survive long enough, by [[alcor|cryopreservation]] if necessary, and superintelligent medicine will do the rest. This is why singularitarianism and [[transhumanism]] overlap so heavily in membership despite being logically independent, and why the same people who organized around [[max-more]]'s principles and later around [[humanity-plus]] also produced the first serious treatments of [[whole-brain-emulation]] and [[mind-uploading]]. In its strongest form the package is a single conditional chain: reach the singularity, or be preserved until it arrives, and [[longevity-escape-velocity]] follows from it rather than from biology.

```timeline
[
  { "year": "1991", "title": "The term appears", "text": "'Singularitarian' is coined within the extropian community as a self-description for those who take the singularity seriously." },
  { "year": "1999–2000", "title": "The safety framing", "text": "Eliezer Yudkowsky publishes essays defining singularitarianism as a commitment to a beneficial transition rather than merely a prediction." },
  { "year": "2000", "title": "First dedicated institution", "text": "The Singularity Institute for Artificial Intelligence is founded to work on the problem directly." },
  { "year": "2005", "title": "The popular version", "text": "Kurzweil's The Singularity Is Near reaches a mass audience and fixes 2045 in public memory as the date." },
  { "year": "2008", "title": "The acceleration wing", "text": "Singularity University is founded by Kurzweil and Peter Diamandis, oriented toward entrepreneurship rather than risk." },
  { "year": "2013", "title": "Renaming and divergence", "text": "The Singularity Institute becomes the Machine Intelligence Research Institute, dropping the word as its associations became a liability." },
  { "year": "2014", "title": "Entry to the mainstream", "text": "Nick Bostrom's Superintelligence recasts the argument in the language of decision theory and policy, and is taken up well outside the movement." }
]
```

## The two wings

### Safety-focused

The safety wing holds that an intelligence explosion is possible, that its outcome depends on properties of the first sufficiently capable system, and that those properties are not automatically good. The Singularity Institute for Artificial Intelligence, founded in 2000, was the first organization created on this premise; it renamed itself the Machine Intelligence Research Institute in 2013, in part because the word "singularity" had acquired connotations the organization did not want.

This wing's arguments were substantially absorbed into the mainstream by [[nick-bostrom]]'s *Superintelligence* in 2014,[^bostrom2014] and from there into a research field — AI alignment — that is now pursued inside major laboratories and funded by governments. Most people working in that field today would not describe themselves as singularitarians, and many reject the fast-takeoff assumptions that motivated the founders. The intellectual descent is nonetheless direct, and it runs through the same [[existential-risk]] literature that produced [[differential-technological-development]].

### Acceleration-focused

The acceleration wing treats the transition as desirable and the main task as bringing it closer. Singularity University, founded in 2008 by Kurzweil and Peter Diamandis with corporate sponsorship, is its most visible institution: an executive-education and entrepreneurship organization built around exponential-technology framing rather than a research body. Its posture — that accelerating technology is the solution to human problems, and that the appropriate response is to build — reappears with different vocabulary in [[effective-accelerationism]].

> [!debate] The same premise, opposite conclusions
> Both wings accept Good's argument. They differ on the conditional probability of a good outcome given a fast transition. The safety wing treats that probability as low by default and the alignment problem as unsolved; the acceleration wing treats it as high, on the grounds that capability and benevolence tend to correlate and that delay has its own body count. Neither number is measurable, and the disagreement has proven resistant to evidence.

## Reception

Mainstream reception has been consistently cool, for reasons that are worth separating.

Scientific criticism focuses on the trend-extrapolation version. Curve-fitting across incommensurable technologies, treating computational throughput as a proxy for intelligence, and dating an event by the intersection of two extrapolated lines are all methodologically weak, and Kurzweil's own prediction scorecard is disputed by his critics on the grounds that he grades leniently. The wider case for [[accelerating-change]] as a law rather than a description of one industry's good half-century has never been established. The economist William Nordhaus's tests for a growth singularity in the data mostly came back negative.[^nordhaus2021]

Sociological criticism focuses on the movement's shape. Critics have noted that singularitarianism reproduces the structure of eschatology — an approaching transformative event, a select group that understands it, personal salvation for those who prepare — and that this structure is doing rhetorical work independent of the technical argument. The philosopher John Gray and the computer scientist Jaron Lanier have both pressed versions of this objection. Adherents respond that structural resemblance to a religion is not an argument against a claim's truth, which is correct but does not address the question of why the belief is attractive.

A third criticism, from within the broader AI community, is that the movement's early focus on a hypothetical superintelligence displaced attention from the concrete harms of systems that already exist. That charge became sharper after 2022, when deployed models began producing measurable effects on labour, information, and safety without anything resembling recursive self-improvement.

## State of play

The word has largely fallen out of use even where the belief persists. The safety wing's concerns are now discussed under "AI risk" or "alignment", inside institutions that would not use the older label; [[anders-sandberg]] and others who worked on the movement's technical questions publish under ordinary academic headings. The acceleration wing's are discussed under techno-optimist banners of more recent coinage. Kurzweil's *The Singularity Is Nearer*, published in 2024, restated the original position and its 2045 date largely unchanged.

What has changed is the standing of the underlying question. Between 2000 and 2020, arguing that machine intelligence might exceed human capability within decades placed a person outside serious discussion. As of 2026 the question is treated as legitimate by national governments and by the laboratories building the systems, though the specific singularitarian claim — a discontinuous, self-driven capability explosion — remains a minority view among AI researchers rather than a consensus. The movement won the argument about whether the topic deserves attention and has not won the argument about its own mechanism.

## Legacy

Singularitarianism's durable contribution is institutional rather than theoretical. It created the first organizations to work full-time on the consequences of advanced AI, at a time when doing so carried professional cost, and it trained a cohort of people who now populate alignment teams, forecasting organizations, and philanthropic funders. It also supplied a vocabulary — takeoff speed, recursive self-improvement, [[artificial-general-intelligence]] — that the mainstream adopted while discarding the movement's confidence about timing.

The unresolved question is whether the movement's early framing helped or hindered. A discontinuous-takeoff model concentrated attention on a single dramatic scenario, and if capability growth turns out to be continuous and diffuse, as Robin Hanson and others have long argued, then two decades of preparation will have been aimed at the wrong shape of problem.

## See also

- [[technological-singularity]]
- [[artificial-general-intelligence]]
- [[ray-kurzweil]]
- [[nick-bostrom]]
- [[extropianism]]
- [[effective-accelerationism]]
- [[existential-risk]]
- [[mind-uploading]]

## References

[^good1965]: `paper` Good, I.J. "Speculations Concerning the First Ultraintelligent Machine." *Advances in Computers*, vol. 6, 1965. {The source of the phrase intelligence explosion. The argument is entirely a priori: it describes a mechanism and offers no evidence that one can be built.}
[^vinge1993]: `paper` Vinge, V. "The Coming Technological Singularity: How to Survive in the Post-Human Era." VISION-21 Symposium, 1993. {An essay presented at a NASA-sponsored symposium rather than a reviewed paper; it predicted the means to create superhuman intelligence within thirty years.}
[^kurzweil2005]: `book` Kurzweil, R. *The Singularity Is Near: When Humans Transcend Biology*. Viking, 2005.
[^bostrom2014]: `book` Bostrom, N. *Superintelligence: Paths, Dangers, Strategies*. Oxford University Press, 2014.
[^nordhaus2021]: `paper` Nordhaus, W. "Are We Approaching an Economic Singularity? Information Technology and the Future of Economic Growth." *International Economic Review*, 2021. {The tests run on macroeconomic aggregates such as productivity and capital share, so they bear on economic growth rather than on machine capability.}

==============================================================================
ARTICLE: sleep-and-longevity
TITLE:   Sleep and longevity
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/sleep-and-longevity
SOURCE:  https://futurehumanwiki.com/raw/sleep-and-longevity
==============================================================================

---
title: "Sleep and longevity"
slug: "sleep-and-longevity"
type: "intervention"
status: "established"
horizon: "present"
trl: 9
categories: ["longevity"]
tags: ["sleep", "circadian rhythm", "epidemiology", "aging", "healthspan", "causal inference"]
summary: "Habitual sleep duration and timing, an exposure tied to mortality in large observational cohorts and almost untested as a deliberate intervention to extend life."
updated: "2026-07-27"
humanEvidence: "Large prospective cohorts link both short and long habitual sleep to higher mortality, and short randomized trials of sleep extension have moved metabolic measures; no trial has tested survival."
access: "Free and in principle universal, but bounded by shift schedules, caregiving, noise, light, housing, and untreated sleep disorders, none of which a recommendation reaches."
reversibility: "reversible"
issues: ["Insomnia and hypnotic-drug evidence is summarized in a single paragraph", "No coverage of adolescent sleep timing or school-start-time studies"]
---

```infobox
{
  "caption": "Behavioural and physiological exposure",
  "rows": [
    { "label": "Type", "value": "Behavioural, circadian" },
    { "label": "Consensus adult guidance", "value": "7 or more hours nightly" },
    { "label": "Strongest evidence", "value": "U-shaped mortality association" },
    { "label": "Randomized survival evidence", "value": "None" },
    { "label": "Night shift work", "value": "IARC Group 2A" },
    { "label": "Cost", "value": "Zero" },
    { "label": "Readiness", "value": "TRL 9" }
  ]
}
```

**Sleep and longevity** describes the relationship between how long and how well a person sleeps
and how long that person lives. It is one of the most consistently reproduced associations in
observational epidemiology: across large prospective cohorts, both short and long habitual sleep
track higher all-cause mortality. It is also one of the thinnest interventional literatures in
[[geroscience-hypothesis|geroscience]]. No randomized trial has tested whether making people sleep
more makes them live longer, and the mechanism most often invoked to explain why it should is
currently disputed.

```keyfacts
[
  { "value": "U-shaped", "label": "Duration–mortality curve", "note": "both short and long self-reported sleep track higher mortality in pooled cohorts" },
  { "value": "0", "label": "Sleep-extension trials with a survival endpoint", "note": "as of mid-2026" },
  { "value": "2A", "label": "IARC class for night shift work", "note": "probably carcinogenic to humans; a hazard statement, not a risk magnitude" }
]
```

## The epidemiological signal

A meta-analysis of prospective cohorts published in 2010 established the shape that later work has
mostly confirmed: mortality is elevated among people reporting short sleep, variously defined as
below about six or seven hours, and also among those reporting long sleep, above about eight or
nine.[^cappuccio2010] The curve is not a straight line, and the two tails do not carry the same
interpretation. The relative risks are modest — tens of percent above the reference category rather
than multiples of it — small enough that residual confounding remains a live explanation for either
tail.

The long-sleep tail is widely read as reverse causation. Cancer, heart failure, depression,
infection, and the drugs used to treat them all increase time spent asleep or in bed, so a person
sleeping ten hours is often a person who is ill. Analyses that exclude participants who die early in
follow-up, or that adjust for baseline disease, weaken the long-sleep association without
eliminating it. The short-sleep tail is less easily dismissed but is confounded in a different
direction: short sleep is concentrated among shift workers, carers, people in pain, people with
untreated apnoea, and people in insecure housing. This is where
[[access-and-inequality|unequal access]] enters a topic that looks free at first glance.

Measurement is the quieter problem. Nearly every cohort in the mortality literature rests on a
single self-reported question, and self-reported sleep correlates only moderately with
polysomnography or actigraphy. Accelerometer sub-studies within large biobanks, built on
[[wearable-health-sensors|wrist-worn sensors]], have begun to report mortality associations, but
they are more recent and far smaller than the pooled self-report literature they are meant to check.
Mendelian randomization studies, which use genetic variants associated with sleep traits as
instruments, have generally found more support for insomnia symptoms as a causal contributor to
cardiometabolic disease than for sleep duration itself.

## What sleep is thought to do

Sleep is regulated by two interacting processes: a homeostatic pressure that builds with time awake
and a circadian oscillator that sets when that pressure is discharged. Sleep loss is not benign in
animals. Rats kept continuously awake by the disk-over-water method died within roughly two to
three weeks, with hypermetabolism and failing thermoregulation, and no single proximate cause of
death was ever established.[^rechtschaffen1983] Nothing comparable has been done, or could be done,
in people.

Human experiments run for days rather than weeks and measure physiology rather than survival.
Restricting eleven healthy young men to four hours in bed for six nights impaired glucose tolerance
and altered evening cortisol, effects that were attenuated after a period of sleep
recovery.[^spiegel1999] Later restriction experiments have reproduced the direction of the glucose
finding, and [[continuous-glucose-monitoring|continuous glucose sensors]] have made the same
physiology observable outside a laboratory. A single night of total sleep deprivation in healthy
adults increased PET-measured amyloid-beta signal in the hippocampus and thalamus, a human result
often cited for a sleep–dementia link, though it measures a tracer over one night rather than a
disease over decades.[^shokri2018] Disturbed sleep also appears years before an Alzheimer's
diagnosis, and separating symptom from contributor has proved difficult.

Chronic short sleep is associated with higher circulating inflammatory markers, connecting it to
[[inflammaging]]. What it does not map onto cleanly is the molecular list of
[[hallmarks-of-aging]]: sleep is a whole-organism state rather than a pathway, which is why it sits
awkwardly in a field organized around druggable targets. Sleep studies increasingly report
[[epigenetic-clock|methylation clock]] and other [[aging-biomarkers|biomarker]] readouts, but a
movement in a [[biological-age|composite age estimate]] is a change in a prediction, not an outcome.

### The clearance dispute

In 2013, a study in mice reported that the interstitial space of the brain expanded during natural
sleep and under anaesthesia, and that clearance of injected tracers, including amyloid-beta, rose
accordingly.[^xie2013] Framed as glymphatic clearance, that result became the mechanistic story
behind a decade of coverage asserting that sleep washes the brain.

> [!debate] Does sleep clear the brain, or not?
> A 2024 study in *Nature Neuroscience*, also in mice, used a different tracer approach and reported
> that clearance from brain tissue was *reduced* during sleep and under anaesthesia rather than
> increased.[^miao2024] The two have not been reconciled; they differ in tracer, anaesthetic, and in
> what counts as clearance. Fluid-dynamics modelling has separately questioned whether bulk flow of
> the magnitude the glymphatic model requires is physically plausible. The popular version of the
> mechanism is firmer than the literature it rests on.

## How the evidence accumulated

```timeline
[
  { "year": "1953", "title": "REM sleep described", "text": "Aserinsky and Kleitman report cyclical eye movements during sleep at the University of Chicago, turning sleep from a uniform state into a measurable, staged process." },
  { "year": "1983", "title": "Sleep deprivation kills rats", "text": "Rechtschaffen's group reports that rats prevented from sleeping die within weeks, with hypermetabolism and thermoregulatory failure and no identified proximate cause." },
  { "year": "1999", "title": "Metabolic cost shown in humans", "text": "Spiegel, Leproult and Van Cauter restrict eleven healthy young men to four hours in bed for six nights and report impaired glucose tolerance." },
  { "year": "2007", "title": "Shift work classified as a probable carcinogen", "text": "An IARC working group places shift work involving circadian disruption in Group 2A, tying disturbed sleep timing to cancer risk as a hazard rather than a quantified one." },
  { "year": "2010", "title": "U-shaped mortality curve pooled", "text": "A meta-analysis of prospective cohorts confirms elevated all-cause mortality at both short and long self-reported sleep durations." },
  { "year": "2013", "title": "Glymphatic clearance reported in mice", "text": "Xie and colleagues report expanded interstitial space and increased solute clearance during sleep in mice." },
  { "year": "2016", "title": "CPAP misses a cardiovascular endpoint", "text": "The SAVE trial finds no reduction in cardiovascular events from continuous positive airway pressure in adults with both moderate-to-severe apnoea and established cardiovascular disease, with low average adherence." },
  { "year": "2017", "title": "Nobel for the circadian clock", "text": "Hall, Rosbash and Young share the prize in physiology or medicine for working out the molecular feedback loop of the circadian clock, first in fruit flies." },
  { "year": "2024", "title": "Clearance result challenged", "text": "A study in mice reports reduced rather than increased brain clearance during sleep and anaesthesia, leaving the mechanism unsettled." }
]
```

## The interventional gap

The central honest point about sleep and longevity is that the intervention has never been tested
against the outcome. Randomizing adults to a durable increase in sleep and following them for
decades has not been attempted, and the design problems are severe: sleep cannot be blinded,
adherence cannot be enforced, and the people whose sleep is most compressed are constrained by work
and circumstance rather than by choice.

What exists instead are short randomized trials with surrogate endpoints. A trial of sleep-extension
counselling in adults with overweight and habitual short sleep reported a fall in objectively
assessed energy intake over two weeks, a genuine randomized effect on a metabolic
proxy.[^tasali2022] Cognitive behavioural therapy for insomnia is first-line treatment for the
disorder and reliably improves its symptoms; whether it alters survival or the incidence of
age-related disease has not been established.

> [!caution] The comparator makes the gap visible
> [[exercise-and-aging|Structured exercise]] has a randomized functional endpoint in older adults.
> [[caloric-restriction]] has decades of rodent lifespan data and controlled human trials on
> intermediate measures. Sleep has neither: no comparable mammalian lifespan experiment has
> manipulated sleep as the independent variable, and no human trial has been powered on events. The
> absence is not evidence that sleep does not matter. It is evidence that the confidence of most
> public advice about it is unearned.

Part of the reason is structural. Sleep is free and unpatentable, so no sponsor has a commercial
reason to fund a decade-long outcome trial, the same problem that has stalled generic-drug
geroscience trials of [[metformin]]. Melatonin, sold over the counter as a
[[dietary-supplements|sleep supplement]] in some countries and prescription-only in others, has
better evidence as a circadian phase-shifter than as a sedative, and consumer devices built on
[[non-invasive-neuromodulation|acoustic stimulation]] of slow oscillations have laboratory
electrophysiology behind them and no outcome data.

## Circadian disruption

The strongest institutional statement here is about timing, not duration. The International Agency
for Research on Cancer classified shift work involving circadian disruption as probably carcinogenic
to humans in 2007, and a later working group reaffirmed Group 2A for night shift work, citing
limited evidence in humans, sufficient evidence in experimental animals, and strong mechanistic
evidence.[^iarc2020] The classification concerns hazard, not magnitude: it says night work probably
belongs on the list of things capable of causing cancer, not how much cancer it causes, and the
human epidemiology remains inconsistent across cohorts.

Proposed mechanisms include suppression of nocturnal melatonin by light at night and misalignment
between the central clock and peripheral clocks in liver, gut, and adipose tissue. Shift work is
also entangled with income, smoking, diet, and access to care, which is why the human evidence stays
"limited" rather than "sufficient". Misalignment is a standing problem in
[[space-medicine|spaceflight medicine]] too, where crews in low orbit sleep short and fragmented
against many light–dark transitions a day.

## Disordered sleep

Obstructive sleep apnoea has the clearest disease associations and the clearest interventional test.
Continuous positive airway pressure reliably reduces daytime sleepiness and improves quality of
life. It did not reduce cardiovascular events in the SAVE trial, which randomized adults who had
both moderate-to-severe apnoea and established cardiovascular disease, and in which average use was
around three hours a night.[^mcevoy2016] Whether that is a failure of the therapy or of adherence is
the field's standing argument, and it is the closest thing sleep medicine has to a hard randomized
result.

Observational studies associating hypnotic prescriptions with higher mortality are confounded by
indication severely enough that the finding is not treated as established. Fatal familial insomnia,
a rare inherited prion disease in which sleep is progressively lost and death follows within months
to a couple of years, is sometimes offered as proof that sleep loss kills; the thalamic degeneration
causes both the insomnia and the fatal course, so the case does not isolate sleep.

Running the other way are rare families of natural short sleepers. A mutation in the transcriptional
repressor *DEC2*, found in one such family, tracks with shorter sleep in carriers, and mice
engineered to carry the same mutation also slept less.[^he2009] Mutations in other genes, including
the β1-adrenergic receptor gene *ADRB1*, have since been reported in other short-sleeping families,
whose members sleep roughly four to six hours without the apparent deficits that curtailed sleep
produces in most people. Sleep need, rather than clock time in bed, may be the variable
that matters, and those families are the biological starting point for proposals to engineer
[[sleep-reduction|shorter sleep]]. [[human-hibernation|Induced torpor]] does not substitute:
hibernating mammals arouse periodically and sleep during those arousals.

## Outlook

Three developments would move this subject from association toward evidence. The first is objective
measurement at cohort scale, which consumer sensors and the [[quantified-self]] tradition have made
possible and which would replace a self-reported question with a recorded signal — though those
devices estimate how long a person slept far better than they stage the sleep, and score motionless
wakefulness as sleep, limits set out under [[wearable-health-sensors]]. The second is
randomized trials against intermediate hard endpoints such as incident hypertension, glycaemic
control, falls, or measured cognitive decline, reachable in years rather than decades. The third is
treating disorders rather than exposures, where randomization is ethical and adherence measurable.

Sleep appears in almost every popular account of long-lived populations, including the
[[blue-zones|Blue Zones]] literature whose underlying demographic data have themselves been
questioned, and in public-health arguments for a [[longevity-dividend|longevity dividend]] and
[[compression-of-morbidity|compressed morbidity]]. It deserves that place: the association is
consistent across cohorts, and the short-term human physiology is unambiguous. What is not
established is direction. The possibility the epidemiology cannot exclude is that habitual sleep
duration is mostly a gauge of health rather than a lever on it, and that what is worth testing is
not sleep but the treatment of whatever disturbs it. Until a trial addresses that, claims that better sleep extends
[[healthspan|healthy life]] are extrapolations from correlation.

## See also

- [[exercise-and-aging]]
- [[sleep-reduction]]
- [[healthspan]]
- [[aging-biomarkers]]
- [[inflammaging]]
- [[geroscience-hypothesis]]
- [[human-hibernation]]
- [[blue-zones]]

## References

[^cappuccio2010]: `paper` Cappuccio, F.P. et al. "Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies." *Sleep*, 2010. {Sleep duration is self-reported in nearly every cohort pooled, which is the main limitation of the whole literature.}
[^rechtschaffen1983]: `paper` Rechtschaffen, A. et al. "Physiological correlates of prolonged sleep deprivation in rats." *Science*, 1983. {In rats, using forced waking on a rotating disk; the method also imposes stress, which complicates attribution to sleep loss alone.}
[^spiegel1999]: `paper` Spiegel, K., Leproult, R. and Van Cauter, E. "Impact of sleep debt on metabolic and endocrine function." *The Lancet*, 1999. {Eleven healthy young men over six nights; the design shows short-term physiology, not long-term risk.}
[^shokri2018]: `paper` Shokri-Kojori, E. et al. "Beta-amyloid accumulation in the human brain after one night of sleep deprivation." *PNAS*, 2018. {A human PET study, but a single night in a small sample; it measures tracer binding, not disease.}
[^xie2013]: `paper` Xie, L. et al. "Sleep drives metabolite clearance from the adult brain." *Science*, 2013. {In mice, with much of the clearance measurement made under anaesthesia rather than natural sleep.}
[^miao2024]: `paper` Miao, A. et al. "Brain clearance is reduced during sleep and anesthesia." *Nature Neuroscience*, 2024. {Also in mice; reports the opposite direction to the 2013 result and the two have not been reconciled.}
[^iarc2020]: `report` IARC Monographs Working Group. *Night Shift Work*. IARC Monographs on the Identification of Carcinogenic Hazards to Humans, Volume 124, 2020.
[^tasali2022]: `paper` Tasali, E. et al. "Effect of sleep extension on objectively assessed energy intake among adults with overweight in real-life settings: a randomized clinical trial." *JAMA Internal Medicine*, 2022.
[^mcevoy2016]: `paper` McEvoy, R.D. et al. "CPAP for prevention of cardiovascular events in obstructive sleep apnea." *New England Journal of Medicine*, 2016. {Average adherence was roughly three hours a night, which is the standard objection to reading it as a null result for the therapy.}
[^he2009]: `paper` He, Y. et al. "The transcriptional repressor DEC2 regulates sleep length in mammals." *Science*, 2009. {The human finding came from a single small family; the causal test was made in transgenic mice and flies carrying the same mutation.}

==============================================================================
ARTICLE: somatic-gene-therapy
TITLE:   Somatic gene therapy
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/somatic-gene-therapy
SOURCE:  https://futurehumanwiki.com/raw/somatic-gene-therapy
==============================================================================

---
title: "Somatic gene therapy"
slug: "somatic-gene-therapy"
type: "intervention"
status: "established"
horizon: "present"
trl: 9
categories: ["genetics", "bodies"]
tags: ["gene therapy", "rare disease", "aav", "drug pricing", "clinical trials", "cell therapy"]
summary: "Genetic modification of a living person's body cells to treat disease, producing a change that is not passed on to their children."
updated: "2026-07-27"
humanEvidence: "Licensed products treat inherited retinal dystrophy, sickle cell disease and spinal muscular atrophy in people; durability varies, and clotting-factor expression after haemophilia gene therapy declines over the years after treatment."
access: "Approved in the US and Europe for a short list of rare diseases and blood cancers, delivered at a small number of specialist centres; list prices run from roughly $400,000 to $4.25 million per patient."
reversibility: "context"
issues: ["The count of licensed products and the price range carry no citation.", "The 2025 Elevidys deaths and the dose-ceiling review are stated without a source."]
---

```infobox
{
  "caption": "Class of medical intervention",
  "rows": [
    { "label": "Type", "value": "Genetic medicine" },
    { "label": "First human trial", "value": "1990 (ADA-SCID)" },
    { "label": "First Western approval", "value": "Glybera, 2012" },
    { "label": "Main delivery routes", "value": "AAV, lentivirus, LNP", "link": "/wiki/aav-vectors" },
    { "label": "Heritable", "value": "No" },
    { "label": "List prices", "value": "$0.4m–$4.3m" },
    { "label": "Readiness", "value": "TRL 9 (licensed products)" }
  ]
}
```

**Somatic gene therapy** is the deliberate alteration of genetic material in a living person's body cells — as distinct from their eggs, sperm, or embryos — in order to treat or prevent disease. Because the modified cells lie outside the germ line, the change ends with the patient and is not inherited, which is the line that separates the practice legally and ethically from [[germline-editing]]. As of 2026 more than twenty gene and gene-modified cell therapies hold marketing authorisation in the United States, nearly all of them for rare single-gene disorders or for blood cancers.

```figure
{"key": "crispr-cas9-schematic", "caption": "The PAM motif is the constraint that bites in practice: no PAM near the site you want, no edit there."}
```

## How it works

Three problems have to be solved: what to change, how to get the change into the right cells, and how to make it last. The first is the easiest. Thousands of diseases trace to a defined mutation, and for many the therapeutic logic is simply to supply a working copy of the gene, silence a toxic one, or repair the sequence in place with [[crispr-cas9]], [[base-editing]], or [[prime-editing]].

Delivery is where most programmes fail. Two broad strategies exist. *Ex vivo* therapy removes a patient's cells — usually CD34+ haematopoietic stem cells or T cells — modifies them in a laboratory, and returns them. This gives complete control over the editing step and allows the modified cells to be characterised before infusion, at the cost of an intensive clinical procedure. *In vivo* therapy injects the vector into the patient directly, most often [[aav-vectors|adeno-associated virus]] into the bloodstream, muscle, eye, or spinal fluid, or increasingly [[lipid-nanoparticles]] carrying mRNA to the liver.

Durability depends on what is modified. Editing a stem cell produces a permanent change, because every daughter cell inherits it; the same logic underlies the use of [[induced-pluripotent-stem-cells]] as a starting material for engineered grafts. AAV genomes, by contrast, persist mostly as episomes that are diluted each time a cell divides — a serious limitation in growing children and in tissues that turn over. Genome editing sidesteps this by making the change in the DNA itself and letting the delivery vehicle disappear. Changing gene expression without touching the sequence at all, through [[epigenome-editing]], is a third option now entering trials.

## Development history

```timeline
[
  { "year": "1990", "title": "First authorised human gene transfer", "text": "W. French Anderson, R. Michael Blaese, and colleagues at the US National Institutes of Health infuse retrovirally modified T cells into a four-year-old with ADA deficiency." },
  { "year": "1999", "title": "Jesse Gelsinger dies", "text": "An 18-year-old with partial ornithine transcarbamylase deficiency dies of a systemic inflammatory response four days after receiving a high dose of adenoviral vector at the University of Pennsylvania. Trials are halted and the field contracts for a decade." },
  { "year": "2000–2003", "title": "Cure, then cancer", "text": "Paris and London trials restore immunity in boys with X-linked severe combined immunodeficiency, but several later develop T-cell leukaemia after the retroviral vector integrates near the LMO2 proto-oncogene." },
  { "year": "2012", "title": "Glybera approved in Europe", "text": "The first gene therapy licensed in a Western market, for lipoprotein lipase deficiency. It is priced near one million euros, treats almost nobody, and is withdrawn in 2017." },
  { "year": "2017", "title": "Luxturna", "text": "Voretigene neparvovec, an AAV therapy for RPE65-associated retinal dystrophy, becomes the first in vivo gene therapy approved in the United States, at roughly $850,000 for both eyes." },
  { "year": "2019", "title": "Zolgensma", "text": "A single intravenous AAV9 dose for spinal muscular atrophy is approved at about $2.1 million, resetting expectations for what a one-time medicine can cost." },
  { "year": "2023", "title": "First CRISPR medicine", "text": "The UK authorises exagamglogene autotemcel for both sickle cell disease and beta thalassaemia in November; the US follows for sickle cell disease in December, alongside the lentiviral therapy Lyfgenia." },
  { "year": "2025", "title": "A therapy for one patient", "text": "A bespoke base-editing treatment is designed, manufactured, and dosed for an infant with CPS1 deficiency in roughly six months, the first personalised in vivo editing therapy." }
]
```

The Gelsinger death is the field's defining trauma. Investigations found not only that the adenoviral dose provoked an overwhelming innate immune reaction, but that adverse events in earlier patients had been under-reported and that the consent process had understated the risk.[^raper2003] Funding and enthusiasm collapsed. Recovery came from unglamorous engineering: self-inactivating lentiviral vectors that no longer carried strong enhancers into the genome, AAV serotypes with better tissue targeting, and much more conservative dosing. The X-SCID trials of the early 2000s taught the complementary lesson, curing children of a lethal immunodeficiency and then giving several of them leukaemia through insertional activation of an oncogene.[^hacein2003]

> [!key] Why a 1999 death still shapes practice
> Dose escalation in small cohorts, long-term follow-up requirements measured in years, and intense scrutiny of vector immunogenicity are all direct inheritances from a single fatality in a phase 1 trial of a non-lethal condition. No comparable event has yet reshaped the norms around [[dual-use-research]] in genome engineering.

## What is approved

The licensed products cluster in three groups. *Ex vivo* haematopoietic stem cell therapies treat inherited blood and metabolic disease: [[casgevy]] and Lyfgenia for sickle cell disease, Zynteglo for beta thalassaemia, Skysona for cerebral adrenoleukodystrophy, and Lenmeldy for metachromatic leukodystrophy. Engineered T-cell products — the CAR-T therapies beginning with Kymriah and Yescarta in 2017 — are gene therapies by mechanism, even though they are usually filed under cancer immunotherapy. *In vivo* AAV products cover the eye, the motor neuron, the liver, and skeletal muscle: Luxturna, Zolgensma, Hemgenix and Beqvez for haemophilia B, Roctavian for haemophilia A, and Elevidys for Duchenne muscular dystrophy.

A fourth category does not fit the one-shot model at all. Vyjuvek, a topical herpes-simplex-based vector for dystrophic epidermolysis bullosa, is applied repeatedly to wounds. Redosability changes the economics and the risk profile completely, and much of the current engineering effort in delivery aims at vehicles that can be given more than once.

## Limitations

Immunity is the recurring obstacle. Many adults carry neutralising antibodies against common AAV serotypes and are excluded from treatment outright; those who are treated generate antibodies that make a second dose ineffective. Capsid-specific T cells can destroy transduced liver cells weeks after infusion, which is why transient steroid cover is standard.

Efficacy has often faded. Factor VIII expression after Roctavian declines substantially over the years following treatment, and the durability question now attaches to every in vivo liver programme. In children, dilution of episomal vector as the organ grows means a therapy that works at age one may not still be working at fifteen — a problem with no clean solution short of editing the stem cell compartment.

Safety failures continue. High systemic AAV doses have caused fatal liver failure, complement activation, and thrombotic microangiopathy; deaths in a trial for X-linked myotubular myopathy and, in 2025, among recipients of Elevidys prompted regulators to revisit dose ceilings for systemic administration. Insertional mutagenesis remains a live concern for integrating vectors, and [[crispr-off-target-effects|off-target and on-target structural edits]] are monitored in every editing programme. The bar is high but not the same one applied to heritable modification, where an error propagates to descendants who cannot consent.

## Cost and access

List prices run from roughly $400,000 to $4.25 million per patient. The arithmetic that produces them is straightforward: development costs comparable to an ordinary drug, spread across a few hundred or few thousand eligible patients, with no repeat revenue. Health systems built to pay for chronic medication monthly are poorly designed to pay for a cure once, and outcome-based instalment contracts have spread unevenly. Insurers' interest in who is likely to need such a payment also keeps [[genetic-discrimination]] on the policy agenda.

The commercial results have been harsh. Glybera was withdrawn; bluebird bio pulled Zynteglo out of Europe in 2021 after failing to agree reimbursement, and the company was later taken private at a fraction of its peak valuation; Pfizer discontinued Beqvez for lack of demand. Several therapies that work are sold nowhere, because no viable business can be built around them. That failure mode — an effective medicine withdrawn for commercial reasons — is close to unique to this class and sits at the centre of arguments about [[access-and-inequality]].

> [!stat] Where the patients are
> The largest single-gene disease burdens, sickle cell disease foremost, fall overwhelmingly on countries without apheresis units, transplant wards, or the budget for a seven-figure infusion. Approval in Boston and London does not translate into treatment in Kano or Kinshasa.

## Outlook

The technical direction is clear: move from *ex vivo* to *in vivo*, from viral to non-viral delivery, and from gene addition to precise correction. In vivo editing of blood stem cells, which would remove the need for chemotherapy conditioning altogether, is the single change that would do most to widen access; it has been shown in animals and is not yet a treatment. Personalised therapy for an ultra-rare mutation, demonstrated once in 2025, poses a regulatory question nobody has answered — how to license a medicine that will only ever have one recipient, and who pays for it.[^musunuru2025] Whether the [[precautionary-principle]] as applied to gene transfer has been calibrated correctly is itself contested; the same caution that followed 1999 also kept effective therapies out of reach for years. What is not contested is that a field able to cure inherited blindness but unable to sell the cure has a distribution failure rather than a discovery failure, and no advance in capsid engineering fixes that.

## See also

- [[germline-editing]]
- [[casgevy]]
- [[aav-vectors]]
- [[lipid-nanoparticles]]
- [[gene-therapy-for-aging]]
- [[governance-of-genome-editing]]
- [[access-and-inequality]]
- [[crispr-off-target-effects]]

## References

[^raper2003]: `paper` Raper, S.E. et al. "Fatal systemic inflammatory response syndrome in an ornithine transcarbamylase deficient patient following adenoviral gene transfer." *Molecular Genetics and Metabolism*, 2003. {Written by the University of Pennsylvania investigators; the failures in consent and adverse-event reporting were established by the separate FDA and NIH inquiries.}
[^hacein2003]: `paper` Hacein-Bey-Abina, S. et al. "LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1." *Science*, 2003. {Reports the first two leukaemia cases in the Paris trial; further cases in this and other trials of the same vector design followed.}
[^musunuru2025]: `paper` Musunuru, K. et al. "Patient-Specific In Vivo Gene Editing to Treat a Rare Genetic Disease." *New England Journal of Medicine*, 2025. {A single infant treated with a bespoke base editor. One patient establishes feasibility and cannot establish efficacy, durability, or safety.}

==============================================================================
ARTICLE: space-medicine
TITLE:   Space medicine
PORTAL:  Space & Extreme Environments
URL:     https://futurehumanwiki.com/wiki/space-medicine
SOURCE:  https://futurehumanwiki.com/raw/space-medicine
==============================================================================

---
title: "Space medicine"
slug: "space-medicine"
type: "concept"
status: "established"
horizon: "present"
categories: ["space", "bodies"]
tags: ["spaceflight", "microgravity", "radiation", "bone loss", "physiology", "aging", "countermeasures"]
summary: "The medical discipline concerned with how spaceflight alters human physiology and with keeping crews functional in an environment the body did not evolve for."
updated: "2026-07-27"
issues: ["The chest-CT-per-week comparison for orbital radiation dose carries no source.", "Carotid stiffening and the disputed Apollo cardiovascular analyses are uncited."]
---

```infobox
{
  "caption": "Medical and physiological discipline",
  "rows": [
    { "label": "Scope", "value": "Human physiology in flight" },
    { "label": "Principal stressors", "value": "Weightlessness, radiation, isolation" },
    { "label": "Longest single flight", "value": "437 days (Valeri Polyakov)" },
    { "label": "Bone loss", "value": "~1–1.5% per month" },
    { "label": "Landmark study", "value": "NASA Twins Study, 2019" },
    { "label": "Human subjects flown", "value": "Fewer than a thousand" },
    { "label": "Status", "value": "Operational clinical practice" }
  ]
}
```

**Space medicine** is the branch of medicine that studies how spaceflight changes the human body and designs the countermeasures, selection standards, and clinical procedures that keep crews alive and capable. It is unusual among medical specialties in that its entire evidence base rests on a cohort of fewer than a thousand people, almost all of them screened for exceptional health before flying. The physiological changes it documents — bone loss, muscle wasting, immune dysregulation, vascular stiffening, ocular remodelling — resemble an accelerated version of ordinary [[hallmarks-of-aging|biological aging]], which is why the field has become quietly relevant to biogerontology.

```keyfacts
[
  { "value": "1–1.5%", "label": "Bone mineral density lost per month", "note": "weight-bearing sites, in the absence of full countermeasures" },
  { "value": "437 days", "label": "Longest continuous spaceflight", "note": "Valeri Polyakov aboard Mir, 1994–95" },
  { "value": "~2 h/day", "label": "Exercise prescribed aboard the ISS", "note": "resistive, treadmill and cycle ergometer combined" }
]
```

## Overview

Human physiology is calibrated to a gravitational field of 1 g, a magnetosphere, a 24-hour light cycle, and a microbial and social environment that has been constant for the species' entire history. Orbital flight removes or alters all four at once. The body responds not by breaking down but by adapting — and the adaptations, appropriate to weightlessness, are maladaptive on return to a planet.

Space medicine therefore has two jobs that pull in different directions. Operationally, it must keep a crew functional for the duration of a mission and land them able to walk. Scientifically, it must characterise a set of exposures no ethics board would ever approve as an experiment, using a sample size that would be considered inadequate in any terrestrial trial.

The stakes rise sharply with mission duration and distance. Aboard the International Space Station, a sick crew member can be home within hours and the vehicle sits inside Earth's magnetic field. On a Mars transit neither is true: [[microgravity-adaptation|deconditioning]] runs for years, the [[radiation-hardening-humans|radiation environment]] is the full galactic cosmic ray flux, and evacuation is physically impossible.

## The stressor environment

Four exposures dominate, and they interact rather than add.

**Weightlessness.** Free fall removes the mechanical loading that maintains bone and postural muscle, and removes the hydrostatic gradient that the cardiovascular system is built around. Every organ system that uses gravity as a reference signal — skeletal, vestibular, cardiovascular, ocular — reorganises.

**Ionising radiation.** Low Earth orbit crews receive roughly the equivalent of a chest CT scan per week, mostly trapped protons and secondary particles. Beyond the magnetosphere the exposure changes in kind as well as amount: galactic cosmic rays include high-energy heavy nuclei that deposit dense tracks of ionisation through cells and cannot be shielded away with any practical mass.

**Isolation, confinement, and altered rhythms.** Crews live in a volume smaller than a modest apartment with the same few people, on a lighting cycle set by orbital mechanics rather than by the sun. Sleep is reliably shortened. Ground analogues such as the Mars-500 chamber study and NASA's CHAPEA habitat isolate volunteers for a year or more specifically to study this.

**Distance from definitive care.** There is no surgeon, no blood bank, no imaging beyond ultrasound, and on interplanetary missions no return option. This constrains crew selection, drives the recurring argument over prophylactic surgery such as pre-mission appendectomy, and forces a crew to be able to diagnose and treat itself.

## Physiological effects

### Bone and muscle

Weight-bearing bone demineralises at roughly 1 to 1.5 per cent per month at the hip and lumbar spine, an order of magnitude faster than postmenopausal osteoporosis. Loss is regional, tracking mechanical unloading: the skull is spared or even gains density. Recovery after return is slow and, at some sites, incomplete years later.[^sibonga2007] Released calcium raises the risk of renal stones during flight, a serious concern when definitive urological care is unavailable.

Muscle loss concentrates in the antigravity muscles — soleus, gastrocnemius, spinal extensors — with a shift toward faster, more fatigable fibre types. Modern exercise protocols preserve much of the mass but not all of the function, and the pattern of decline resembles [[stem-cell-exhaustion|age-related sarcopenia]] compressed into months.

### Cardiovascular system and fluid shifts

Within hours of reaching orbit, several litres of fluid redistribute headward. Faces swell, legs thin, nasal congestion is near-universal, and the sense of smell dulls. The body reads the central volume increase as overhydration and excretes it; plasma volume falls by roughly a tenth within days. On landing, that reduced volume plus a blunted baroreflex produces orthostatic intolerance — the reason returning crews are carried from the capsule rather than walking out.

Longer-term findings are more concerning and less settled. Carotid artery stiffening has been documented in flight. One study using ultrasound found stagnant and in one case reversed flow in the internal jugular vein, with a clot identified in a crew member and treated in orbit.[^goebel2019] Whether spaceflight raises long-term cardiovascular mortality is genuinely disputed; analyses of the small Apollo cohort have reached opposite conclusions.

### Eye and brain

Spaceflight-associated neuro-ocular syndrome, or SANS, is the most clinically alarming finding of the ISS era. It comprises optic disc oedema, flattening of the posterior globe, choroidal folds, and a hyperopic refractive shift, and it affects a substantial fraction of long-duration crew members.[^mader2011] Some changes persist for years after landing. The mechanism is unresolved; the leading hypothesis involves chronically altered cerebrospinal fluid dynamics from the headward fluid shift, but no countermeasure has been validated.

Structural MRI shows upward displacement of the brain within the skull, narrowing of the vertex sulci, and expansion of the cerebral ventricles that only partly reverses over months.[^roberts2017] Neurocognitive testing shows modest post-flight decrements in speed and accuracy that recover, though separating true central effects from fatigue and readaptation is difficult.

### Immune function and the microbiome

Immune dysregulation in flight is consistent and well documented: altered T-cell function, shifted cytokine profiles, and reactivation of latent herpesviruses, with Epstein–Barr, varicella-zoster and cytomegalovirus shedding detectable in a large share of crew.[^crucian2018] Reactivation is usually asymptomatic but occasionally produces shingles. The pattern — chronic low-grade immune activation with impaired cell-mediated responses — is close enough to [[inflammaging]] that the two literatures now cite each other.

> [!caution] Small n, healthy volunteers
> Every figure in this article comes from a cohort selected for above-average health, screened repeatedly, and numbering in the hundreds across six decades. Effects smaller than large ones are frequently undetectable, and null results are rarely informative. Space physiology is a field where the mechanism is often better understood than the epidemiology.

## Measuring the astronaut

The NASA Twins Study is the field's most-cited single investigation: Scott Kelly spent close to a year aboard the ISS while his identical twin Mark remained on the ground, with both sampled intensively before, during, and after.[^twins2019] Its findings were more nuanced than the coverage suggested. Most measured changes — gene expression, cytokines, metabolites — moved during flight and returned to baseline within six months. A minority persisted. Telomeres unexpectedly lengthened in flight and shortened rapidly on return, leaving more critically short telomeres than before, a pattern subsequently seen in other astronauts.[^luxton2020] The result cuts against any simple reading of [[telomeres-and-telomerase|telomere length]] as a straightforward aging readout.

The Twins Study established a template that later work extended. The Space Omics and Medical Atlas, published as a package in 2024, aggregated multi-omic data across missions including short commercial flights, and reported that most molecular changes from brief spaceflight resolve within months.[^soma2024] Private missions have widened the subject pool beyond career astronauts, which slightly improves the statistics and considerably complicates the medical standards.

## Spaceflight as an aging model

The comparison to aging is more than rhetorical. Long-duration flight produces bone loss, sarcopenia, immune remodelling, vascular stiffening, insulin resistance, mitochondrial stress, and cognitive slowing over months rather than decades, in genetically diverse subjects, with dense before-and-after sampling and a defined exposure start. That is a study design biogerontology cannot otherwise obtain.

Mitochondrial dysfunction in particular has emerged as a candidate common node across spaceflight tissues, connecting to the [[mitochondrial-dysfunction|mitochondrial theory]] literature. [[epigenetic-clock|Epigenetic clock]] measurements have been applied to astronaut samples, though the sample sizes are too small to support strong conclusions and the clocks' own [[biological-age|biological age]] interpretation remains contested.

The analogy has limits that are easy to elide. Spaceflight deconditioning is largely a disuse phenomenon and largely reversible; aging is neither. Radiation exposure adds a damage mechanism that ordinary aging does not feature at comparable intensity. Treating a returning astronaut and treating an eighty-year-old are not the same clinical problem, and claims that spaceflight "ages" people by a specific number of years are not supported by the [[aging-biomarkers|biomarker]] evidence.

> [!key] Why the field matters beyond spaceflight
> Very few interventions can be tested against a rapid, reversible, well-instrumented model of multi-system decline in humans. Space medicine has one. Countermeasures validated against orbital bone and muscle loss feed directly back into terrestrial disuse osteoporosis, ICU-acquired weakness, and the [[geroscience-hypothesis|geroscience]] programme.

## Countermeasures and clinical practice

Daily loaded exercise is the backbone of the flight regime and, as on Earth, the best-evidenced intervention available — the same conclusion reached in [[exercise-and-aging]]. The hardware and its dose-response data belong to the deconditioning literature. What belongs to space medicine is the prescription: it is written per crew member, adjusted in flight against bone and fitness markers, and carried as a medical order rather than a fitness programme, which is why crew time for it is protected against operational pressure.

Pharmacology is narrower and mostly borrowed from terrestrial practice. Adding an antiresorptive drug to the exercise regime protected bone measurably better than exercise alone in flown crew, which is among the few pharmacological results in the field with in-flight evidence behind it.[^leblanc2013] Nothing comparable exists for SANS, for immune dysregulation, or for radiation. Nutritional standards target vitamin D, energy and protein adequacy, and sodium restriction. Drug stability is a problem of its own: formulations degrade in a warm cabin under chronic irradiation, and no pharmacy has been qualified for a mission of Mars duration.

In-flight care is practised at the level of a well-equipped remote clinic and no further. Crew medical officers are trained to a defined scope rather than being physicians, and the jugular thrombosis described above — managed with the anticoagulant already stocked aboard until a resupply flight could deliver an alternative — is the standard illustration of how thin the margin is. Artificial gravity would address unloading at a stroke and has never flown with a crew; its rotation-rate and Coriolis constraints are treated under microgravity adaptation.

More speculative approaches sit further out. Induced [[human-hibernation|torpor]] has been proposed to reduce consumable demand and psychological load on long transits, with the additional argument that hibernating mammals resist bone and muscle loss. Genetic approaches, including the transfer of [[tardigrade-genes|tardigrade damage-suppressor proteins]] into human cells, belong to the [[pantropy]] programme rather than to current clinical practice, and nothing of the kind has been attempted in a person.

## Limits of the evidence base

The field's central weakness is statistical. Astronaut cohorts are tiny, self-selected, and subject to a strong healthy-worker effect; the NASA Longitudinal Study of Astronaut Health has struggled for decades to detect effects that theory predicts should exist. Bed rest, dry immersion, and hindlimb-unloaded rodents substitute for flight, and each analogue reproduces some stressors while omitting radiation entirely.

Rodent data dominate the radiation literature and are usually generated at dose rates far above anything a crew would experience, using single-ion beams rather than the mixed spectrum of deep space. Extrapolating cognitive deficits from an acutely irradiated mouse to a Mars crew is a long inferential chain, and the field says so.

## Outlook

The decisive test would be a first crewed Mars mission, which combines every stressor at maximum duration with no evacuation option and no prior human data at that exposure. Nothing in the current evidence base establishes that a crew would arrive able to perform surface operations, and the SANS problem in particular has no validated countermeasure after more than a decade of study.

Two questions dominate the research agenda as of 2026. The first is whether partial gravity — 0.38 g on Mars, 0.17 g on the Moon — is protective, threshold-like, or nearly as damaging as zero, a question no experiment has yet answered for humans or for any mammal over a meaningful period. The second is whether the cluster of changes grouped under the accelerated-aging analogy shares a mechanism that could be targeted, or whether it is several independent disuse and damage processes that happen to co-occur.

## See also

- [[microgravity-adaptation]]
- [[radiation-hardening-humans]]
- [[human-hibernation]]
- [[closed-loop-life-support]]
- [[pantropy]]
- [[generation-ship-biology]]
- [[hallmarks-of-aging]]
- [[exercise-and-aging]]

## References

[^twins2019]: `paper` Garrett-Bakelman, F.E. et al. "The NASA Twins Study: A multidimensional analysis of a year-long human spaceflight." *Science*, 2019. {One flown subject compared with his identical twin on the ground, so every result is a case study; the design controls genotype and nothing else.}
[^mader2011]: `paper` Mader, T.H. et al. "Optic disc edema, globe flattening, choroidal folds, and hyperopic shifts observed in astronauts after long-duration space flight." *Ophthalmology*, 2011.
[^sibonga2007]: `paper` Sibonga, J.D. et al. "Recovery of spaceflight-induced bone loss: bone mineral density after long-duration missions as fitted with an exponential function." *Bone*, 2007.
[^roberts2017]: `paper` Roberts, D.R. et al. "Effects of Spaceflight on Astronaut Brain Structure as Indicated on MRI." *New England Journal of Medicine*, 2017.
[^crucian2018]: `paper` Crucian, B.E. et al. "Immune System Dysregulation During Spaceflight: Potential Countermeasures for Deep Space Exploration Missions." *Frontiers in Immunology*, 2018.
[^goebel2019]: `paper` Marshall-Goebel, K. et al. "Assessment of Jugular Venous Blood Flow Stasis and Thrombosis During Spaceflight." *JAMA Network Open*, 2019. {Ultrasound in a small group of ISS crew; the clot was found incidentally during the study rather than as a prospectively measured endpoint.}
[^luxton2020]: `paper` Luxton, J.J. et al. "Temporal Telomere and DNA Damage Responses in the Space Radiation Environment." *Cell Reports*, 2020.
[^leblanc2013]: `paper` LeBlanc, A. et al. "Bisphosphonates as a supplement to exercise to protect bone during long-duration spaceflight." *Osteoporosis International*, 2013.
[^soma2024]: `paper` Overbey, E.G. et al. "The Space Omics and Medical Atlas (SOMA) and international astronaut biobank." *Nature*, 2024. {Much of the new data comes from short commercial flights, so the finding that changes resolve applies to brief missions rather than year-long ones.}

==============================================================================
ARTICLE: speech-neuroprosthesis
TITLE:   Speech neuroprosthesis
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/speech-neuroprosthesis
SOURCE:  https://futurehumanwiki.com/raw/speech-neuroprosthesis
==============================================================================

---
title: "Speech neuroprosthesis"
slug: "speech-neuroprosthesis"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 5
categories: ["cybernetics"]
tags: ["bci", "speech decoding", "als", "anarthria", "neural decoding", "communication"]
summary: "A brain–computer interface that reconstructs intended speech from motor cortex activity, producing text or synthesized voice for people who cannot speak."
updated: "2026-07-27"
humanEvidence: "Published results come from one or a few implanted participants, mostly with ALS or brainstem stroke; no multi-participant efficacy trial has been run and no device is approved anywhere."
access: "Not obtainable: a handful of research implants at academic centres under investigational device exemptions, with no approved or purchasable system."
reversibility: "difficult"
issues: ["The 2024 UC Davis result is given as accuracy only; no words-per-minute figure is stated."]
---

```infobox
{
  "caption": "Assistive neurotechnology",
  "rows": [
    { "label": "Target population", "value": "Anarthria, ALS, brainstem stroke" },
    { "label": "Recording site", "value": "Ventral sensorimotor cortex" },
    { "label": "Sensors used", "value": "ECoG grids, penetrating arrays" },
    { "label": "First sentence decoding", "value": "2021" },
    { "label": "Best reported rate", "value": "~60–80 words per minute" },
    { "label": "Participants per study", "value": "One to a few" },
    { "label": "Regulatory status", "value": "Investigational" }
  ]
}
```

**Speech neuroprosthesis** is a brain–computer interface that recovers intended speech from
neural activity and renders it as text, synthesized voice, or an animated face. It is aimed at
people who retain the intention and the neural machinery for speech but have lost the motor
pathway: those with amyotrophic lateral sclerosis, brainstem stroke, or severe anarthria from other
causes. Between 2021 and 2025 the field moved from decoding a fifty-word vocabulary at a
conversational crawl to near-real-time output over vocabularies of a hundred thousand words.

Crucially, these systems decode attempted articulation, not thought. The signals come from the
region of cortex that would drive the lips, jaw, tongue, and larynx; the decoder recovers a motor
plan, and language models convert that noisy plan into plausible sentences. That distinction
governs both what the technology can do and what privacy risks it does and does not create.

```keyfacts
[
  { "value": "~15 wpm", "label": "First sentence decoding", "note": "2021, 50-word vocabulary, roughly a quarter of words wrong" },
  { "value": "62 wpm", "label": "Penetrating-array result", "note": "2023, Stanford, 125,000-word vocabulary" },
  { "value": "78 wpm", "label": "Surface-array result", "note": "2023, UCSF, with synthesized voice and avatar" }
]
```

## How it works

Speech production has a somatotopic map. Along the ventral part of the precentral gyrus, cortical
populations encode the movements of individual articulators, and this organization persists in
people who have been unable to speak for years — the map does not disappear when the output pathway
fails. That persistence is the enabling fact for the whole field.

A speech neuroprosthesis records from that cortex, extracts features (high-gamma power for
[[ecog-interfaces]], threshold crossings and band power for a penetrating [[utah-array]]), and
passes them to a sequence model that emits phonemes or subword units. A language model then
converts the phoneme stream into words, using the statistics of English to resolve ambiguity that
the neural signal alone leaves open. These are among the few clinical research systems in which a
statistical language model of the kind central to debates about
[[artificial-general-intelligence]] sits inside a control loop with a person's nervous system. The
user attempts to speak, silently or with whatever residual movement remains, and the pipeline runs
in near real time.

Output can be text on a screen, a synthesized voice, or both. Some systems personalize the voice
from recordings made before the person lost speech, and one has driven a digital avatar whose face
moves with the decoded articulation. This is the same [[neural-decoding]] architecture used for
cursor control, applied to a much higher-dimensional output space.

> [!key] The language model does a lot of the work
> Raw phoneme decoding accuracy is well below the headline word accuracy. A substantial share of
> performance comes from the language model choosing the most probable sentence consistent with a
> noisy neural signal. This is legitimate engineering — human listeners do the same thing — but it
> means the reported word error rate is a property of the whole system, and that unusual or
> unpredictable utterances are decoded worse than the aggregate figure suggests.

## Development history

The 2021 report that established feasibility involved a man with anarthria following a brainstem
stroke, implanted with a surface grid over sensorimotor cortex. He produced sentences from a
fifty-word vocabulary at roughly fifteen words per minute, with about a quarter of words
wrong.[^moses2021] Slow, error-prone, and unmistakably speech.

Two 2023 papers changed the scale. A Stanford group using four penetrating arrays in a participant
with ALS reported 62 words per minute over a 125,000-word vocabulary, with a word error rate near
24 percent — and under 10 percent when the vocabulary was restricted to fifty words.[^willett2023] A
UCSF and Berkeley team, using a 253-channel surface array in a woman paralysed by a brainstem
stroke, reported a median 78 words per minute over a 1,024-word vocabulary, together with a
synthesized voice built from pre-injury recordings and an avatar that moved its
face.[^metzger2023]

A 2024 report from a UC Davis and Brown collaboration pushed accuracy rather than speed, reporting
word accuracy near 98 percent over a very large vocabulary in a participant with ALS, with the
system usable from the first session and improving over months of daily self-directed
use.[^card2024] Subsequent work has attacked latency: streaming architectures that emit words as
they are formed rather than at the end of a sentence, and voice synthesis fast enough to preserve
conversational turn-taking and some prosody.[^littlejohn2025]

A related line uses imagined handwriting rather than speech. Decoding attempted pen strokes from
motor cortex produced roughly 90 characters per minute in one participant, which remains the
fastest character-level result and demonstrates that the choice of imagined movement is itself a
design parameter.[^willett2021]

## What has and has not been shown

Shown: fluent, large-vocabulary decoding in individual participants; stable enough performance for
daily use over months; voice synthesis with recognisable identity; decoding from both surface and
penetrating electrodes; usable results from the first day of calibration in at least one case.

Not shown: any of this in more than a handful of people. Nearly every result above comes from a
single participant, and different participants have different implant sites, different disease
stages, and different residual capabilities. There is no multi-participant efficacy trial, no
approved device, no data on how the systems perform in noisy real-world settings with fatigue and
changing medication, and almost no work outside English. Most systems remain tethered through a
percutaneous connector to laboratory hardware.

A further limitation is that these decoders address speech *production*. They do not help people
whose language comprehension or formulation is impaired, which excludes many stroke and dementia
patients. A [[cochlear-implant]] restores an input channel; a speech neuroprosthesis restores an
output channel; neither touches the linguistic system between them.

Within the wider field of [[neuroprosthetics]], speech decoding is unusual in producing an output
that is unambiguously interpretable. Efforts at cognitive prosthetics further upstream — the
hippocampal work described under [[memory-prosthesis]] — report effects that are small, contested,
and difficult to evaluate precisely because there is no equivalent of a transcript to score.

## Inner speech and privacy

The obvious worry — that a device might transcribe thoughts a person did not choose to utter — is
neither science fiction nor imminent. Attempted speech and imagined speech produce overlapping but
distinguishable patterns in motor cortex, and a 2025 study reported that imagined speech is
decodable from the same electrodes, while also demonstrating a keyword-gated mode designed so that
the decoder ignores neural activity until the user internally produces a chosen
password.[^kunz2025]

That is a reassuring engineering answer to a real problem, and it establishes the shape of the
issue: the safeguard has to be designed in, because the signal is there whether or not the system
is asked to use it. The broader questions belong to [[mental-privacy]] and the developing
[[neurorights]] frameworks — who holds the neural recordings, whether decoded text has legal status
as speech, and what happens when a decoder outputs a sentence the user did not intend. Error
correction in a speech prosthesis is not a cosmetic matter; a wrong word attributed to a person who
cannot verbally repudiate it is a distinctive harm.

## Outlook

The near-term path is a pivotal trial in ALS, the population with the clearest need and the
strongest existing results, using a fully implanted wireless system rather than a percutaneous one.
Progress there depends less on decoding accuracy, which is already adequate, than on hardware
longevity, surgical throughput, and the unglamorous work of making a device that a clinician can
implant and a family can support at home — the same constraints that shape every implanted
[[brain-computer-interface]], from the endovascular [[stentrode]] to the penetrating threads of
[[neuralink]].

These systems are also the strongest existing evidence for the modest version of
[[human-ai-merger]]: a machine-learning model sitting between cortex and the world, doing
inferential work the user's damaged nervous system cannot, with the user unable to distinguish
their own contribution from the model's.

The scientific question underneath is how much of language is reachable this way at all. Decoding
articulation works because articulation is motor. Semantic content, the words a person has not yet
converted into a motor plan, is represented in distributed cortical patterns that no implanted
electrode array currently samples, and it is not known whether a device could recover it without
covering far more of the brain than any surgery would justify.

## See also

- [[brain-computer-interface]]
- [[ecog-interfaces]]
- [[utah-array]]
- [[neural-decoding]]
- [[neuroprosthetics]]
- [[mental-privacy]]
- [[neurorights]]
- [[brain-to-brain-interface]]

## References

[^moses2021]: `paper` Moses, D. A. et al. "Neuroprosthesis for decoding speech in a paralyzed person with anarthria." *New England Journal of Medicine*, 2021.

[^willett2023]: `paper` Willett, F. R. et al. "A high-performance speech neuroprosthesis." *Nature*, 2023.

[^metzger2023]: `paper` Metzger, S. L. et al. "A high-performance neuroprosthesis for speech decoding and avatar control." *Nature*, 2023.

[^card2024]: `paper` Card, N. S. et al. "An accurate and rapidly calibrating speech neuroprosthesis." *New England Journal of Medicine*, 2024.

[^willett2021]: `paper` Willett, F. R., Avansino, D. T., Hochberg, L. R., Henderson, J. M. and Shenoy, K. V. "High-performance brain-to-text communication via handwriting." *Nature*, 2021. {The participant had a spinal cord injury rather than a speech disorder, so the result tests decoding rather than the clinical use case for a speech device.}

[^littlejohn2025]: `paper` Littlejohn, K. T. et al. "A streaming brain-to-voice neuroprosthesis to restore naturalistic communication." *Nature Neuroscience*, 2025.

[^kunz2025]: `paper` Kunz, E. M. et al. "Inner speech in motor cortex and implications for speech neuroprostheses." *Cell*, 2025. {The password-gated mode is a laboratory demonstration in already implanted research participants, not a safeguard validated in any deployed device.}

==============================================================================
ARTICLE: stem-cell-exhaustion
TITLE:   Stem cell exhaustion
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/stem-cell-exhaustion
SOURCE:  https://futurehumanwiki.com/raw/stem-cell-exhaustion
==============================================================================

---
title: "Stem cell exhaustion"
slug: "stem-cell-exhaustion"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity", "bodies"]
tags: ["aging", "stem cells", "regeneration", "clonal hematopoiesis", "niche", "mechanisms"]
summary: "The age-related decline in the number and regenerative function of adult stem cells across blood, muscle, gut and skin, and the loss of repair capacity that follows."
updated: "2026-07-27"
humanEvidence: "Human sequencing establishes clonal hematopoiesis after 60 and a sharp fall in blood clonal diversity after about 70; the senolytic and reprogramming rejuvenation results are from mice."
issues: ["The two conflicting 2018 human hippocampal neurogenesis studies are described without citations."]
---

```infobox
{
  "caption": "Hallmark of aging",
  "rows": [
    { "label": "Tier", "value": "Integrative hallmark" },
    { "label": "Tissues most affected", "value": "Blood, muscle, gut, skin" },
    { "label": "Blood signature", "value": "Myeloid bias, clonal hematopoiesis" },
    { "label": "Causes", "value": "Cell-intrinsic and niche-derived" },
    { "label": "Clinical marker", "value": "CHIP, detectable by sequencing" },
    { "label": "Established therapy", "value": "Haematopoietic stem cell transplant" }
  ]
}
```

**Stem cell exhaustion** is the progressive failure of adult stem and progenitor cells to maintain and repair the tissues they serve. It is classified as an integrative hallmark in the [[hallmarks-of-aging]] framework, meaning it is understood as a consequence of upstream damage that becomes a cause of visible decline in its own right. The term is slightly misleading: in several tissues the number of stem cells stays constant or rises with age while their per-cell function falls.

```keyfacts
[
  { "value": "~70", "label": "Age at which blood clonal diversity falls sharply", "note": "from whole-genome sequencing of single human stem cell colonies" },
  { "value": "6", "label": "Genes carrying most CHIP driver mutations", "note": "DNMT3A, TET2, ASXL1, JAK2, TP53, PPM1D" },
  { "value": "2", "label": "Sources of decline", "note": "cell-intrinsic damage and a changed niche" }
]
```

## Blood

The haematopoietic system is the best-characterized case because its stem cells can be purified, counted, and functionally tested by transplantation. Aged mouse haematopoietic stem cells are more numerous than young ones but reconstitute an irradiated recipient less well, and they produce a skewed output: more myeloid cells, fewer lymphoid.[^rossi2005] The human correlate is a shrinking naive T-cell pool, weakened responses to new vaccines, and thymic involution that begins in childhood and is largely complete by middle age.

### Clonal hematopoiesis

The most consequential discovery in this area is clonal hematopoiesis of indeterminate potential, or CHIP: the expansion of a single mutant stem cell clone to a detectable fraction of blood cells in people with no blood disorder. Two 2014 studies identified it in large sequencing datasets, found that it becomes common after age 60, and showed that carriers face elevated risks of haematological malignancy and, unexpectedly, of death from cardiovascular causes.[^jaiswal2014][^genovese2014] The driver mutations cluster in a small set of genes, chiefly *DNMT3A*, *TET2*, *ASXL1*, *JAK2*, *TP53* and *PPM1D* — mostly epigenetic regulators and DNA damage response genes.

The cardiovascular association is thought to run through inflammation: macrophages descended from a *TET2*-mutant clone secrete more interleukin-1β and interleukin-6, accelerating atherosclerotic plaque development in mice. That connects CHIP directly to [[inflammaging]] and offers one of the few cases where a specific somatic mutation in an aged human is linked to a specific inflammatory mechanism and a specific disease.

Whole-genome sequencing of individual stem cell colonies has since shown that the change is not confined to people with detectable CHIP. Clonal diversity in human haematopoiesis stays high through midlife and then falls sharply after about age 70, with the blood of the very old produced by a small number of expanded clones.[^mitchell2022] The proximate cause appears to be a shift in the balance between drift and selection rather than a sudden loss of stem cells.

> [!stat] By the numbers
> CHIP is detectable by standard sequencing in a minority of people in their sixties and in a substantially larger fraction over 70. With ultra-sensitive assays, some degree of clonal expansion can be found in nearly everyone past middle age, which makes the clinical threshold a matter of convention rather than biology.

## Muscle

Skeletal muscle repair depends on satellite cells, quiescent progenitors under the basal lamina of each fibre. Their number declines modestly with age; their capacity to activate, proliferate and fuse declines more. Some geriatric satellite cells lose quiescence not by activating but by entering [[cellular-senescence|senescence]], a switch associated with de-repression of the p16INK4a locus.[^sousa2014] Sarcopenia — the loss of muscle mass and strength with age — is the visible outcome, and it is the single functional decline most responsive to [[exercise-and-aging|resistance training]].

## Gut, skin and pigment

Intestinal epithelium turns over every few days, driven by Lgr5-expressing crypt base columnar cells supported by neighbouring Paneth cells. Aged mouse intestinal stem cells form organoids less efficiently, and both the stem cells and their niche contribute. A day of fasting improves their function in young and old mice through a shift to fatty acid oxidation, which is one of the more direct demonstrations that [[caloric-restriction|dietary restriction]] acts on a stem compartment.[^mihaylova2018]

Hair greying is the most visible instance of the hallmark. It results from incomplete maintenance of melanocyte stem cells in the hair follicle bulge, which are progressively lost rather than merely inactivated.[^nishimura2005]

## Nervous system

Whether the adult human brain generates new neurons at all remains disputed. Two studies published in 2018 reached opposite conclusions from human hippocampal tissue, one finding that neurogenesis becomes undetectable in adults and the other that it persists into old age; subsequent work has reported abundant immature neurons in healthy adults with a sharp decline in Alzheimer's disease. The disagreement turns substantially on tissue fixation and antibody specificity. Rodent hippocampal neurogenesis clearly declines with age. Whether that finding carries over to humans is precisely what the conflicting human studies leave open.

## Niche versus cell-intrinsic causes

Heterochronic experiments separate the two contributions. Transplanting old haematopoietic stem cells into young recipients does not restore balanced lineage output, indicating that blood stem cell aging is largely cell-intrinsic and written into the epigenome. Exposing old muscle to a young systemic environment through [[parabiosis-and-young-blood|heterochronic parabiosis]] substantially restores satellite cell activation, indicating that muscle stem cell aging is substantially extrinsic.[^conboy2005] Both results are from mice, and the parabiosis literature has since become entangled in disputes over which circulating factor, if any, is responsible.

Intrinsic causes include accumulated DNA damage, [[telomeres-and-telomerase|telomere attrition]] in high-turnover compartments, epigenetic drift, declining [[autophagy]] that leaves damaged organelles to be inherited at division, and [[mitochondrial-dysfunction|mitochondrial]] impairment. Niche causes include fibrosis and matrix stiffening, altered Wnt and Notch signalling, marrow adiposity, and the secretory output of accumulated senescent cells.

## Interventions

Haematopoietic stem cell transplantation is a mature therapy, but it replaces a compartment rather than rejuvenating one, and it carries conditioning toxicity that makes it unsuitable for anything short of life-threatening disease. Among experimental approaches, [[senolytics|senolytic]] clearance improves muscle stem cell function in aged and progeroid mice, and [[partial-reprogramming|transient reprogramming]] restores regenerative capacity in mouse muscle and other tissues. Neither has been shown to improve regeneration in an aged human.

> [!caution] Unproven clinics
> A large international market sells "stem cell therapy" for aging, joint pain and neurological disease, typically using autologous adipose-derived cells that have not been shown to engraft or differentiate. Regulators including the US Food and Drug Administration have taken enforcement action against several such providers. Three patients suffered severe and permanent bilateral vision loss after intravitreal injection of adipose-derived cells at one Florida clinic, a case series reported in the *New England Journal of Medicine* in 2017.[^kuriyan2017]

## Open problems

The framing question is whether stem cell decline is a cause worth targeting or a readout of everything upstream of it. If aged stem cells fail mainly because their epigenomes have drifted, [[epigenetic-reprogramming|reprogramming]] is the appropriate intervention. If they fail mainly because the niche has become fibrotic and inflamed, the target is the niche. If clonal selection is the dominant process in blood, then the goal is not to restore stem cell function but to prevent the expansion of clones with a fitness advantage — a problem closer to cancer prevention than to regeneration, and one that no current intervention addresses.

## See also

- [[hallmarks-of-aging]]
- [[cellular-senescence]]
- [[parabiosis-and-young-blood]]
- [[partial-reprogramming]]
- [[induced-pluripotent-stem-cells]]
- [[inflammaging]]

## References

[^rossi2005]: `paper` Rossi, D. J. et al. "Cell intrinsic alterations underlie hematopoietic stem cell aging." *PNAS*, 2005.
[^jaiswal2014]: `paper` Jaiswal, S. et al. "Age-related clonal hematopoiesis associated with adverse outcomes." *New England Journal of Medicine*, 2014.
[^genovese2014]: `paper` Genovese, G. et al. "Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence." *New England Journal of Medicine*, 2014. {The clonal expansions turned up in exome data collected for a schizophrenia study, so the cohort was never assembled to study blood or aging.}
[^mitchell2022]: `paper` Mitchell, E. et al. "Clonal dynamics of haematopoiesis across the human lifespan." *Nature*, 2022. {Reconstructed from whole-genome sequencing of colonies grown from a small number of donors across the age range, not from following anyone over time.}
[^sousa2014]: `paper` Sousa-Victor, P. et al. "Geriatric muscle stem cells switch reversible quiescence into senescence." *Nature*, 2014.
[^mihaylova2018]: `paper` Mihaylova, M. M. et al. "Fasting activates fatty acid oxidation to enhance intestinal stem cell function during homeostasis and aging." *Cell Stem Cell*, 2018.
[^nishimura2005]: `paper` Nishimura, E. K., Granter, S. R., Fisher, D. E. "Mechanisms of hair graying: Incomplete melanocyte stem cell maintenance in the niche." *Science*, 2005.
[^conboy2005]: `paper` Conboy, I. M. et al. "Rejuvenation of aged progenitor cells by exposure to a young systemic environment." *Nature*, 2005.
[^kuriyan2017]: `paper` Kuriyan, A. E. et al. "Vision loss after intravitreal injection of autologous 'stem cells' for AMD." *New England Journal of Medicine*, 2017.

==============================================================================
ARTICLE: synthetic-embryos
TITLE:   Stem-cell-based embryo models
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/synthetic-embryos
SOURCE:  https://futurehumanwiki.com/raw/synthetic-embryos
==============================================================================

---
title: "Stem-cell-based embryo models"
slug: "synthetic-embryos"
type: "technology"
status: "experimental"
horizon: "late 2020s"
trl: 3
categories: ["reproduction", "genetics"]
tags: ["embryo models", "stem cells", "developmental biology", "14-day rule", "bioethics", "regulation"]
summary: "Structures grown by aggregating stem cells that reproduce stages of early embryonic development without fertilisation, and the regulatory categories they fall outside."
updated: "2026-07-27"
humanEvidence: "Nothing has been done in a person. Human work is confined to cultured cells, and no embryo model in any species has produced a live animal after transfer to a uterus."
access: "Not a product and not for sale: the models are made in research laboratories from pluripotent stem cell lines, and none is offered clinically or reproductively."
issues: ["The 2023 human integrated embryo model reports are described without citations.", "The 2016 thirteen-day human embryo culture result needs a citation."]
---

```infobox
{
  "caption": "Developmental biology research model",
  "rows": [
    { "label": "Also called", "value": "SCBEMs, embryoids" },
    { "label": "Main types", "value": "Blastoids, gastruloids, integrated models" },
    { "label": "Built from", "value": "Pluripotent stem cells", "link": "/wiki/induced-pluripotent-stem-cells" },
    { "label": "First human blastoids", "value": "2021" },
    { "label": "Live offspring produced", "value": "None, in any species" },
    { "label": "Regulatory status", "value": "Outside most embryo statutes" }
  ]
}
```

**Stem-cell-based embryo models** are structures assembled from cultured stem cells that reproduce features of early embryonic development without an egg, a sperm, or fertilisation. They are widely called synthetic embryos, a term nearly every researcher in the field objects to. None has produced a live animal in any species, and the most advanced human versions self-organise for a period roughly equivalent to the first two weeks of development before deteriorating. Their significance is that they fall outside the legal definition of an embryo almost everywhere, while resembling one closely enough that the exemption has become difficult to defend.

## What they are

All of them start from pluripotent cells: embryonic stem cell lines, or [[induced-pluripotent-stem-cells]] made by [[yamanaka-factors|factor-based reprogramming]] of adult tissue. The models divide by which parts of the conceptus they include.

- **Blastoids** model the blastocyst. Mouse versions were made in 2018 by combining embryonic and trophoblast stem cells; human blastoids followed in 2021 from two directions, one starting from naive pluripotent cells and one arising during the reprogramming of fibroblasts.[^rivron2018][^yu2021] They form an outer trophectoderm-like layer, an inner cell mass–like cluster, and a cavity, and some attach to cultured endometrial cells in a way that resembles the first step of implantation.
- **Gastruloids** model the body plan without extraembryonic tissue. Aggregates of pluripotent cells, given a pulse of Wnt signalling, break symmetry, establish an anterior–posterior axis, and express the segmentation gene programme in the correct order. They have no yolk sac, no placenta, and no head, and are not blastocyst-derived, so they cannot be confused with an embryo at any stage.
- **Integrated post-implantation models** are the contentious category. Mouse versions reported in 2022 by two groups combined embryonic stem cells with cells pushed toward trophoblast and primitive endoderm fates, and produced structures with a beating heart tube, a neural tube, somites, a gut, and a yolk sac.[^tarazi2022] Human counterparts reported in 2023 recapitulated aspects of the amniotic cavity, yolk sac, and bilaminar disc to about the fourteenth day of development.

Efficiency is low throughout. Typically a small percentage of aggregates form structures with correct morphology, the rest producing disorganised tissue, which is itself a reason to doubt that the successful ones are following the same developmental logic as an embryo rather than arriving at a similar appearance by another route.

> [!note] Terminology
> "Synthetic" is doubly misleading: nothing is chemically synthesised, and the cells are ordinary human or mouse cells. "Embryo model" is the term used by the International Society for Stem Cell Research and in the UK code of practice. Popular coverage in 2022 and 2023 that described mouse models as embryos "created without eggs or sperm" prompted public corrections from several of the researchers involved.

## Why the field wants them

The period between implantation and the fourth week of human gestation is the least understood phase of human development, for a straightforward reason: it happens inside a uterus, is inaccessible, and is when a large share of pregnancies fail. Donated IVF embryos are scarce, ethically constrained, and cannot legally be cultured past fourteen days in most jurisdictions.

Embryo models supply an alternative that is renewable, genetically tractable, and produced in quantity. Concrete applications include studying the causes of early miscarriage, which rise steeply with maternal age for reasons examined in [[reproductive-longevity]]; mapping the signalling that specifies the body axes; testing whether a drug or environmental exposure is teratogenic at a stage no other model reaches; and examining the origin of congenital anomalies. They also intersect with [[embryo-selection]], since much of what a clinic infers from an embryo's morphology rests on developmental assumptions that models can now test directly.

The models are close cousins of [[organoids]], which self-organise from stem cells into tissue-like structures, and share both their strengths and their ceiling: real spatial organisation, real gene expression programmes, no vasculature, and limited maturity.

## The fourteen-day rule and its erosion

The fourteen-day limit on culturing human embryos originates in the 1984 Warnock report in the United Kingdom and was written into the Human Fertilisation and Embryology Act 1990.[^warnock1984] It was chosen because the primitive streak appears around day fourteen, after which twinning is no longer possible and individuation can be said to have begun. It was also chosen because at the time nobody could culture an embryo anywhere near that long, which made it a rule that cost nothing. The rule has since been adopted, formally or in practice, in most countries that fund human embryo research, and it is one of the few pieces of bioethical line-drawing that survived four decades intact.

Two developments undermined it. Human embryos were cultured to around thirteen days in 2016, making the limit binding for the first time. And embryo models arrived, which are not embryos under the statutory definitions and are therefore not covered by the limit at all. The UK Act defines an embryo as a live human embryo "where fertilisation is complete"; a structure that was never fertilised does not qualify.

The International Society for Stem Cell Research revised its guidelines in 2021, moving culture beyond fourteen days from a prohibited category into one requiring case-by-case specialised review and public consultation, and placing integrated embryo models under oversight while leaving non-integrated models such as gastruloids largely unrestricted.[^isscr2021] In 2024 a UK-based group published a voluntary code of practice for stem-cell-based embryo models with an independent oversight committee, an instance of the self-governance approach whose founding precedent is the [[asilomar-conference]] and whose weakness is the same: it binds only those who agree to be bound. Whether a research community should be able to relax a limit it originally proposed is the recurring objection, and it is the point at which arguments from the [[precautionary-principle]] enter the debate.

> [!debate] What would make a model an embryo
> Proposals for a bright line fall into three families. A *provenance* test asks whether fertilisation occurred, which is clear but arbitrary. A *morphological* test asks whether the structure has the parts of an embryo, which is vague. A *potentiality* test asks whether it could develop into a person if transferred, which is the criterion most philosophers regard as morally relevant and the hardest to apply, since testing it directly requires the transfer nobody permits.

## Limitations

No embryo model has produced a live animal. Mouse models transferred to a uterus have implanted in some experiments and then failed. Whether this reflects a fixable technical deficit or a categorical difference between a model and an embryo is unresolved, and it is the single most consequential open question in the field, because a model that could develop to term would collapse the regulatory distinction entirely.

Fidelity is uneven. Single-cell transcriptomic comparisons with real primate and human embryos show that some cell types in models match their in vivo counterparts closely and others do not, with extraembryonic lineages generally poorer matches than embryonic ones. Timing is often wrong, and models frequently produce cells in the wrong proportions. The structures also lack a maternal interface, so nothing about implantation biology that depends on the endometrium can be modelled properly, which is a limitation shared with attempts at [[ectogenesis]] and every version of the [[artificial-womb]].

## Ethics and governance

The moral-status question is unavoidable and unsettled. If moral status attaches to a developmental potential rather than to an origin, then a model that grows better will at some point acquire whatever status an embryo has, and the field will have arrived there gradually and without a decision. If status attaches to fertilisation, models never acquire it however lifelike they become. Neither position is fringe, and the disagreement is not empirical.

Downstream concerns are more tractable. Consent from the donors of the original stem cell lines rarely mentioned embryo models, since the lines predate them. Public trust is affected by terminology, and the field has been damaged by coverage it did not write; the reception of [[human-cloning]] in the late 1990s is the standing example of how a research programme can be defined by a word it did not choose. Transfer to a human or animal uterus is prohibited under every existing guideline, and the enforcement question resembles the one described in [[governance-of-genome-editing]]: rules that bind funded academic laboratories in a handful of countries do not bind a well-resourced private effort elsewhere.

A further prospect complicates the picture. If [[in-vitro-gametogenesis]] succeeds, gametes could be made from stem cells and used to create real embryos, at which point the interesting entity is unambiguously an embryo and the model debate becomes moot for reproductive purposes. Some researchers have suggested models could instead be a source of tissue, connecting them to [[lab-grown-organs]] and interspecies chimera work, though nothing at the current level of organisation supports that.

The near-term question is narrower than the philosophy suggests: whether funders and journals will require embryo-model work to pass through review bodies designed for embryo research, and whether a voluntary code with no statutory backing survives contact with a laboratory that decides not to participate.

## See also

- [[embryo-selection]]
- [[in-vitro-gametogenesis]]
- [[organoids]]
- [[induced-pluripotent-stem-cells]]
- [[ectogenesis]]
- [[governance-of-genome-editing]]
- [[human-cloning]]
- [[artificial-womb]]
- [[uterus-transplantation]]

## References

[^rivron2018]: `paper` Rivron, N. C. et al. "Blastocyst-like structures generated solely from stem cells." *Nature*, 2018.
[^yu2021]: `paper` Yu, L. et al. "Blastocyst-like structures generated from human pluripotent stem cells." *Nature*, 2021. Published alongside Liu, X. et al., "Modelling human blastocysts by reprogramming fibroblasts into iBlastoids," *Nature*, 2021.
[^tarazi2022]: `paper` Tarazi, S. et al. "Post-gastrulation synthetic embryos generated ex utero from mouse naive ESCs." *Cell*, 2022. {Mouse cells only, and the paper's own title uses the word synthetic that most of the field now objects to.}
[^isscr2021]: `report` International Society for Stem Cell Research. *Guidelines for Stem Cell Research and Clinical Translation*, 2021, subsequently revised in its provisions on stem-cell-based embryo models. {Guidelines written by the research community they govern; they carry no statutory force and bind only laboratories that choose to follow them.}
[^warnock1984]: `report` Warnock, M. *Report of the Committee of Inquiry into Human Fertilisation and Embryology*. HMSO, 1984.

==============================================================================
ARTICLE: stentrode
TITLE:   Stentrode
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/stentrode
SOURCE:  https://futurehumanwiki.com/raw/stentrode
==============================================================================

---
title: "Stentrode"
slug: "stentrode"
type: "technology"
status: "experimental"
horizon: "late 2020s"
trl: 6
categories: ["cybernetics"]
tags: ["bci", "endovascular", "implants", "paralysis", "als", "neurotechnology"]
summary: "An endovascular brain–computer interface in which a stent carrying electrodes is delivered through the jugular vein into a vein overlying the motor cortex."
updated: "2026-07-27"
humanEvidence: "A small number of people with severe paralysis, mostly from ALS, have used the implant for discrete computer control over months to years; the published safety series covers four Australian participants."
access: "Investigational only: implants have been placed solely within the SWITCH and COMMAND feasibility studies at a handful of centres, with no market authorisation anywhere and nothing on sale."
reversibility: "difficult"
issues: ["The COMMAND primary safety endpoint is attributed to the company without a citation.", "The 2023-2025 platform integrations are described without sources."]
---

```infobox
{
  "caption": "Endovascular neural interface",
  "rows": [
    { "label": "Developer", "value": "Synchron", "link": "/wiki/synchron" },
    { "label": "Originated", "value": "University of Melbourne" },
    { "label": "First animal report", "value": "2016" },
    { "label": "First human implant", "value": "2019 (Australia)" },
    { "label": "Access route", "value": "Jugular vein to sagittal sinus" },
    { "label": "Electrodes", "value": "16" },
    { "label": "Regulatory status", "value": "Investigational" }
  ]
}
```

**Stentrode** is a brain–computer interface built into a self-expanding stent. Rather than opening
the skull, a neurointerventional radiologist threads the device through the jugular vein and into
the superior sagittal sinus, the large vein running along the midline of the brain, where it
expands against the vessel wall directly above motor cortex. Electrodes on the stent record cortical
activity through the vein wall; a lead runs under the skin of the neck to a telemetry unit implanted
in the chest, which transmits wirelessly to an external computer.

The design trades signal quality for surgical accessibility. With sixteen electrodes
recording population-level activity rather than individual neurons, the Stentrode captures far less
information than a penetrating [[utah-array]]. It also requires no craniotomy, uses catheter
techniques that thousands of clinicians already practise, and does not place foreign material in
the brain parenchyma.

## How it works

The stent is a nitinol mesh of the type used for treating venous stenosis, with electrode contacts
mounted on its struts. Delivered collapsed inside a catheter, it self-expands on release and
presses the electrodes against the endothelium. Over the following weeks the vessel wall
incorporates the stent — the same endothelialization that makes vascular stents stable, and which
here fixes the electrodes in position without sutures or bone anchors.

Signals reach the electrodes through the vein wall, which acts as a spatial filter. The result
resembles what surface arrays record in [[ecog-interfaces]] but with fewer, larger, and more
distant contacts: local field potentials and band power, not action potentials. Participants
generate control signals by attempting movements — typically of a foot or ankle, since the leg
representation of motor cortex sits along the midline nearest the sagittal sinus — and a decoder
maps the resulting change in band power to a discrete output the company calls a digital motor
output, essentially a click.

Because one reliable click is not enough to run a computer, the system is used with eye tracking or
switch-scanning software: gaze selects the target, the neural signal confirms it. This
division of labour is deliberate. The [[neural-decoding]] problem is much easier when the decoder
only has to distinguish attempted movement from rest.

```compare
{
  "columns": ["Stentrode", "Penetrating microelectrode array"],
  "rows": [
    { "label": "Procedure", "values": ["Catheter via jugular vein", "Craniotomy and cortical insertion"] },
    { "label": "Signal type", "values": ["Local field potentials", "Single-unit spikes and LFP"] },
    { "label": "Electrodes", "values": ["16", "96–1,024"] },
    { "label": "Accessible cortex", "values": ["Tissue adjacent to large veins", "Any surgically reachable area"] },
    { "label": "Demonstrated output", "values": ["Discrete selection", "Continuous multi-dimensional control, speech"] },
    { "label": "Removal", "values": ["Difficult once endothelialized", "Possible, at the cost of repeat surgery"] }
  ]
}
```

## Development history

The concept came from Thomas Oxley and Nicholas Opie's group at the University of Melbourne, who
reasoned that the cerebral venous system already reaches within millimetres of cortex and that
catheter delivery would avoid the complication profile of open neurosurgery. A 2016 report in sheep
showed that an endovascular array could record cortical activity chronically, with signal quality
that remained usable as the stent incorporated into the vessel wall.[^oxley2016] [[synchron]] was
founded to commercialize the device.

The first-in-human study, SWITCH, implanted the device in four participants with severe paralysis,
mostly from amyotrophic lateral sclerosis, at the Royal Melbourne Hospital beginning in
2019. The published results reported no device-related serious adverse events, no vessel occlusion,
and participants able to use the system for texting, email, and online banking in combination with
eye tracking.[^oxley2021][^mitchell2023] A United States early feasibility study, COMMAND, began in
2022 with the first implant at Mount Sinai in New York, and the company has reported that its
primary safety endpoint was met.

```timeline
[
  { "year": "2016", "title": "Chronic recording in sheep", "text": "An endovascular stent-electrode array records cortical activity for months, published in Nature Biotechnology." },
  { "year": "2019–2021", "title": "SWITCH first-in-human study", "text": "Four participants with severe paralysis receive implants in Australia; the trial reports no device-related serious adverse events." },
  { "year": "2022", "title": "First US implant", "text": "The COMMAND early feasibility study begins at Mount Sinai under an FDA investigational device exemption." },
  { "year": "2023–2025", "title": "Integration with consumer platforms", "text": "Synchron demonstrates control of commercial voice assistants and announces work with major computing platforms on native input support for neural devices." }
]
```

## Clinical results and limits

What participants have demonstrated is real but narrow: reliable discrete selection sufficient for
communication and device control, sustained over months to years, in people who otherwise depend on
eye-gaze systems that fail when eye movement deteriorates. That is a meaningful clinical benefit
for a population with few options.

What has not been demonstrated is continuous high-dimensional control. Nothing resembling the
robotic-arm work or the fluent [[speech-neuroprosthesis]] results obtained with penetrating
arrays[^willett2023] has come from an endovascular device, and there is a physical reason to expect
that gap to persist. Speech motor cortex lies on the lateral surface of the brain, far from the sagittal
sinus; the venous anatomy determines which cortex is reachable, and it does not reach everywhere.
Driving a wheelchair or a powered [[exoskeleton]] would likewise require continuous multi-axis
control that a click cannot supply.

> [!debate] Bandwidth versus access
> Synchron's argument is that a device a vascular surgeon can implant in an ordinary catheter lab
> scales to a patient population that craniotomy never will, and that reliable simple control
> covers most of the clinical need. Critics respond that sixteen field-potential channels cap the
> device below the applications that justify an implant at all, and that the comparison should be
> against non-invasive eye trackers rather than against penetrating arrays.

Other limitations follow from the vascular setting. The device is difficult to remove once the vein
wall has grown over it. Participants need antiplatelet therapy, at least initially, and venous
thrombosis or stenosis remains the principal safety concern even though the published studies did
not observe it. Long-term stability data extend to a few years across a small number of
participants, which is not enough to characterize rare complications.

## Strategy and context

Synchron has positioned the Stentrode against both invasive and non-invasive alternatives. Compared
with [[neuralink]] and other penetrating systems, it accepts a much lower ceiling in exchange for a
procedure that generalists can perform. Compared with [[non-invasive-neuromodulation]] and
scalp-electrode devices, it offers a stable, always-available signal that does not require setup or
tolerate hair and sweat artefacts.

The company has also pursued integration with mainstream computing platforms, so that a neural
device appears to an operating system as an input source alongside touch and voice. If that
standardization takes hold, it would matter beyond any one device: the persistent complaint from
users of assistive [[neuroprosthetics]] is that the hardware works and the software ecosystem does
not.

## Outlook

The obvious next step is a pivotal trial in a defined indication — loss of communication in ALS
being the clearest — that could support market authorization. No implanted BCI has cleared that
bar anywhere, and the Stentrode's low channel count may actually help, because a simpler device
with a simpler claim is easier to evaluate. The regulatory precedents that exist are for
stimulating devices such as the [[cochlear-implant]] and [[deep-brain-stimulation]] systems, where
the therapeutic claim is a change in symptoms rather than a measured rate of communication.

The unresolved technical question is whether channel count can grow. Multiple stents, longer arrays
extending further along the sinus, and improved electrode materials have all been proposed. Whether
any of that lifts an endovascular interface out of the switch-like regime, or whether the vein wall
imposes a hard ceiling on what can be resolved, has not been established. As with every implanted
neural device, the governance questions — data ownership, [[mental-privacy]], device support
obligations, the [[neurorights]] frameworks now emerging in several jurisdictions — arrive before
the clinical ones are settled.

## See also

- [[synchron]]
- [[brain-computer-interface]]
- [[utah-array]]
- [[ecog-interfaces]]
- [[neuralink]]
- [[neural-decoding]]
- [[neuroprosthetics]]
- [[speech-neuroprosthesis]]

## References

[^oxley2016]: `paper` Oxley, T. J. et al. "Minimally invasive endovascular stent-electrode array for high-fidelity, chronic recordings of cortical neural activity." *Nature Biotechnology*, 2016. {Recordings in freely moving sheep, assessing signal quality and stability over months rather than any decoded control task.}

[^oxley2021]: `paper` Oxley, T. J. et al. "Motor neuroprosthesis implanted with neurointerventional surgery improves capacity for activities of daily living tasks in severe paralysis: first in-human experience." *Journal of NeuroInterventional Surgery*, 2021. {A first-in-human report from the device developers describing early use in the first participants, with no control group and no blinded assessment.}

[^mitchell2023]: `paper` Mitchell, P. et al. "Assessment of safety of a fully implanted endovascular brain-computer interface for severe paralysis in 4 patients." *JAMA Neurology*, 2023. {Four participants, safety endpoints only. Vessel patency at twelve months is the load-bearing finding; the study was not designed to show clinical benefit.}

[^willett2023]: `paper` Willett, F. R. et al. "A high-performance speech neuroprosthesis." *Nature*, 2023.

==============================================================================
ARTICLE: steve-horvath
TITLE:   Steve Horvath
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/steve-horvath
SOURCE:  https://futurehumanwiki.com/raw/steve-horvath
==============================================================================

---
title: "Steve Horvath"
slug: "steve-horvath"
type: "person"
status: "established"
horizon: "present"
categories: ["people", "longevity"]
tags: ["epigenetics", "biomarkers", "aging", "dna methylation", "biostatistics", "measurement"]
summary: "German-American biostatistician who built the first multi-tissue epigenetic clock, giving aging research a quantitative readout it could test and argue about."
updated: "2026-07-27"
issues: ["Needs a sourced birth year and education detail; the infobox omits Born and Education rows."]
---

```infobox
{
  "caption": "Biostatistician and geneticist",
  "rows": [
    { "label": "Nationality", "value": "German-American" },
    { "label": "Training", "value": "Mathematics, then biostatistics" },
    { "label": "Known for", "value": "Multi-tissue epigenetic clock, 2013", "link": "/wiki/epigenetic-clock" },
    { "label": "Also known for", "value": "WGCNA network methods" },
    { "label": "Field", "value": "Biostatistics; epigenetics of aging" },
    { "label": "Long affiliation", "value": "UCLA" },
    { "label": "Current affiliation", "value": "Altos Labs", "link": "/wiki/altos-labs" }
  ]
}
```

**Steve Horvath** is a German-American biostatistician who in 2013 published a predictor that
estimates a person's chronological age from DNA methylation at a few hundred positions in the
genome, and does so across most human tissues. The model, generally known as the Horvath clock,
turned an intuition that bodies age at different rates into a number that could be measured,
compared between cohorts, and disputed. Most subsequent [[epigenetic-clock|methylation clocks]] are
either refinements of it or reactions against its design.

## Overview

Horvath is a statistician by trade rather than a bench biologist, and the clock reflects that
origin. It was not derived from a theory of what ages cells. It was fitted — a penalised regression
run over methylation arrays until a small subset of sites reproduced the calendar. That method is
the source of both the clock's reach and its central weakness, and Horvath has generally been more
explicit about the second than his field's publicity has been.

Before the clock he was known for network methods in genomics. Weighted gene co-expression network
analysis, developed with colleagues at UCLA in the mid-2000s, groups genes into correlated modules
and remains one of the most widely used tools in transcriptomics.[^zhang2005] The habit of looking
for structure in high-dimensional measurement, rather than for a mechanism, carried directly into
the aging work.

## Career

Horvath trained first in mathematics in Germany and then in biostatistics, and spent most of his
academic career at UCLA, holding appointments in human genetics and in biostatistics. In 2022 he
moved to [[altos-labs]], the reprogramming company that had launched publicly that January, where he
works on measurement for rejuvenation programmes — a natural destination, since a company attempting
to reset cellular age needs some way to say whether it has.

```timeline
[
  { "year": "2005", "title": "Co-expression network methods", "text": "Bin Zhang and Horvath publish a general framework for weighted gene co-expression network analysis, later packaged as WGCNA." },
  { "year": "2013", "title": "The multi-tissue clock", "text": "A single-author paper in Genome Biology describes a 353-site methylation predictor of chronological age that works across most human tissues." },
  { "year": "2015", "title": "Acceleration linked to mortality", "text": "Cohort analyses report that blood reading older than the calendar predicts all-cause mortality, independent of standard risk factors." },
  { "year": "2018", "title": "Second-generation clocks", "text": "DNAm PhenoAge, trained on clinical laboratory measures rather than on age alone, is published by Morgan Levine with Horvath and colleagues." },
  { "year": "2019", "title": "GrimAge", "text": "A clock built from methylation surrogates for plasma proteins and smoking history, regressed on time to death." },
  { "year": "2022", "title": "Move to industry", "text": "Horvath joins Altos Labs, taking the measurement problem into a company built around epigenetic rejuvenation." },
  { "year": "2023", "title": "Universal mammalian clock", "text": "The Mammalian Methylation Consortium reports clocks built on conserved CpG sites that estimate age across dozens of species." }
]
```

## The 2013 clock

The paper that made his name is sole-authored and methodologically plain.[^horvath2013] It assembled
thousands of publicly available methylation samples covering dozens of tissue and cell types, fitted
an elastic-net regression against chronological age, and reported that 353 CpG sites sufficed to
predict age with a median error of roughly three and a half years.

The number was not the point. Single-tissue predictors already existed, and a blood clock is a
reasonable thing to expect: leukocyte populations shift with age in ways an assay can see. What
Horvath showed was that a single set of coefficients, fitted once, worked on brain, breast, kidney,
liver, saliva and buccal cells alike. That implied the methylation changes it tracks are not a
property of one cell lineage's turnover history but something closer to a shared clock running in
most of the body.

One side result in the same paper did as much work as the main one: the predictor returned close to
zero for [[induced-pluripotent-stem-cells]], meaning that reprogramming a cell resets the reading
along with everything else. A second oddity came out of his group's later tissue comparisons. The
cerebellum reads consistently younger than the rest of the body, and no settled explanation for that
has emerged.

> [!key] Why a number changed the field
> Before 2013, arguments about whether an intervention slowed aging in people had no endpoint short
> of waiting for deaths. A cheap molecular readout gave the field something it could put in a trial
> protocol. Whether it deserves that role is a separate question from whether it transformed the
> conversation, and it did.

## Age acceleration

Horvath's second contribution is conceptual and follows from the first. If a model predicts
chronological age well, the interesting quantity is where it fails: the residual between predicted
and actual age, which he called epigenetic age acceleration. He and his collaborators split it into
an intrinsic component, adjusted for blood cell composition, and an extrinsic component that
retains it.

Cohort studies then tested whether the residual carried information. It does, modestly. Blood
reading older than the birth certificate predicts all-cause mortality across multiple cohorts after
adjustment for conventional risk factors.[^marioni2015] Acceleration has since been reported in
association with obesity, chronic infection, socioeconomic position and several genetic conditions,
findings that gave the measure clinical face validity well before anyone could say what it was
measuring. It sits alongside older candidates such as [[telomeres-and-telomerase|telomere length]]
in the search for usable [[aging-biomarkers]], and outperforms them on precision by a wide margin.

> [!note] What "the Horvath clock" refers to
> The phrase is used for the 2013 multi-tissue model and, loosely, for any clock from his group.
> They differ in kind. The 2013 clock and the 2018 skin-and-blood clock predict chronological age;
> PhenoAge and GrimAge predict health and death. Papers reporting that "the epigenetic clock"
> responded to a treatment often do not say which one, and the answer usually matters.

## The correlate problem

The limitation is structural, not a matter of better data. A model fitted to reproduce chronological
age is by construction a correlate of chronological age. Nothing in the fitting procedure can
distinguish a site that drives aging from one that merely records it, and no epigenetic clock has
been shown to track a causal driver rather than a downstream consequence. Horvath and Kenneth Raj
have said as much in review, describing the clock as plausibly reading out an epigenetic maintenance
system whose identity is unknown.[^horvath2018]

Second-generation clocks are the field's answer to that criticism, and Horvath and his collaborators
built the main ones. DNAm PhenoAge, led by Morgan Levine, was trained on a composite of clinical
laboratory measures rather than on age alone.[^levine2018] GrimAge went further still, constructing
methylation surrogates for plasma proteins and for smoking pack-years and regressing those on time
to death.[^lu2019] Both predict mortality substantially better than the 2013 clock. Both are also
further from being mechanistic: GrimAge carries a large share of its weight in a smoking surrogate,
which is a strong predictor of dying and a weak claim about aging.

> [!caution] What a moved clock does not establish
> Showing that a treatment lowers a methylation age reading does not show it slowed aging. The
> reading could shift without any change in function, and running the same sample twice can move the
> estimate by more than a year on some platforms, which is large relative to reported intervention
> effects.[^higgins2022] Horvath was senior author on the TRIIM study, in which nine men receiving
> growth hormone alongside two other drugs showed reduced clock readings; the study had no control
> group, and it remains among the most cited human results in the area.[^fahy2019]

## Beyond humans

The Mammalian Methylation Consortium, which Horvath helped organise, profiled tissues from dozens of
mammalian species on a shared array targeting CpG sites conserved across mammals, and built clocks
that estimate age in a mouse, a bat and a bowhead whale from one set of coefficients.[^lu2023] The
comparative programme is the closest thing the clock literature has to a test of whether methylation
change is fundamental to mammalian aging or an artefact of how any one species' cells divide. It
also supplies a practical instrument for animals whose lifespans can actually be measured, a use
described under [[epigenetic-clock]]; whether the same relationships hold in animals of
[[negligible-senescence|negligible senescence]] remains an open comparative question.

## Reception

The 2013 paper is among the most cited in modern aging research, and the clock is now standard
apparatus in cohort epidemiology, in [[senolytics]] and [[partial-reprogramming|reprogramming]]
studies, and in the endpoint discussions that surround the [[geroscience-hypothesis]]. As of 2026 no
regulator has accepted any clock as a surrogate endpoint, and the [[xprize-healthspan]] competition
deliberately chose functional measures instead.

Horvath's own public posture has been unusually restrained for the field. He has repeatedly framed
the clock as a measurement in search of a mechanism rather than as a diagnosis, which distinguishes
him from the consumer market that grew up around his work — direct-to-consumer methylation age
tests, discussed under [[epigenetic-clock]] and [[consumer-blood-testing]], routinely report a
single number with a confidence the underlying assay reliability does not support.

The open question his work sets up is whether the methylation changes he found are part of the
machinery of aging or its exhaust. Both readings are consistent with everything measured so far, and
they imply completely different research programmes: one in which [[epigenetic-reprogramming]] is a
therapy and one in which it is cosmetics for a biomarker.

## See also

- [[epigenetic-clock]]
- [[biological-age]]
- [[aging-biomarkers]]
- [[epigenetic-reprogramming]]
- [[hallmarks-of-aging]]
- [[david-sinclair]]
- [[altos-labs]]
- [[geroscience-hypothesis]]

## References

[^zhang2005]: `paper` Zhang, B. and Horvath, S. "A General Framework for Weighted Gene Co-Expression Network Analysis." *Statistical Applications in Genetics and Molecular Biology*, 2005.
[^horvath2013]: `paper` Horvath, S. "DNA methylation age of human tissues and cell types." *Genome Biology*, 2013. {Sole-authored, and built entirely from previously published datasets rather than new sample collection.}
[^marioni2015]: `paper` Marioni, R.E. et al. "DNA methylation age of blood predicts all-cause mortality in later life." *Genome Biology*, 2015.
[^levine2018]: `paper` Levine, M.E. et al. "An epigenetic biomarker of aging for lifespan and healthspan." *Aging*, 2018.
[^lu2019]: `paper` Lu, A.T. et al. "DNA methylation GrimAge strongly predicts lifespan and healthspan." *Aging*, 2019.
[^horvath2018]: `paper` Horvath, S. and Raj, K. "DNA methylation-based biomarkers and the epigenetic clock theory of ageing." *Nature Reviews Genetics*, 2018.
[^higgins2022]: `paper` Higgins-Chen, A.T. et al. "A computational solution for bolstering reliability of epigenetic clocks." *Nature Aging*, 2022.
[^fahy2019]: `paper` Fahy, G.M. et al. "Reversal of epigenetic aging and immunosenescent trends in humans." *Aging Cell*, 2019. {Nine men, no control arm; the clock analysis was the study's headline result rather than its prespecified aim.}
[^lu2023]: `paper` Lu, A.T. et al. "Universal DNA methylation age across mammalian tissues." *Nature Aging*, 2023.

==============================================================================
ARTICLE: substrate-independence
TITLE:   Substrate independence
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/substrate-independence
SOURCE:  https://futurehumanwiki.com/raw/substrate-independence
==============================================================================

---
title: "Substrate independence"
slug: "substrate-independence"
type: "concept"
status: "contested"
horizon: "indefinite"
categories: ["minds", "foundations"]
tags: ["functionalism", "consciousness", "philosophy of mind", "uploading", "multiple realizability", "qualia"]
summary: "The claim that mental states depend on a system's functional organization rather than on the material implementing it, so that a mind could run on non-biological hardware."
updated: "2026-07-27"
issues: ["Aaronson's expander-graph objection and the 2023 open letter on IIT are described without citations."]
---

```infobox
{
  "caption": "Thesis in philosophy of mind",
  "rows": [
    { "label": "Also called", "value": "Computational functionalism" },
    { "label": "Ancestor claim", "value": "Multiple realizability, 1967" },
    { "label": "Key defence", "value": "Fading and dancing qualia" },
    { "label": "Main rivals", "value": "Biological naturalism, IIT" },
    { "label": "Empirical status", "value": "No decisive test proposed" },
    { "label": "Load-bearing for", "value": "Mind uploading, machine consciousness" }
  ]
}
```

**Substrate independence** is the thesis that what makes a physical system have mental states is the
pattern of causal relations among its parts, not the stuff those parts are made of. If it is true,
a system built from silicon, or from any other medium that reproduces the relevant organization,
would have the same mental states as a brain reproducing the same organization — including, on the
strong reading, the same conscious experience. The thesis is the philosophical load-bearing wall
under [[mind-uploading]], [[whole-brain-emulation]], and most arguments that a machine could be
conscious. It is not established, and the strong reading is actively disputed.

## The argument

The ancestor of the thesis is *multiple realizability*, proposed by Hilary Putnam in 1967: pain
cannot be identical to a specific brain-state type, because octopuses, humans, and hypothetical
silicon creatures could all be in pain while sharing no neural type.[^putnam1967] What they share is
a functional role — a state caused by tissue damage, causing avoidance and complaint. If mental
kinds are functional kinds, then any system that fills the role has the mental state, whatever it is
made of.

Applied to a whole mind, the claim becomes that reproducing the causal organization of a brain at
sufficient grain reproduces the mind. Two clarifications matter. First, "sufficient grain" is
doing enormous work; nobody claims that reproducing the organization of a brain at the level of
brain regions would suffice. The thesis is compatible with the required grain being molecular, which
would make it useless in practice. Second, substrate independence about *behaviour* is far less
contentious than substrate independence about *experience*. Almost nobody denies that a sufficiently
detailed simulation would behave like a person; the argument is about whether the lights are on.

[[nick-bostrom]] states substrate independence explicitly as a premise of the simulation
argument.[^bostrom2003] It appears, usually unstated, wherever emulation is discussed as a route to
survival, and it is assumed by most versions of [[transhumanism]] that treat biology as replaceable
hardware.

## Fading and dancing qualia

David Chalmers offers the most-cited defence.[^chalmers1995] Imagine replacing neurons one at a time
with silicon devices that reproduce each neuron's input-output behaviour exactly. Behaviour is
preserved by construction, so at every stage the subject reports normal experience. If experience
nonetheless drains away as the substrate changes, there must be a point at which the subject has
faint or absent experience while sincerely reporting rich experience — a systematic dissociation
between consciousness and judgement that Chalmers argues is implausible. A variant, *dancing qualia*,
imagines switching between biological and silicon circuits during operation: if experience changed
with each switch, the subject would be unable to notice enormous swings in their own perception.

The argument does not prove substrate independence; it argues that the alternatives are strange. It
also assumes precisely what is at issue in the engineering — that a device can reproduce a neuron's
input-output behaviour exactly, in context, including the chemical signalling that
[[connectomics]] cannot see.

> [!key] What follows if it is true
> Substrate independence does not by itself make anyone survive an upload. It says the copy would
> have a mind; whether the copy is the original person is a separate question handled under
> [[personal-identity-and-continuity]] and sharpened by the [[teleportation-problem]]. Conflating
> the two is the most common error in popular treatments.

## Objection: biological naturalism

John Searle argues that computation is not sufficient for mind, because computation is defined
syntactically and mental states have content.[^searle1980] The Chinese Room presents a system that
manipulates symbols correctly without understanding them; Searle's conclusion is that the brain
causes consciousness through specific biological powers, in the way that the stomach digests, and
that a simulation of those powers no more produces consciousness than a simulation of digestion
produces a meal. The argument has been attacked from every direction for four decades — most
prominently by the reply that the system as a whole understands even if the person inside it does
not — and it retains adherents.

A subtler version comes from Peter Godfrey-Smith, who argues that the metabolic and dynamical
properties of living tissue may not be incidental packaging for a computation but part of what makes
minds possible.[^gs2016] Brains are not clocked digital systems; they are chemical processes with no
clean separation between signal and machinery. On this view the question is not whether silicon is
special but whether the abstraction that treats a neuron as an input-output device discards
something essential.

Ned Block's earlier objection makes the intuition vivid: if functional organization is all that
matters, then the population of China, connected by radios so as to implement the organization of a
brain for an hour, would collectively have experiences.[^block1978] Functionalists generally accept
this conclusion; critics take it as a reductio.

## Objection: integrated information theory

Integrated information theory, developed by Giulio Tononi, is a rare theory of consciousness that
denies substrate independence outright. It holds that consciousness is identical to a system's
intrinsic cause-effect power, quantified as Φ, and that Φ depends on the physical architecture
rather than the input-output function. A digital computer executing a simulation of a brain, running
transistors in a largely feed-forward, sequentially addressed architecture, would on this account
have very low Φ regardless of how convincingly it talks. Christof Koch's formulation is that
simulating a black hole does not warp spacetime around the computer.

This is a substantive empirical-metaphysical disagreement rather than a quibble: IIT and
computational functionalism make opposite predictions about whether an emulation is conscious, and
no experiment currently distinguishes them, since both predict identical behaviour. IIT is itself
contested — Scott Aaronson has argued that simple mathematical structures such as expander graphs
achieve enormous Φ while plainly not being conscious, and in 2023 more than a hundred researchers
signed an open letter arguing that the theory's distinctive claims have not been empirically tested
and objecting to its presentation as an established account. The relevant point
here is that the leading theory-driven challenge to substrate independence comes from inside
consciousness science, not from vitalism.[^tononi2016] It also cuts the other way for biological
systems: on IIT's account, a cortical [[organoids|organoid]] with rich recurrent connectivity could
have more integrated information than a language model with far more parameters.

## Objection: computation is not intrinsic

A third line questions whether "implementing a computation" is a fact about a physical system at all.
Putnam later argued that any open physical system can be interpreted as implementing any
finite-state automaton, and Searle made a similar point: computation is observer-relative, so a
system does not have mental states in virtue of a computation unless something fixes which
computation it performs. Chalmers's reply is that implementation requires a *causal* correspondence
between physical state transitions and computational ones, which is a fact about the system's
counterfactual structure and not an interpretation.[^chalmers1996] Whether that reply succeeds is
still argued.

## Where the argument stands

There is no experiment on offer. The thesis makes no prediction that differs, behaviourally, from
its denial — which is why the debate is conducted with thought experiments rather than data, and why
progress in [[neural-correlates-of-consciousness]] research has not resolved it. The most likely
route to indirect evidence runs through theory: if a theory of consciousness earns enough empirical
support to be trusted, its verdict on non-biological systems inherits that support. Approaches to
[[machine-consciousness]] that assess systems against indicator properties drawn from multiple
theories are an attempt to make progress under that uncertainty.

Meanwhile the thesis functions as a premise in arguments it cannot support on its own. Claims that
[[cryonics]] or [[brain-preservation]] preserve a person, that an emulation would deserve moral
consideration, that progress toward [[artificial-general-intelligence]] bears on whether machines
have experiences, or that a [[posthuman]] future includes non-biological people all assume it.
Even the gradualist framings collected under [[human-ai-merger]] assume it at the limit, since a
sufficiently replaced person is a non-biological one. Treating the thesis as settled, in either
direction, is the error.

## See also

- [[mind-uploading]]
- [[machine-consciousness]]
- [[neural-correlates-of-consciousness]]
- [[whole-brain-emulation]]
- [[personal-identity-and-continuity]]
- [[teleportation-problem]]
- [[digital-immortality]]
- [[posthuman]]

## References

[^putnam1967]: `book` Putnam, H. "Psychological Predicates." In *Art, Mind, and Religion*, University of Pittsburgh Press, 1967. Later reprinted as "The Nature of Mental States." {Putnam later abandoned functionalism and argued against the position this paper introduced.}

[^chalmers1995]: `book` Chalmers, D. J. "Absent Qualia, Fading Qualia, Dancing Qualia." In T. Metzinger (ed.), *Conscious Experience*, Imprint Academic, 1995.

[^searle1980]: `paper` Searle, J. R. "Minds, brains, and programs." *Behavioral and Brain Sciences*, 1980. {Searle states and answers the standard objections, the systems reply among them, inside the paper itself; the journal's format then printed peer commentary alongside it.}

[^block1978]: `paper` Block, N. "Troubles with Functionalism." *Minnesota Studies in the Philosophy of Science*, 1978.

[^gs2016]: `paper` Godfrey-Smith, P. "Mind, Matter, and Metabolism." *The Journal of Philosophy*, 2016.

[^chalmers1996]: `paper` Chalmers, D. J. "Does a Rock Implement Every Finite-State Automaton?" *Synthese*, 1996.

[^bostrom2003]: `paper` Bostrom, N. "Are You Living in a Computer Simulation?" *Philosophical Quarterly*, 2003.

[^tononi2016]: `paper` Koch, C., Massimini, M., Boly, M. and Tononi, G. "Neural correlates of consciousness: progress and problems." *Nature Reviews Neuroscience*, 2016. {A review of the neural correlates literature by proponents of the theory, not a primary statement of integrated information theory or of the Phi formalism.}

==============================================================================
ARTICLE: synchron
TITLE:   Synchron
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/synchron
SOURCE:  https://futurehumanwiki.com/raw/synchron
==============================================================================

---
title: "Synchron"
slug: "synchron"
type: "organization"
status: "emerging"
horizon: "late 2020s"
categories: ["organizations", "cybernetics"]
tags: ["brain-computer interface", "endovascular", "neurotech", "paralysis", "clinical trials", "implants"]
summary: "A neurotechnology company developing an endovascular brain–computer interface implanted through a blood vessel rather than by opening the skull."
updated: "2026-07-27"
issues: ["The funding section names investors but gives no amounts and carries no citation.", "The 2025 Apple accessibility and NVIDIA collaborations are stated without sources."]
---

```infobox
{
  "caption": "Neurotechnology company",
  "rows": [
    { "label": "Founded", "value": "2012, Melbourne" },
    { "label": "Founders", "value": "Thomas Oxley, Nicholas Opie" },
    { "label": "Headquarters", "value": "New York, United States" },
    { "label": "Device", "value": "Stentrode", "link": "/wiki/stentrode" },
    { "label": "Delivery route", "value": "Jugular vein to sagittal sinus" },
    { "label": "Electrodes", "value": "16" },
    { "label": "Key trials", "value": "SWITCH (Australia), COMMAND (US)" }
  ]
}
```

**Synchron** is a neurotechnology company developing the [[stentrode]], a [[brain-computer-interface]] built into a vascular stent and delivered to a vein on the surface of the motor cortex through the jugular, avoiding craniotomy entirely. Founded in Melbourne in 2012 and now headquartered in New York, it began implanting patients in Australia in 2019 and in the United States in 2022, and it has pursued a deliberately low-bandwidth strategy in contrast to competitors chasing channel count.

## Overview

The company's strategic premise is that the binding constraint on BCI deployment is not the number of neurons recorded but the willingness of surgeons, regulators and patients to accept an implant. Open-skull electrode arrays require neurosurgery, carry infection and haemorrhage risk, and are performed at a small number of centres. Endovascular delivery is a procedure that interventional neuroradiologists already perform thousands of times a year for stroke and aneurysm, using equipment that already exists in most large hospitals. The argument is explicitly modelled on the diffusion of the [[cochlear-implant]] and of [[deep-brain-stimulation]], both of which reached large patient populations only once the implantation procedure became routine.

The cost of that choice is signal. The Stentrode carries sixteen electrodes sitting inside a vessel wall, separated from cortex by the vessel and by tissue, and records local field potentials rather than single-unit activity. It cannot approach the information rate of a penetrating array. Synchron's position is that a small, reliable set of control signals is sufficient for the applications that matter first — switch control, cursor movement, and communication for people with severe paralysis.

```keyfacts
[
  { "value": "16", "label": "Electrodes on the Stentrode", "note": "against 96 on a Utah array and over 1,000 on some research devices" },
  { "value": "2019", "label": "First human implants", "note": "the SWITCH trial in Melbourne" },
  { "value": "2022", "label": "First US implant", "note": "under an FDA investigational device exemption granted in 2021" }
]
```

## History

Thomas Oxley, a neurologist, and Nicholas Opie, a biomedical engineer, developed the endovascular approach at the University of Melbourne with support from Australian and United States defence research funding. The founding insight was that veins run close to the cortical surface, that stents are routinely delivered to them, and that a stent could carry electrodes and become endothelialized into the vessel wall over weeks, stabilizing the recording. The group's 2016 report of chronic cortical recordings from a stent-mounted array in freely moving sheep established that the approach could work at all.[^oxley2016]

The SWITCH trial implanted the device in four Australian patients with severe paralysis, mostly from amyotrophic lateral sclerosis. The first-in-human report described patients using the system for digital device control at home,[^oxley2021] and a subsequent safety analysis found no device-related serious adverse events and no vessel occlusion at twelve months.[^mitchell2023] Synchron then received a United States investigational device exemption and began the COMMAND early feasibility study, implanting its first American patient in 2022 at Mount Sinai; COMMAND enrolled a small cohort across several centres and has been reported as meeting its primary safety endpoint.

```timeline
[
  { "year": "2012", "title": "Founded", "text": "Thomas Oxley and Nicholas Opie establish the company in Melbourne around an endovascular electrode concept developed with defence research funding." },
  { "year": "2016", "title": "Animal validation", "text": "Recordings from a stent-mounted array in sheep demonstrate that a vessel-borne electrode can pick up cortical signals over months." },
  { "year": "2019–2021", "title": "SWITCH trial", "text": "Four Australian patients with severe paralysis receive the device; the trial reports vessel patency and usable control signals at twelve months." },
  { "year": "2021", "title": "US regulatory clearance to trial", "text": "The FDA grants an investigational device exemption, permitting the first US implants." },
  { "year": "2022–2024", "title": "COMMAND", "text": "A small US early feasibility study across several centres reports acceptable safety and functional digital control." },
  { "year": "2025", "title": "Platform integrations", "text": "Apple's accessibility framework adds native support for BCI input with Synchron as a partner, and the company announces work with NVIDIA on decoding models." }
]
```

## Technology

The device is a self-expanding nitinol stent with electrodes mounted on its struts. It is delivered by catheter through the jugular vein into the superior sagittal sinus, the large vein running along the midline above the motor cortex, and deployed so that the electrodes press against the vessel wall adjacent to motor cortex, whose leg representation lies nearest the midline. Over several weeks the stent endothelializes, becoming incorporated into the vessel wall, which both stabilizes it mechanically and, Synchron argues, reduces the chronic foreign-body response that degrades [[utah-array|penetrating array]] signals over years.

A lead runs down the vein to a subclavian transmitter unit implanted in the chest, which powers the array and relays data wirelessly. The architecture is closer to a pacemaker than to a research rig, a design choice shared with most deployed [[neuroprosthetics]]. The system decodes attempted movements — typically of a foot or ankle — into discrete switch events, which are mapped onto interface actions. No stimulation is delivered, so the device provides no sensory feedback and is not a bidirectional interface.

The signal ceiling follows from the recording site: local field potentials rather than the spike trains that [[neural-decoding]] work on penetrating arrays exploits, which limits the system to a small number of distinguishable commands and makes the interaction paradigm resemble switch scanning rather than continuous multi-degree-of-freedom control. [[stentrode]] sets out the mechanism in full.

> [!key] The tradeoff stated plainly
> A Utah array records individual neurons and supports high-dimensional control, at the cost of a craniotomy and progressive signal loss from glial scarring. The Stentrode records population activity through a vessel wall and supports a handful of commands, at the cost of no craniotomy and a stable interface. Which is the better device depends entirely on how many patients would accept brain surgery, a question the field has not answered empirically.

## Partnerships and funding

Synchron has raised venture funding from investors including ARCH Venture Partners, with participation reported from vehicles associated with Jeff Bezos and Bill Gates. It has pursued integration partnerships rather than building its own application layer: a demonstration of Amazon Alexa control, and, in 2025, participation in Apple's addition of a native brain–computer interface input protocol to its accessibility framework — the first time a major consumer platform treated neural input as a first-class modality alongside touch and switch control. A separate collaboration with NVIDIA concerns machine-learning models for decoding, using the accumulated recordings from its trial participants.

The platform strategy matters more than it appears. A BCI that presents itself to an operating system as a standard input device inherits every application already written for accessibility, which removes a substantial part of the software problem from the company's own scope.

## Reception

Clinical reception has been favourable on safety, which is the metric an early feasibility study is designed to address. Independent commentators have noted that endovascular delivery genuinely lowers the barrier to implantation and that the reported vessel patency results are the most important finding, since thrombosis in a major cerebral vein would be a catastrophic failure mode.

Criticism concentrates on capability. Sixteen channels of field potential cannot support the [[speech-neuroprosthesis|speech decoding]] results that [[ecog-interfaces|electrocorticographic]] and penetrating arrays have produced, and cannot approach the cursor performance demonstrated by BrainGate participants. Some researchers argue that the company's approach solves the wrong constraint: that patients with locked-in syndrome will accept a craniotomy, and that the value of a BCI scales steeply with bandwidth. Synchron's counter is that the field has produced two decades of impressive results in a few dozen research participants and almost no deployed devices, and that reaching many patients requires a procedure that ordinary hospitals can perform.

The comparison with [[neuralink]] is unavoidable and largely unhelpful, since the two are pursuing different points on the same tradeoff curve. Neuralink's robot-inserted thread arrays record far more channels and require skull penetration; Synchron's device records far fewer and does not. Both remain in early clinical studies, and neither has an approved product.

## Outlook

The decisive question is whether a pivotal trial can demonstrate benefit against an endpoint a regulator accepts. Early feasibility studies establish safety in a handful of patients; approval requires a defined patient population, a validated outcome measure, and a comparator, none of which the BCI field has settled. Synchron has said it intends to run a larger trial, and the design of that trial will reveal what claim the company believes it can support.

The second question is durability. Endothelialization stabilizes the device but also fixes it permanently in a vessel; explantation has not been demonstrated in humans and the long-term consequences of a foreign body in a major cerebral sinus over decades are unknown.

The third is governance. The [[neurorights]] campaign and the [[mental-privacy]] literature both concern who owns neural recordings and what may be inferred from them, and a device that streams cortical signals to a consumer operating system sits directly in the path of that argument. Field potentials from motor cortex are crude by the standards of what laboratory decoding work can extract, which is a weak reassurance rather than a strong one: the history of the field is that inference from neural data consistently outruns what the raw signal appears to contain. Whether the bandwidth ceiling that limits the device also limits the [[human-ai-merger]] ambitions that motivate much BCI investment is a question Synchron's own strategy answers in the affirmative, and its competitors dispute.

## See also

- [[stentrode]]
- [[brain-computer-interface]]
- [[neuralink]]
- [[utah-array]]
- [[ecog-interfaces]]
- [[neural-decoding]]
- [[speech-neuroprosthesis]]
- [[neurorights]]

## References

[^oxley2016]: `paper` Oxley, T.J. et al. "Minimally invasive endovascular stent-electrode array for high-fidelity, chronic recordings of cortical neural activity." *Nature Biotechnology*, 2016.
[^oxley2021]: `paper` Oxley, T.J. et al. "Motor neuroprosthesis implanted with neurointerventional surgery improves capacity for activities of daily living tasks in severe paralysis: first in-human experience." *Journal of NeuroInterventional Surgery*, 2021. {Authored by the company's founders about the trial they designed; a first-in-human report records early use and does not test efficacy.}
[^mitchell2023]: `paper` Mitchell, P. et al. "Assessment of Safety of a Fully Implanted Endovascular Brain-Computer Interface for Severe Paralysis in 4 Patients: The Stentrode With Thought-Controlled Digital Switch (SWITCH) Study." *JAMA Neurology*, 2023. {A four-patient safety series with no comparison group; twelve-month vessel patency is what it establishes, not clinical benefit.}

==============================================================================
ARTICLE: synthetic-genomes
TITLE:   Synthetic genomes
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/synthetic-genomes
SOURCE:  https://futurehumanwiki.com/raw/synthetic-genomes
==============================================================================

---
title: "Synthetic genomes"
slug: "synthetic-genomes"
type: "technology"
status: "experimental"
horizon: "2040s"
trl: 4
categories: ["genetics"]
tags: ["synthetic biology", "dna synthesis", "genome engineering", "minimal cell", "biosecurity"]
summary: "Genomes designed on a computer and assembled from chemically synthesized DNA, then installed in a cell to see whether the design lives."
updated: "2026-07-27"
humanEvidence: "Nothing has been done in a person. The built genomes are bacterial and yeast, human-directed work is confined to cultured cells, and no synthetic human chromosome exists as of 2026."
access: "Nothing to obtain as a product: genome-scale writing is done by a small number of funded laboratories, while commercial suppliers sell only short DNA fragments, subject to order screening."
reversibility: "irreversible"
issues: ["The 2025 UK synthetic human chromosome programme is described without a citation.", "The recoded Escherichia coli results carry no citation."]
---

```infobox
{
  "caption": "Field of synthetic biology",
  "rows": [
    { "label": "First synthetic bacterial genome", "value": "2010, JCVI-syn1.0" },
    { "label": "Minimal cell", "value": "2016, 473 genes" },
    { "label": "First synthetic eukaryote project", "value": "Sc2.0, yeast" },
    { "label": "Largest recoded genome", "value": "4 Mb, Escherichia coli" },
    { "label": "Human-scale effort", "value": "SynHG, announced 2025" },
    { "label": "Binding constraint", "value": "Assembly, not synthesis" },
    { "label": "Readiness", "value": "TRL 4" }
  ]
}
```

**Synthetic genomes** are chromosomes designed as text and then built as molecules: an organism's genetic instructions written in a design file, assembled from short synthetic DNA fragments, and installed in a cell to see whether it functions. The point is not to copy nature but to test it. A designed genome states exactly which genes an experimenter believes are necessary and which changes are tolerable, and the cell either grows or it does not. That makes genome writing a comprehension test as much as an engineering achievement — the claim of the field's founders being that what cannot be built is not understood.

```figure
{"key": "dna-double-helix", "caption": "Writing a genome means assembling this molecule to a specification, rather than copying one that already exists."}
```

## How genomes are built

Chemical DNA synthesis produces reliable strands only up to a few hundred bases, so every genome-scale project is an exercise in hierarchical assembly. Short oligonucleotides are joined into fragments of a few kilobases, those into chunks of tens of kilobases, and chunks into megabase assemblies, usually by exploiting yeast's efficient homologous recombination to stitch overlapping pieces together inside a living cell. Errors accumulate at every stage and must be sequenced out.

The final step is the hard one. A completed synthetic chromosome must displace the native one. In bacteria this has been done by transplanting a whole genome into a recipient cell emptied of its own, the approach that produced the first cell running on a synthesized chromosome in 2010;[^gibson2010] in yeast, by replacing native chromosome segments piece by piece with synthetic equivalents, selecting at each round, so that the cell is converted gradually rather than rebooted. There is no general method for either operation in animal cells, which is the main reason mammalian genome writing remains at the stage of individual chromosome fragments.

> [!note] Terminology
> "Synthetic genome" describes provenance, not novelty. JCVI-syn1.0 was a near-copy of a natural sequence assembled from synthesized DNA; a heavily redesigned genome that is partly natural in origin may be far more synthetic in the ordinary sense of the word.

## Development history

The field has advanced by roughly one order of magnitude of genome size per decade, and each step has been limited by assembly and installation rather than by the chemistry of making DNA. The sequence below is also a sequence of increasing ambition: from copying a viral genome, to copying a bacterial one, to redesigning a eukaryotic one.

```timeline
[
  { "year": "2002", "title": "Poliovirus from published sequence", "text": "Eckard Wimmer's group assembles an infectious poliovirus genome from mail-order oligonucleotides, the first demonstration that a viral genome can be built from sequence data alone." },
  { "year": "2008", "title": "A bacterial genome synthesized", "text": "The J. Craig Venter Institute assembles the Mycoplasma genitalium genome, though it is not installed in a cell." },
  { "year": "2010", "title": "JCVI-syn1.0", "text": "A synthesized Mycoplasma mycoides genome is transplanted into a recipient cell, which then replicates under the control of the built chromosome." },
  { "year": "2014", "title": "First synthetic eukaryotic chromosome", "text": "The Sc2.0 consortium reports a redesigned, functional yeast chromosome III, stripped of transposons and introns." },
  { "year": "2016", "title": "Minimal cell", "text": "JCVI-syn3.0 is reduced to 473 genes, of which a substantial fraction have no known function." },
  { "year": "2019–2021", "title": "Recoded Escherichia coli", "text": "A 4-megabase E. coli genome is rebuilt with three codons removed, and the corresponding transfer RNAs and release factor deleted." },
  { "year": "2023", "title": "Sc2.0 chromosomes completed", "text": "All sixteen synthetic yeast chromosomes are reported, along with a designed neochromosome carrying relocated transfer RNA genes." },
  { "year": "2025", "title": "Human chromosome project announced", "text": "A UK-funded programme is announced to develop the tools for building a synthetic human chromosome in cultured cells, with an attached social science strand." }
]
```

## Minimal cells

The most informative result to come out of genome writing is negative. In building a stripped-down *Mycoplasma*, the Venter Institute found that a genome could be cut to 473 genes and still sustain a free-living cell — but that roughly a third of those genes had no assigned function.[^hutchison2016] They were necessary and unexplained. A decade of sequencing and annotation had not identified them, and only the requirement to build a working cell exposed the gap.

Later versions restored a small number of genes to fix abnormal cell division, producing a strain that grows and divides more normally and that now serves as a chassis for studying how a minimal set of parts organises a cell. The lesson generalises: gene-by-gene knockout studies systematically miss redundancy, and only a whole-genome design forces the question of sufficiency.

## Sc2.0 and the rewritable genome

The synthetic yeast project set out to redesign rather than copy, beginning with a rebuilt chromosome III reported in 2014.[^annaluru2014] Its chromosomes have transposable elements and introns removed, all instances of one stop codon swapped for another to free a codon for later reassignment, and transfer RNA genes relocated to a dedicated artificial chromosome. The most distinctive design choice was to build in a mechanism for scrambling: recombination sites placed after every non-essential gene allow an inducible enzyme to delete, invert and rearrange the genome at random, generating enormous structural diversity from a single strain.

This makes the synthetic yeast genome an evolutionary instrument rather than a fixed product. It can be shuffled deliberately to find configurations that improve a trait, an approach not available with the targeted edits of [[crispr-cas9]] or the sequence-level rewriting of [[prime-editing]].

## Recoding and virus resistance

Rebuilding a genome allows changes no editing tool could make efficiently, because they must be applied at thousands of positions at once. Removing every instance of a codon from a bacterial genome, then deleting the machinery that reads it, frees that codon for reassignment to a non-standard amino acid — and incidentally renders the cell unreadable to viruses, whose genomes still use the original code. These consequences are treated in detail in [[recoded-organisms]].

## Human genome writing

Genome Project–write, proposed in 2016 by a group including Jef Boeke and [[george-church]] as a successor to the Human Genome Project, aimed to reduce the cost of designing and building large genomes by orders of magnitude.[^boeke2016] Its announcement was clumsy: an invitation-only meeting closed to press prompted public criticism from synthetic biologists and bioethicists that a project of this kind should not begin behind a closed door, and the organisers subsequently reframed the goal around cells rather than organisms. The flagship proposal became an "ultra-safe" human cell line, recoded for resistance to viral infection, which would be valuable for biomanufacturing and would sidestep the question of synthetic people entirely. The same reasoning drives the multiply edited donor animals used in [[xenotransplantation]], where inactivating endogenous retroviruses is a prerequisite rather than a bonus.

In 2025 a British funder committed to a programme aimed at developing the tools to build a synthetic human chromosome, with an explicit remit confined to cells in culture and a parallel social science strand. As of 2026 no synthetic human chromosome exists. The gap is quantitative and large: a human genome is close to a thousand times the size of the largest genome yet assembled from synthetic DNA, mammalian cells have no equivalent of yeast's chromosome-replacement machinery, and chromatin structure, imprinting and long-range regulation are design variables nobody knows how to specify.

> [!debate] What the project would mean
> Supporters argue that synthetic human chromosomes would give a clean testbed for regulatory genomics and virus-resistant manufacturing cells, and that the work stops at cell culture. Critics respond that the tools do not stop where the stated goal does, and that a capability to write human chromosomes eventually meets the questions raised by [[germline-editing]] and by [[synthetic-embryos]] whether or not anyone intends it to. Neither side disputes that the capability is decades away.

## Risks and governance

Genome writing is the clearest current case of [[dual-use-research]], and the reason engineered pathogens dominate the biological section of the [[existential-risk]] literature. The 2018 reconstruction of horsepox virus from commercially ordered DNA, undertaken partly to demonstrate feasibility, showed that a pathogen of a size that matters could be built for a sum well within a small laboratory's budget.[^noyce2018] The principal control is screening at the point of synthesis: providers check orders against sequences of concern and check the identity of customers. The International Gene Synthesis Consortium operates such screening voluntarily, and United States federal policy has moved toward steering research funding to providers that screen, though the requirements have been revised more than once and the arrangement is not statutory. Coverage is incomplete, benchtop synthesis devices complicate enforcement, and sequence variants generated by [[ai-protein-design|protein design models]] that evade string-matching screens are an active concern.

The biosafety questions differ from those raised by self-spreading constructs such as [[gene-drive|gene drives]] or by the chirally inverted organisms discussed in [[mirror-life]]. A synthetic *Mycoplasma* is fastidious and fragile. The risk is in the capability and in what the falling cost of synthesis makes available to whom, which is why the governance debate here borrows less from [[governance-of-genome-editing]] than from arms control and from the norms established at the [[asilomar-conference]].

## Outlook

The near-term work is unglamorous: cheaper and more accurate synthesis, enzymatic rather than phosphoramidite chemistry, better assembly in mammalian cells, and design rules that predict whether a rewritten sequence will function before it is built. The deeper question is what genome writing is for. In bacteria and yeast the answers are concrete — chassis organisms, virus resistance, non-standard chemistry. For human cells the strongest case is a virus-resistant manufacturing line, which is a narrow industrial goal supporting a very broad capability, and the field has not settled whether that asymmetry is acceptable.

## See also

- [[recoded-organisms]]
- [[crispr-cas9]]
- [[dual-use-research]]
- [[mirror-life]]
- [[gene-drive]]
- [[prime-editing]]
- [[germline-editing]]
- [[george-church]]

## References

[^hutchison2016]: `paper` Hutchison, C. A. III et al. "Design and synthesis of a minimal bacterial genome." *Science*, 2016.
[^gibson2010]: `paper` Gibson, D. G. et al. "Creation of a bacterial cell controlled by a chemically synthesized genome." *Science*, 2010.
[^annaluru2014]: `paper` Annaluru, N. et al. "Total synthesis of a functional designer eukaryotic chromosome." *Science*, 2014.
[^boeke2016]: `paper` Boeke, J. D. et al. "The Genome Project–Write." *Science*, 2016. {A short proposal piece rather than an experimental report; it announces a programme and states an aim, and contains no results.}
[^noyce2018]: `paper` Noyce, R. S., Lederman, S., Evans, D. H. "Construction of an infectious horsepox virus vaccine from chemically synthesized DNA fragments." *PLoS ONE*, 2018. {The reconstruction was done for a company developing a smallpox vaccine, and the cost claim widely drawn from it is the authors' own estimate.}

==============================================================================
ARTICLE: tardigrade-genes
TITLE:   Tardigrade genes and human cells
PORTAL:  Space & Extreme Environments
URL:     https://futurehumanwiki.com/wiki/tardigrade-genes
SOURCE:  https://futurehumanwiki.com/raw/tardigrade-genes
==============================================================================

---
title: "Tardigrade genes and human cells"
slug: "tardigrade-genes"
type: "concept"
status: "experimental"
horizon: "2040s"
categories: ["space", "genetics"]
tags: ["tardigrade", "radiation", "dna repair", "anhydrobiosis", "gene transfer", "dsup"]
summary: "The attempt to transfer stress-tolerance proteins from tardigrades into human cells, and the gap between what those experiments showed and how they were reported."
updated: "2026-07-27"
humanEvidence: "No person has received Dsup by any route. Protection has been shown only in cultured human cell lines against acute X-rays; no organism-level radioprotection has been demonstrated in a mammal."
issues: ["The Dsup mRNA delivery work is described without a citation."]
---

```infobox
{
  "caption": "Cross-species gene transfer research",
  "rows": [
    { "label": "Source organism", "value": "Ramazzottius varieornatus" },
    { "label": "Key protein", "value": "Dsup (damage suppressor)" },
    { "label": "First human-cell result", "value": "2016" },
    { "label": "Effect shown", "value": "Reduced X-ray DNA damage" },
    { "label": "Tested in animals", "value": "Mice, preclinical only" },
    { "label": "Tested in humans", "value": "No" }
  ]
}
```

**Tardigrade genes and human cells** describes a small but heavily publicised research line in which proteins from tardigrades — millimetre-scale invertebrates that survive desiccation, freezing, vacuum, and enormous radiation doses — are expressed in mammalian cells to see whether their stress tolerance transfers. One protein, Dsup, does confer measurable protection against radiation-induced DNA damage in cultured human cells. That result is real and has been replicated. Almost everything the popular coverage attached to it is not.

## What tardigrades actually survive

Tardigrades enter cryptobiosis: a reversible, near-total suspension of metabolism triggered by drying, freezing, or oxygen deprivation. It is a far deeper shutdown than the mammalian torpor discussed under [[human-hibernation]], and unlike torpor it involves no maintained circulation at all. In the desiccated "tun" state, body water falls to a few per cent and detectable metabolism approaches zero. Animals have been revived from tuns after years, and dried specimens flown on the FOTON-M3 mission survived direct exposure to space vacuum, with survival falling sharply when solar ultraviolet was added.[^jonsson2008]

Their radiation tolerance is the headline property. Lethal doses for tardigrades are measured in thousands of grays, roughly a thousand times the dose that kills a human, and hydrated animals tolerate nearly as much as dried ones.

> [!key] Tardigrades are not radiation-adapted
> Nothing in a tardigrade's environment has ever delivered a kilogray. The consensus explanation is that radiation tolerance is a side effect of desiccation tolerance: drying generates reactive oxygen species and double-strand breaks by a similar route, so machinery evolved to survive drying happens to survive irradiation. This matters for transfer efforts, because it means the relevant adaptations are anti-desiccation adaptations that a human cell has no context for.

## Dsup and the human-cell experiment

The 2016 genome of *Ramazzottius varieornatus* identified a tardigrade-unique nuclear protein its describers named Dsup, for damage suppressor. Expressed stably in cultured human HEK293 cells, Dsup reduced X-ray-induced DNA fragmentation by roughly forty per cent and improved the cells' proliferation after irradiation and after hydrogen peroxide exposure.[^hashimoto2016]

Later work clarified the mechanism. Dsup is an intrinsically disordered protein that binds nucleosomes and coats chromatin, physically shielding DNA from hydroxyl radicals rather than accelerating repair.[^chavez2019] It is a molecular umbrella, not a better repair enzyme. That distinction predicts the limits of the effect: shielding should help most against the indirect, radical-mediated component of damage and least against the direct ionisation tracks laid down by heavy nuclei, which are precisely the component that makes [[radiation-hardening-humans|deep-space radiation]] dangerous.

Dsup has since been expressed in other systems, including plant cells, with reports of improved stress tolerance. A separate line of work delivers Dsup-encoding messenger RNA rather than a gene, using the [[lipid-nanoparticles|nanoparticle chemistry]] developed for mRNA vaccines, on the argument that transient expression in tissue at risk is enough to shield it during a known exposure such as a course of radiotherapy. This sidesteps permanent genetic modification entirely. It is a preclinical animal approach, and how long protection lasts in that format is unestablished.

```timeline
[
  { "year": "Early 20th c.", "title": "Extreme tolerance documented", "text": "Experimenters revive desiccated tardigrades and expose tuns to extreme temperatures, establishing cryptobiosis as a reproducible phenomenon rather than an anecdote." },
  { "year": "2007–2008", "title": "Survival in open space", "text": "Dried tardigrades flown on FOTON-M3 survive vacuum exposure in low Earth orbit; solar ultraviolet, not vacuum, proves to be the dominant killer." },
  { "year": "2015–2016", "title": "The horizontal gene transfer claim and its correction", "text": "A genome paper reports that a large fraction of tardigrade genes were acquired from bacteria; a competing assembly shows the signal is contamination." },
  { "year": "2016", "title": "Dsup in human cells", "text": "The Ramazzottius varieornatus genome yields Dsup, which reduces radiation-induced DNA damage when expressed in cultured human cells." },
  { "year": "2017", "title": "Disordered proteins explain drying", "text": "CAHS and related intrinsically disordered proteins are shown to vitrify tardigrade cytoplasm during desiccation, displacing trehalose as the assumed mechanism." },
  { "year": "2019", "title": "Mechanism resolved", "text": "Dsup is shown to bind nucleosomes and protect chromatin from hydroxyl radicals, making it a shield rather than a repair factor." }
]
```

## Disordered proteins and vitrified cytoplasm

Dsup is not the only transferable candidate. Many anhydrobiotic organisms accumulate trehalose, a disaccharide that replaces water around membranes and proteins, but several tardigrade species carry very little of it. Instead they produce families of intrinsically disordered proteins — cytoplasmic abundant heat soluble proteins and their relatives — that form a glass as the animal dries, immobilising cellular contents and preventing the aggregation that would otherwise destroy them.[^boothby2017]

The connection to [[proteostasis|protein quality control]] is direct: these proteins solve a folding and aggregation problem by freezing the whole system in place. The connection to [[cryonics|vitrification]] is also direct, and not coincidental — both depend on reaching a glassy state without crystallisation. Expressing CAHS proteins in yeast and bacteria has conferred partial desiccation tolerance, which is a stronger transfer result than anything achieved in mammalian cells.

## What did not transfer

The gap between the 2016 result and the claims made about it is instructive, and worth stating plainly.

Human cells expressing Dsup are not desiccation tolerant, not freeze tolerant, and not vacuum tolerant. Cryptobiosis is a whole-organism programme involving coordinated water loss, membrane restructuring, metabolic shutdown, and controlled rehydration; no single gene delivers it. The protection measured was partial, obtained under strong overexpression in an immortalised cell line, against acute X-rays and chemical oxidants rather than against the HZE particles and chronic low dose rates of the space environment. Reducing DNA breaks is also not the same as reducing their consequences, since the genomic instability listed among the [[hallmarks-of-aging]] accumulates through repair errors as well as through damage. No organism-level radioprotection has been demonstrated in a mammal, and no human has received Dsup by any route.

There are also reasons for caution about long-term expression. A foreign protein that coats chromatin is, by construction, positioned to interfere with transcription, replication, and repair complexes that need access to the same DNA. Cells expressing Dsup have shown altered growth in some reports. Any somatic strategy would face the delivery problems set out in [[somatic-gene-therapy]] and [[aav-vectors]]; any heritable strategy would face the objections set out under [[germline-editing]] and [[crispr-off-target-effects]].

## The horizontal gene transfer episode

A separate controversy shaped the field's credibility. A 2015 genome paper reported that a large fraction of tardigrade genes — on the order of a sixth — had been acquired horizontally from bacteria, and framed this as an explanation for extreme tolerance. An independent assembly published shortly afterwards showed the foreign sequences were contamination from the animals' bacterial food and environment, and put the genuine horizontal transfer fraction at a level typical for animals.[^koutsovoulos2016]

The correction was rapid, public, and civil, and it is now used as a teaching case in genome assembly. It also illustrates a pattern: tardigrades attract claims that outrun their evidence, and the organism's charisma makes those claims travel.

> [!caution] What "tardigrade DNA in humans" would and would not mean
> Coverage of the 2016 experiment frequently described human cells as having been made "part tardigrade" or as gaining the animal's survival abilities. Expressing one protein of roughly 400 amino acids in a cell line is a routine transfection experiment. It changes one property, partially, in vitro. It does not make a cell, still less a person, tolerant of anything a tardigrade tolerates.

## Reception and status

Within radiation biology the Dsup work is regarded as a legitimate and interesting finding about chromatin protection, and the mechanism has generated useful basic science about how nucleosome-binding proteins modulate radical damage. Within [[space-medicine]] it is treated as a distant possibility rather than a countermeasure. Within transhumanist and [[pantropy]] discussions it is often cited as a worked example that human radiation hardening is a live engineering option, which overstates what a single in-vitro result supports.

The broader research programme — mining [[negligible-senescence|extreme-tolerance organisms]] for transferable protective mechanisms — is sound and has precedent, and comparative biology has produced real leads in other areas. Whether tardigrades in particular will yield anything clinically useful is unresolved. The most plausible near-term application is not spaceflight at all but transient shielding of healthy tissue during cancer radiotherapy, where the exposure is acute, the target tissue is known in advance, and permanent modification is unnecessary.

## See also

- [[radiation-hardening-humans]]
- [[space-medicine]]
- [[pantropy]]
- [[negligible-senescence]]
- [[cryonics]]
- [[human-hibernation]]
- [[germline-editing]]
- [[human-enhancement]]

## References

[^hashimoto2016]: `paper` Hashimoto, T. et al. "Extremotolerant tardigrade genome and improved radiotolerance of human cultured cells by tardigrade-unique protein." *Nature Communications*, 2016. {The human-cell result is one experiment inside a genome paper, using strong overexpression in an immortalised line rather than a dedicated study.}
[^chavez2019]: `paper` Chavez, C., Cruz-Becerra, G., Fei, J., Kassavetis, G.A., Kadonaga, J.T. "The tardigrade damage suppressor protein binds to nucleosomes and protects DNA from hydroxyl radicals." *eLife*, 2019. {Biochemistry with purified Dsup and nucleosomes; it establishes the shielding mechanism and not how much protection a living cell would get.}
[^jonsson2008]: `paper` Jönsson, K.I., Rabbow, E., Schill, R.O., Harms-Ringdahl, M., Rettberg, P. "Tardigrades survive exposure to space in low Earth orbit." *Current Biology*, 2008.
[^boothby2017]: `paper` Boothby, T.C. et al. "Tardigrades Use Intrinsically Disordered Proteins to Survive Desiccation." *Molecular Cell*, 2017.
[^koutsovoulos2016]: `paper` Koutsovoulos, G. et al. "No evidence for extensive horizontal gene transfer in the genome of the tardigrade Hypsibius dujardini." *PNAS*, 2016. {An independent assembly of the same species, which is how the foreign sequences were traced to bacterial contamination rather than to the animal.}

==============================================================================
ARTICLE: targeted-drug-delivery
TITLE:   Targeted drug delivery
PORTAL:  Nanomedicine
URL:     https://futurehumanwiki.com/wiki/targeted-drug-delivery
SOURCE:  https://futurehumanwiki.com/raw/targeted-drug-delivery
==============================================================================

---
title: "Targeted drug delivery"
slug: "targeted-drug-delivery"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["nanomedicine"]
tags: ["nanomedicine", "drug delivery", "oncology", "antibody-drug conjugates", "pharmacokinetics", "nanoparticles"]
summary: "The engineering of carriers and conjugates that concentrate a drug where it is needed, a field whose clinical successes have come from chemistry rather than from nanoparticle targeting."
updated: "2026-07-27"
humanEvidence: "Antibody-drug conjugates, sugar-conjugated siRNAs and reformulated carriers are approved and in routine use in patients; no approved nanomedicine's benefit is clearly attributable to EPR targeting."
access: "Approved conjugates and reformulated carriers are prescribed routinely in oncology and rare disease at specialty-biologic prices; carriers targeted to tissues outside the liver exist only in research."
reversibility: "reversible"
issues: ["The count of approved antibody-drug conjugates in the keyfacts block carries no source."]
---

```infobox
{
  "caption": "Pharmaceutical engineering discipline",
  "rows": [
    { "label": "Origin of the idea", "value": "Paul Ehrlich's 'magic bullet', c. 1900" },
    { "label": "First approved nanocarrier", "value": "Liposomal doxorubicin, 1995" },
    { "label": "Dominant modality", "value": "Antibody–drug conjugates" },
    { "label": "Typical tumour delivery", "value": "Under 1% of injected dose" },
    { "label": "Default destination", "value": "Liver and spleen" },
    { "label": "Key failure mode", "value": "Clearance by phagocytes" },
    { "label": "Readiness", "value": "TRL 9 (approved products)" }
  ]
}
```

**Targeted drug delivery** is the attempt to concentrate a drug in diseased tissue and keep it out of healthy tissue, either by packaging it in a carrier with favourable biodistribution or by attaching it to a molecule that binds a marker on the target cell. It is the oldest ambition in pharmacology and the most persistently disappointing. The therapies that have actually reached patients work by chemistry and antibody biology rather than by the nanoparticle targeting the field spent four decades pursuing.

```keyfacts
[
  { "value": "0.7%", "label": "Median share of injected nanoparticle dose reaching a solid tumour", "note": "across a decade of published studies surveyed in 2016" },
  { "value": "1995", "label": "First approved nanoparticle drug", "note": "liposomal doxorubicin, which reduced cardiac toxicity without improving survival in most settings" },
  { "value": "12+", "label": "Antibody–drug conjugates approved worldwide", "note": "more than a dozen as of 2026; the modality that delivered on the targeting premise" }
]
```

## How it works

Two mechanisms are available, and they are usually confused.

Passive targeting relies on where a particle goes by default. Carriers between roughly 10 and 200 nanometres circulate longer than free drug, avoid renal filtration, and accumulate preferentially in tissues with leaky vasculature. Tumours were long thought to be such tissues: the enhanced permeability and retention effect, described by Matsumura and Maeda in 1986, holds that fenestrated tumour vessels let macromolecules out and poor lymphatic drainage keeps them there.[^matsumura1986] EPR became the organising rationale for cancer nanomedicine.

Active targeting attaches a ligand — an antibody, an aptamer, a peptide, a sugar — that binds a receptor enriched on the target cell. The intuitive picture is that the ligand steers the carrier. It does not. Studies comparing antibody-targeted and untargeted liposomes found that targeting made no difference to how much drug reached the tumour; what it changed was how much of the drug that arrived got inside cells rather than sitting in the interstitium.[^kirpotin2006] Targeting improves uptake at the destination; it does not improve navigation.

Clearance dominates both. Any particle entering blood adsorbs plasma proteins within seconds, forming a corona that determines its biological identity more than its engineered surface does.[^cedervall2007] Opsonised particles are removed by macrophages of the liver and spleen. Polyethylene glycol coatings slow this, and pre-existing anti-PEG antibodies, common in the general population, can accelerate it again.

## The reckoning over EPR

In 2016 a survey of a decade of published animal studies found that the median proportion of an injected nanoparticle dose reaching a solid tumour was about 0.7 percent.[^wilhelm2016] The number was contested — critics argued that percentage of injected dose is the wrong metric, since what matters is concentration at the target relative to toxicity elsewhere — but the central point survived the argument: the accumulation the field had assumed was an order of magnitude smaller than advertised.

A 2020 study then challenged the mechanism itself. Using models in which endothelial gaps could be counted, it concluded that the great majority of nanoparticles entering tumours did so by active transport through endothelial cells rather than by leaking through gaps between them.[^sindhwani2020] If entry is an energy-dependent cellular process rather than passive diffusion through holes, then the design rules derived from EPR — optimise size and circulation time, wait for leakage — were aimed at the wrong variable.

EPR is also far more variable in humans than in the mouse xenografts where it was characterised. Rodent tumours grow fast, are highly vascularised, and are large relative to the animal; human tumours are heterogeneous, often fibrotic, and frequently have high interstitial pressure that opposes convective entry.

> [!caution] Contested
> Whether EPR is a usable clinical phenomenon remains disputed. Some groups argue it is real but patient-specific and could be exploited by selecting patients whose tumours show high permeability, using the imaging and circulating-marker methods described in [[nanoparticle-diagnostics]]. Others argue that the effect is too weak and too inconsistent in humans to build a therapeutic strategy on. No approved nanomedicine's benefit is clearly attributable to EPR.

## What actually worked

Antibody–drug conjugates are the modality that delivered. An antibody against a tumour antigen carries a cytotoxic payload too potent to be given systemically, joined by a linker designed to hold in blood and release inside the cell. Gemtuzumab ozogamicin was approved in 2000, withdrawn in 2010 over toxicity and lack of demonstrated benefit, and reapproved in 2017 at a lower, fractionated dose — a history that illustrates how narrow the therapeutic window is. Brentuximab vedotin, ado-trastuzumab emtansine, and trastuzumab deruxtecan followed, the last extending benefit to tumours expressing only low levels of its target because the released payload diffuses into neighbouring cells.

Conjugation to a sugar ligand achieved something similar for RNA drugs. Attaching a triantennary N-acetylgalactosamine group to a small interfering RNA directs it to the asialoglycoprotein receptor, which hepatocytes display in enormous numbers. The result is subcutaneous dosing, liver-specific silencing, and durable effect from infrequent injections — the most complete example of receptor-directed delivery in medicine.

Nanoparticle carriers succeeded where the goal was reformulation rather than targeting. Liposomal doxorubicin reduced cardiotoxicity. Albumin-bound paclitaxel eliminated a toxic solvent. [[lipid-nanoparticles]] made mRNA vaccines and the first systemic in vivo genome-editing therapies possible, and they work because they go to the liver reliably, not because they go anywhere selectively.

What none of these resembles is a carrier that navigates. The addressable containers of [[dna-nanotechnology]] come closest to conditional release on a molecular signal, and they have reached mice rather than patients. Where cells must be modified precisely, the field has generally moved the targeting problem out of the body altogether: [[casgevy]] and other ex vivo therapies remove the cells, edit them in a facility, and put them back, which is expensive and immunologically demanding but sidesteps delivery entirely.

## Limitations

The liver is the recurring problem. Its fenestrated endothelium and resident macrophages take up most of what circulates, which is convenient for hepatic targets and an obstacle for everything else. Reformulating a particle to shift its protein corona can redirect it toward lung or spleen, and antibody-decorated particles have reached specific cell types in animals, but no carrier has been approved on the strength of selective delivery to a tissue outside the liver.

The brain is harder still. The blood–brain barrier excludes almost all particles; the approaches with the best evidence use receptor-mediated transcytosis through the transferrin receptor, or temporary mechanical opening of the barrier with focused ultrasound and injected microbubbles, the same modality used for the non-thermal brain stimulation covered in [[non-invasive-neuromodulation]]. Both are in clinical development rather than routine use.

Even at the target, a carrier must escape the endosome. For nucleic acid cargoes this is the dominant inefficiency, with only a small percentage of internalised material reaching the cytosol.

Toxicity does not disappear when a drug is targeted. Conjugates cause on-target damage wherever the antigen is expressed, and payload released prematurely in circulation causes conventional cytotoxicity; interstitial lung disease and ocular toxicity have limited specific ADCs. Complement activation by particle surfaces can cause acute infusion reactions unrelated to the drug.

Cost is a structural limitation rather than a technical one. Conjugates and engineered carriers are biologics with complex manufacturing, priced accordingly, and the diffusion problem they create is the subject of [[access-and-inequality]]. A delivery technology that only works inside well-funded health systems solves a narrower problem than its developers describe.

## Outlook

The lesson the field draws from its own record is that delivery is not a finishing step applied to a finished drug — it determines what drugs are possible. Genome editing is the clearest case: the set of diseases addressable by [[crispr-cas9]] and [[base-editing]] as of 2026 is essentially the set of diseases whose causal gene is expressed in hepatocytes, because that is where the carriers go. Extending editing to muscle, immune cells, or the central nervous system is a delivery problem, not an editing one, and the same is true of proposals such as [[gene-therapy-for-aging]] or the systemic administration of [[senolytics]].

Two directions look most likely to change that. Engineered capsids of the kind described in [[aav-vectors]], and cell-type-selective lipids, are converging on the ability to specify a destination tissue at formulation time, an incremental route that has already produced organ-selective particles in animals and that would widen the reach of [[somatic-gene-therapy]] more than any change to the editing enzymes themselves. Cell-based carriers — red cell membranes, macrophages, engineered bacteria — borrow biology's own trafficking rather than trying to replicate it, and shade into the externally guided devices covered in [[microrobots-in-medicine]]. Neither approach resembles the autonomous targeting machine imagined in [[medical-nanorobots]], and the history of this field is a long argument that the imagined version underrates how thoroughly the body sorts what enters it.

## See also

- [[lipid-nanoparticles]]
- [[medical-nanorobots]]
- [[dna-nanotechnology]]
- [[nanoparticle-diagnostics]]
- [[aav-vectors]]
- [[somatic-gene-therapy]]
- [[microrobots-in-medicine]]
- [[access-and-inequality]]

## References

[^matsumura1986]: `paper` Matsumura, Y. and Maeda, H. "A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs." *Cancer Research*, 1986.
[^kirpotin2006]: `paper` Kirpotin, D.B. et al. "Antibody targeting of long-circulating lipidic nanoparticles does not increase tumor localization but does increase internalization in animal models." *Cancer Research*, 2006.
[^cedervall2007]: `paper` Cedervall, T. et al. "Understanding the nanoparticle–protein corona using methods to quantify exchange rates and affinities of proteins for nanoparticles." *PNAS*, 2007.
[^wilhelm2016]: `paper` Wilhelm, S. et al. "Analysis of nanoparticle delivery to tumours." *Nature Reviews Materials*, 2016. {A survey of published animal studies, so the 0.7 percent figure is a median across animal tumour models rather than a measurement in patients.}
[^sindhwani2020]: `paper` Sindhwani, S. et al. "The entry of nanoparticles into solid tumours." *Nature Materials*, 2020. {The mechanism was established in mouse tumour models in which endothelial gaps could be counted directly; it has not been measured in human tumours.}

==============================================================================
ARTICLE: technological-singularity
TITLE:   Technological singularity
PORTAL:  Foundational Concepts
URL:     https://futurehumanwiki.com/wiki/technological-singularity
SOURCE:  https://futurehumanwiki.com/raw/technological-singularity
==============================================================================

---
title: "Technological singularity"
slug: "technological-singularity"
type: "concept"
status: "speculative"
horizon: "indefinite"
categories: ["foundations", "minds"]
tags: ["singularity", "artificial intelligence", "forecasting", "intelligence explosion", "superintelligence", "recursive self-improvement"]
summary: "The hypothesized point at which machine intelligence begins improving itself fast enough that technological change outruns human ability to predict or steer it."
updated: "2026-07-27"
issues: ["The claim that automated AI research is now measured rather than speculated carries no source."]
---

```infobox
{
  "caption": "Forecasting hypothesis in AI and futures studies",
  "rows": [
    { "label": "Earliest statement", "value": "Stanisław Ulam, 1958" },
    { "label": "Mechanism formalized", "value": "I. J. Good, 1965" },
    { "label": "Term popularized", "value": "Vernor Vinge, 1993" },
    { "label": "Best-known date claim", "value": "2045", "link": "/wiki/ray-kurzweil" },
    { "label": "Core mechanism", "value": "Recursive self-improvement" },
    { "label": "Field", "value": "AI forecasting, philosophy of mind" },
    { "label": "Status", "value": "Undemonstrated" }
  ]
}
```

**Technological singularity** is the name given to a hypothesized transition in which machine intelligence becomes capable of improving itself, sets off a feedback loop of accelerating capability gain, and thereby produces a future that human beings cannot forecast because they cannot model the agents driving it. The idea has three distinguishable versions, which are often conflated: a claim about a specific mechanism, a claim about the shape of long-run technological growth, and a claim about the limits of prediction. Nothing about it has been demonstrated, and the disagreement about it runs between serious people rather than between experts and cranks.

```keyfacts
[
  { "value": "1965", "label": "Intelligence explosion described", "note": "I. J. Good, working alongside the British code-breaking community" },
  { "value": "1993", "label": "Term popularized", "note": "Vernor Vinge, in an essay written for a NASA-sponsored symposium" },
  { "value": "2045", "label": "Kurzweil's stated date", "note": "one forecaster's extrapolation, not a position of any scientific body" }
]
```

## Overview

The word does two jobs. In its mathematical sense a singularity is a point at which a function's value ceases to be defined — a curve that goes vertical. In its use here it names an event horizon: a moment past which the ordinary methods of forecasting stop returning answers. Both senses are metaphors borrowed to describe the same conjecture, which is that intelligence is the input to technological progress, that intelligence is itself a technology, and that a self-improving process therefore has no obvious external brake.

What follows from that conjecture depends heavily on which version is being argued. The mechanism version makes a strong, falsifiable-in-principle claim about what a sufficiently capable system would do. The growth version makes a weaker claim about historical trend lines. The horizon version makes almost no empirical claim at all and is closer to an admission of ignorance. Critics frequently attack one and proponents defend another.

## Origins

The earliest recognizable statement is a remark attributed to John von Neumann and reported by Stanisław Ulam in a 1958 memorial tribute: that the accelerating progress of technology gives "the appearance of approaching some essential singularity in the history of the race beyond which human affairs, as we know them, could not continue."[^ulam1958] No further elaboration survives, and von Neumann himself never published on the subject.

The mechanism was set out by the statistician I. J. Good in 1965. Good defined an "ultraintelligent machine" as one that surpasses all human intellectual activity, observed that machine design is itself an intellectual activity, and concluded that such a machine would design better machines: "there would then unquestionably be an 'intelligence explosion', and the intelligence of man would be left far behind."[^good1965] He added, in the same paragraph, that the first ultraintelligent machine would be the last invention humanity need make, provided it could be kept controllable. The safety caveat is present in the founding text.

Vernor Vinge, a mathematician and science-fiction writer, gave the idea its current name and its horizon framing in a 1993 essay written for a symposium sponsored by NASA.[^vinge1993] Vinge argued that superhuman intelligence could arrive by several routes — machine AI, [[human-ai-merger|human–computer symbiosis]], biological [[intelligence-amplification]], networked collective intelligence, or the [[whole-brain-emulation|emulation]] of a scanned human brain — and that any of them ends the era in which human models of the future have purchase. He wrote that the technological means would exist within thirty years, and added that he would be surprised if the event occurred before 2005 or after 2030.

The idea's social carriers were the [[extropianism|extropian]] community and, from the late 1990s, the movement described under [[singularitarianism]], which is why the singularity is discussed in the same literature as [[cryonics]] and [[transhumanism]] despite having no logical dependence on either.

```timeline
[
  { "year": "1958", "title": "Von Neumann's remark", "text": "Ulam's tribute records a conversation about accelerating progress approaching 'some essential singularity in the history of the race'." },
  { "year": "1965", "title": "Intelligence explosion", "text": "I. J. Good describes recursive self-improvement by an ultraintelligent machine, and notes that control is the open problem." },
  { "year": "1993", "title": "Vinge's essay", "text": "The term 'technological singularity' is popularized, framed as a prediction horizon rather than a growth curve." },
  { "year": "2005", "title": "The curve-fitting version", "text": "Ray Kurzweil's The Singularity Is Near argues from exponential trends across many technologies and dates the event to 2045." },
  { "year": "2008", "title": "The FOOM debate", "text": "Robin Hanson and Eliezer Yudkowsky publicly argue over whether takeoff would be local and fast or distributed and gradual." },
  { "year": "2014", "title": "Superintelligence", "text": "Nick Bostrom's book reframes the discussion around control and alignment, moving it toward mainstream policy attention." },
  { "year": "2020–2022", "title": "Scaling laws", "text": "Empirical relationships between compute, data and loss give AI capability forecasting a quantitative footing it previously lacked." },
  { "year": "2023–2026", "title": "Automated research as the test", "text": "Attention shifts to whether models can meaningfully contribute to AI research itself, the step the mechanism argument requires." }
]
```

## The three arguments

### Intelligence explosion

Good's argument is a claim about a positive feedback loop with a short time constant. If a system's capability at AI research is a function of its intelligence, and its intelligence is the output of AI research, then improvement compounds. The strength of the loop is what matters: if each generation of self-improvement yields less than the previous one, the process converges to a ceiling and there is no explosion. [[nick-bostrom]] formalized this as the ratio of "optimization power" applied to "recalcitrance", and noted that the conclusion is sensitive to assumptions no one can currently measure.[^bostrom2014]

The 2008 exchange between Robin Hanson and Eliezer Yudkowsky, usually called the FOOM debate, is still the clearest statement of the disagreement. Yudkowsky argued that the loop could run inside a single system over days or weeks, because human brains are a poor design and much of the improvement is architectural. Hanson argued that capability is embodied in accumulated knowledge and institutions rather than in a single agent, so gains would diffuse across an economy and look like fast but recognizable growth. Neither position has been settled by anything that has happened since.

### Accelerating returns

[[ray-kurzweil]]'s version replaces the mechanism with a trend. He argues that information technologies improve exponentially, that the exponent itself grows, and that fitting curves to price-performance data across computing, sequencing, and communications yields a date at which machine intelligence exceeds the aggregate of human intelligence.[^kurzweil2005] This is the version most often encountered in popular coverage, and the one most vulnerable to technical objection. It depends on choices about which metrics to plot, on treating [[accelerating-change|exponential trends]] as laws rather than as descriptions of a period, and on equating computational operations per second with intelligence — an equation that no theory of cognition supports.

### Prediction horizon

Vinge's version is epistemic. If the future is driven by agents whose reasoning cannot be simulated by present-day humans, then statements about that future have no evidential basis, including optimistic ones. This is the most defensible formulation and also the least useful: it recommends humility without recommending action, and it is compatible with the singularity never occurring.

> [!debate] Which claim is on trial
> Proponents typically defend the horizon claim, which is nearly unfalsifiable, and critics typically attack the accelerating-returns claim, which is the weakest. The mechanism claim — that recursive self-improvement has a strong enough feedback coefficient to run away — is the one that actually decides the question, and it is the one for which there is no direct evidence in either direction.

## Evidence and state of play

Nothing has demonstrated recursive self-improvement in a machine. What exists as of 2026 is a body of scaling results showing that model loss falls predictably with compute, parameters, and data over many orders of magnitude,[^hoffmann2022] and a record of rapid gains on benchmarks that were designed to be hard. Both are relevant but neither is the claim.

Benchmark progress is contested evidence for a specific reason: a benchmark measures the thing it measures, and saturation can reflect either general capability or contamination and task-specific optimization. François Chollet's argument that intelligence should be measured as skill-acquisition efficiency rather than skill, and the abstraction-and-reasoning benchmark he built on that basis, remains the sharpest statement of why leaderboard movement underdetermines the question.[^chollet2019] Systems that score well on graduate-level examinations while failing at tasks requiring novel abstraction are consistent with several very different underlying stories.

The economic evidence is weaker still. William Nordhaus assembled a set of tests for whether the United States economy is on a path toward an information-technology-driven growth singularity — accelerating productivity, a rising capital share, falling information-goods prices propagating to output — and found that most of them point away from it, though he framed the result as showing the event is not imminent rather than impossible.[^nordhaus2021] Measured total factor productivity growth in advanced economies has been sluggish for two decades, which is difficult to reconcile with a technology base said to be on a vertical curve.

## Criticism

The most common technical objection is the complexity brake, argued by Paul Allen and Mark Greaves: as scientific understanding advances, each further increment requires disproportionately more work, and software progress in particular has never displayed the exponential character of hardware.[^allen2011] Brains are not modular the way the argument assumes, and the difficulty of understanding them may rise faster than the intelligence brought to bear.

A second objection targets the identification of intelligence with a scalar quantity. There is no accepted definition of general intelligence that supports statements like "a thousand times smarter", and the assumption that cognitive capability is a single dimension along which one can travel without bound is an assumption, not a finding. [[artificial-general-intelligence]] as a term inherits the same problem, as does the [[posthuman]] literature's talk of capacities exceeding any current human's.

A third points to physical constraints. Self-improvement in software still requires experiments, fabrication plants, energy, and time; a system cannot think its way past the need to run an assay or build a chip. Biology in particular resists compression: the pace of a drug programme is set by trial duration, which is why [[ai-drug-discovery|machine learning applied to drug discovery]] has put candidates into trials without yet producing an approved medicine. Scales like [[technology-readiness-level]] exist precisely because the gap between a laboratory demonstration and a working system is dominated by physical validation rather than by insight. Proponents reply that a sufficiently capable system would compress those loops rather than escape them, and point to [[ai-protein-design|learned protein structure prediction]] as a problem that resisted fifty years of effort and then fell quickly. That reply is plausible and unquantified.

Finally, the field's forecasting track record is poor. Predictions of human-level machine intelligence have clustered roughly twenty to forty years from whenever they were made, across seven decades of very different technical conditions. That pattern is itself evidence about how these estimates are generated.

> [!caution] What is not established
> No system has improved its own architecture without human direction. No theory predicts the feedback coefficient of such a loop. No consensus definition of general intelligence exists that would let anyone say when the threshold has been crossed. The singularity remains a conjecture with a long intellectual pedigree and no experimental support.

## What large language models changed

Large models moved two things and left the central question untouched. They made the possibility of broadly capable systems concrete enough that governments and mainstream researchers now treat it as a policy subject rather than a genre convention, and they supplied a quantitative forecasting substrate in the form of scaling relationships. Both are real changes from the situation in 2015.

What they did not do is exhibit the mechanism. Contemporary models are trained by humans on human-generated data, and their improvement is driven by capital expenditure on compute and by research conducted by people. Whether models can substantially automate that research — write the code, design the experiments, evaluate the results, and thereby shorten their own development cycle — is the step Good's argument requires, and it is now the object of explicit measurement rather than speculation. Progress on it as of 2026 is real but partial, and its interpretation is disputed by people with access to the same evidence.

The philosophical questions were not moved at all. Whether a system that behaves intelligently is conscious remains open, as [[machine-consciousness]] and the state of research on [[neural-correlates-of-consciousness]] make clear, and it bears directly on whether a post-singularity world contains anyone whose welfare counts. David Chalmers's analysis, which grants the argument its structure and then asks what follows for personal identity and value, remains the most careful treatment.[^chalmers2010]

## Outlook

The practical significance of the hypothesis has shifted from prediction to preparation. [[existential-risk]] research, the [[differential-technological-development]] proposal, and the alignment agenda all take the mechanism seriously enough to plan for it without asserting a date; [[effective-accelerationism]] rejects the whole framing and reaches the opposite policy conclusion from the same premise about capability growth. Any serious treatment of the [[future-of-humanity]] has to take a position on which of these is right. That is a defensible posture: the argument's conclusion does not depend on the timing, and its main policy implication — that the transition would be difficult to reverse — holds whether it arrives in a decade or never.

The open empirical question is narrow and stateable. Does the loop close? If systems can perform the research that improves systems, at a quality and cost that beats the humans currently doing it, then Good's conditional has its antecedent satisfied and the argument runs. If they cannot, the accelerating-returns curves are a description of a growth episode in semiconductors, and the singularity is a hypothesis about a mechanism that was never engaged.

## See also

- [[singularitarianism]]
- [[artificial-general-intelligence]]
- [[accelerating-change]]
- [[whole-brain-emulation]]
- [[nick-bostrom]]
- [[ray-kurzweil]]
- [[existential-risk]]
- [[effective-accelerationism]]

## References

[^ulam1958]: `paper` Ulam, S. "Tribute to John von Neumann." *Bulletin of the American Mathematical Society*, 1958.
[^good1965]: `paper` Good, I.J. "Speculations Concerning the First Ultraintelligent Machine." *Advances in Computers*, vol. 6, 1965.
[^vinge1993]: `paper` Vinge, V. "The Coming Technological Singularity: How to Survive in the Post-Human Era." VISION-21 Symposium, 1993.
[^kurzweil2005]: `book` Kurzweil, R. *The Singularity Is Near: When Humans Transcend Biology*. Viking, 2005.
[^bostrom2014]: `book` Bostrom, N. *Superintelligence: Paths, Dangers, Strategies*. Oxford University Press, 2014.
[^chalmers2010]: `paper` Chalmers, D. "The Singularity: A Philosophical Analysis." *Journal of Consciousness Studies*, 2010.
[^allen2011]: `news` Allen, P. and Greaves, M. "The Singularity Isn't Near." *MIT Technology Review*, 2011. {An opinion essay rather than a study; it argues from the history of scientific and software progress and reports no new measurements.}
[^nordhaus2021]: `paper` Nordhaus, W. "Are We Approaching an Economic Singularity? Information Technology and the Future of Economic Growth." *International Economic Review*, 2021.
[^hoffmann2022]: `paper` Hoffmann, J. et al. "Training Compute-Optimal Large Language Models." *Advances in Neural Information Processing Systems*, 2022. {The predictable quantity is training loss, which is not itself a measure of any downstream capability.}
[^chollet2019]: `preprint` Chollet, F. "On the Measure of Intelligence." arXiv preprint, 2019. {Proposes a definition and a benchmark rather than reporting an experiment, and was posted without peer review.}

==============================================================================
ARTICLE: technology-readiness-level
TITLE:   Technology readiness level
PORTAL:  Foundational Concepts
URL:     https://futurehumanwiki.com/wiki/technology-readiness-level
SOURCE:  https://futurehumanwiki.com/raw/technology-readiness-level
==============================================================================

---
title: "Technology readiness level"
slug: "technology-readiness-level"
type: "concept"
status: "established"
horizon: "present"
categories: ["foundations"]
tags: ["forecasting", "translation", "regulation", "engineering", "clinical trials", "methodology"]
summary: "A nine-point scale, originating at NASA, that rates how far a technology has moved from observed basic principles to proven operational use."
updated: "2026-07-27"
issues: ["Adoption dates for ISO 16290 and the European Commission are stated without citations."]
---

```infobox
{
  "caption": "Assessment scale in engineering management",
  "rows": [
    { "label": "Origin", "value": "NASA, mid-1970s" },
    { "label": "Nine-level form", "value": "Mankins, 1995" },
    { "label": "Range", "value": "TRL 1 to TRL 9" },
    { "label": "Standardised as", "value": "ISO 16290:2013" },
    { "label": "Major adopters", "value": "NASA, US DoD, European Commission" },
    { "label": "Known weakness", "value": "One dimension, assumes linearity" }
  ]
}
```

**Technology readiness level** (TRL) is a nine-point ordinal scale that describes how far a technology has progressed from the observation of basic principles toward demonstrated operation in its intended environment. It was created to give programme managers a common vocabulary for maturity, so that a proposal claiming a component was "ready" could be interrogated about what exactly had been demonstrated, at what scale, and under what conditions. This wiki assigns a TRL to every article typed as a technology or intervention, which makes the scale's assumptions and its failure modes directly relevant to how those articles should be read.

```keyfacts
[
  { "value": "9", "label": "Levels in the standard scale", "note": "expanded from an original seven at NASA" },
  { "value": "~1 in 7", "label": "Drugs entering Phase I that reach approval", "note": "Wong, Siah and Lo, across 2000–2015" },
  { "value": "1,026", "label": "Experimental acute-stroke treatments tested in animals", "note": "O'Collins et al.; essentially none succeeded in humans" }
]
```

## The scale

The canonical definitions are: 1, basic principles observed; 2, technology concept formulated; 3, experimental proof of concept; 4, component validated in a laboratory; 5, component validated in a relevant environment; 6, system or subsystem demonstrated in a relevant environment; 7, prototype demonstrated in an operational environment; 8, system complete and qualified; 9, system proven in operation.

Two words carry most of the weight. *Relevant environment* means conditions that reproduce the stresses of the real setting in the respects that matter — vacuum and thermal cycling for a spacecraft component, a living organism for a therapeutic. *Operational environment* means the actual setting, with all of its uncontrolled variables. The distinction between levels 5 and 7 is therefore a claim about how much of the real world has been let in, and it is where assessments most often disagree.

## Origins and spread

Stan Sadin at NASA Headquarters introduced a seven-level formulation in the mid-1970s to compare the maturity of candidate technologies for future missions. John Mankins expanded and documented the nine-level version in a 1995 NASA white paper that remains the reference text for the definitions.[^mankins1995]

Adoption spread through defence acquisition after the US Government Accountability Office recommended maturity assessment as a corrective to programmes that committed to production before their key technologies worked. The US Department of Defense built TRLs into its acquisition process and added companion scales, notably manufacturing readiness levels, on the observation that a technology can be demonstrated and still be unbuildable at volume. The European Commission adopted the definitions across its research framework programmes in 2014, which made TRL a routine field in grant applications across European science. ISO published a space-systems standard for the levels and their assessment criteria in 2013.

The consequence is that TRL now functions as an administrative currency far outside the domain it was designed for, including in fields where its central metaphor — a component moving toward integration in a system — does not obviously apply.

## Adaptation to biomedicine

Medical product development has its own maturity vocabulary: discovery, lead optimisation, preclinical, Phase I, II and III, and post-market. Mapping this onto TRLs is possible but lossy, and several agencies have published crosswalks for medical countermeasure programmes.

```compare
{
  "columns": ["Aerospace reading", "Biomedical analogue"],
  "rows": [
    { "label": "TRL 3", "values": ["Analytical proof of concept", "Target validated in cells"] },
    { "label": "TRL 4", "values": ["Component validated in the lab", "Efficacy in a small-animal model"] },
    { "label": "TRL 6", "values": ["Subsystem in a relevant environment", "GLP toxicology; large-animal or first-in-human safety"] },
    { "label": "TRL 7", "values": ["Prototype in operation", "Efficacy in a controlled clinical trial"] },
    { "label": "TRL 9", "values": ["Proven in operation", "Approved and in routine clinical use"] }
  ]
}
```

A parallel scheme used in translational medicine divides the pipeline into T0 through T4 phases running from basic discovery through to population health impact, a framing developed for genomic medicine and now used more broadly.[^khoury2007] It has the advantage of naming the last step, dissemination into practice, which TRL 9 elides.

## The valley of death

The gap between laboratory validation and operational prototype — roughly TRL 4 through 7 — is where most technologies die, and the reason is structural rather than scientific. Basic research is funded by science agencies that stop at proof of concept. Product development is funded by firms and investors who require a defined path to revenue. Between the two sits work that is too applied to be publishable and too unproven to be financeable, and it is expensive.

In biomedicine the valley has a measurable shape. Across a large sample of development programmes, roughly one in seven drugs entering Phase I trials reached approval, with the largest single loss at Phase II, where efficacy is tested for the first time in patients.[^wong2019] Fields relevant to this wiki show the pattern clearly. [[senolytics]] cleared TRL 4 convincingly in mice and have produced modest or null results in the small human trials run so far. [[gene-therapy-for-aging]] has strong animal data and no controlled human efficacy evidence. [[organ-bioprinting]] and [[tissue-engineering]] are stuck between validated tissue constructs and anything approaching a vascularised organ. [[artificial-womb]] devices have kept premature lambs alive for weeks in animal studies since the late 2010s, and as of 2026 no first-in-human trial has been publicly reported.

## Limits of the concept

A 2015 review of TRL practice across aerospace, defence and other sectors identified recurring problems: the scale collapses many dimensions of maturity into one number, it assumes a linear progression that real programmes do not follow, and it is routinely used as a risk measure although it says nothing about how difficult the remaining steps are.[^olechowski2015]

That last point is the important one. TRL measures distance travelled, not distance remaining. Two technologies at TRL 4 may be separated by two years and by thirty. Nothing in the scale distinguishes a component awaiting an engineering iteration from one awaiting a scientific result nobody knows how to obtain. Attempts to patch this have produced companion scales — integration readiness, system readiness, manufacturing readiness — each of which adds information and none of which is widely used outside defence.

The scale is also gameable. Because funding decisions depend on it, self-assessed TRLs drift upward, and the boundary terms are elastic enough to support optimistic readings. Independent technology readiness assessments exist precisely because self-reporting is unreliable.

> [!caution] TRL is not a timeline
> A high TRL implies past demonstration; it implies nothing about years to deployment. Some TRL 6 technologies reach routine use in three years and some never do. Any article on this wiki that pairs a TRL with a horizon is pairing an assessment with a guess, and the two should be read differently.

## Why biology fits badly

The scale was designed for engineered systems that can be decomposed. A valve is tested alone, then in a subsystem, then in a vehicle, and each test is informative about the next. Biology resists this. The "component" is embedded in a regulatory network that cannot be held constant, the "relevant environment" is an organism whose relevance to humans is exactly what is in question, and integration effects dominate.

The canonical illustration is acute stroke, where over a thousand candidate treatments showed benefit in animal models and essentially none succeeded in humans.[^ocollins2006] Nothing in a TRL assessment would have distinguished those failures in advance, because each had genuinely been validated in a relevant environment by the scale's own definition. The same worry applies across geroscience: mouse lifespan extension is real, replicable and a weak predictor of human effect, which is why [[aging-biomarkers]] and the surrogate endpoints discussed in the [[geroscience-hypothesis]] are being pursued as substitutes for waiting decades to find out.

A further mismatch is that biological interventions can be simultaneously mature and unproven. [[aav-vectors]] and [[lipid-nanoparticles]] are TRL 9 as delivery platforms and TRL 3 for many of the payloads they might carry. The technology and its application have separate readiness, and the scale offers only one slot.

## Use on this wiki

TRLs on this site are assigned conservatively and mean what the standard definitions say. A rating of 1 to 3 marks a subject that exists as argument, calculation or laboratory proof of concept — the position of [[medical-nanorobots]] and [[cryonics]] among the technologies covered here. Subjects typed as concepts rather than technologies, such as [[respirocytes]] and [[whole-brain-emulation]], carry no rating at all, which is itself informative: there is no artefact to assess. A rating of 4 to 6 marks animal or early human evidence, the position of most of the longevity interventions covered here. A rating of 7 to 9 marks clinical or commercial deployment, which applies to a small minority of the wiki's subjects, among them [[cochlear-implant|cochlear implants]], [[casgevy]] and [[induced-pluripotent-stem-cells|iPSC]]-derived products in their first approved uses.

## Outlook

The scale's persistence despite well-documented flaws is a fact about institutions rather than about measurement. It survives because it is cheap, comparable across proposals and legible to non-specialists, and because no proposed replacement has offered the same three properties. The open question is whether a maturity measure can be built that captures remaining difficulty rather than accumulated progress — which would require forecasting the very thing the scale exists to avoid forecasting, and which is why forty years of criticism have produced supplements rather than successors.

## See also

- [[geroscience-hypothesis]]
- [[aging-biomarkers]]
- [[accelerating-change]]
- [[differential-technological-development]]
- [[tissue-engineering]]
- [[senolytics]]
- [[future-of-humanity]]

## References

[^mankins1995]: `report` Mankins, J. C. "Technology Readiness Levels: A White Paper." NASA Office of Space Access and Technology, 1995. {An agency white paper rather than a ratified standard; it became the reference text by adoption elsewhere.}
[^khoury2007]: `paper` Khoury, M. J. et al. "The Continuum of Translation Research in Genomic Medicine: How Can We Accelerate the Appropriate Integration of Human Genome Discoveries into Health Care and Disease Prevention?" *Genetics in Medicine*, 2007.
[^wong2019]: `paper` Wong, C. H., Siah, K. W. and Lo, A. W. "Estimation of Clinical Trial Success Rates and Related Parameters." *Biostatistics*, 2019. {An estimate pooled across industry pipelines; success rates differ widely between therapeutic areas.}
[^olechowski2015]: `paper` Olechowski, A., Eppinger, S. D. and Joglekar, N. "Technology Readiness Levels at 40: A Study of State-of-the-Art Use, Challenges, and Opportunities." *Proceedings of PICMET*, 2015.
[^ocollins2006]: `paper` O'Collins, V. E. et al. "1,026 Experimental Treatments in Acute Stroke." *Annals of Neurology*, 2006. {A survey of the preclinical literature: the treatments counted were tested in animal models of stroke, not in people.}

==============================================================================
ARTICLE: telomeres-and-telomerase
TITLE:   Telomeres and telomerase
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/telomeres-and-telomerase
SOURCE:  https://futurehumanwiki.com/raw/telomeres-and-telomerase
==============================================================================

---
title: "Telomeres and telomerase"
slug: "telomeres-and-telomerase"
type: "concept"
status: "established"
horizon: "present"
categories: ["longevity"]
tags: ["aging", "telomeres", "cancer", "cell biology", "biomarkers", "mechanisms"]
summary: "Repetitive DNA caps that shorten with each cell division, and the enzyme that rebuilds them, linking division limits, cancer risk, and a contested marker of aging."
updated: "2026-07-27"
humanEvidence: "Human genetics ties telomerase mutations to the short telomere syndromes, and Mendelian randomization ties inherited longer telomeres to higher cancer risk; no trial has tested telomerase therapy for aging in people."
issues: ["The self-experimental telomerase gene therapy episode is described without a source."]
---

```infobox
{
  "caption": "Chromosome structure and enzyme",
  "rows": [
    { "label": "Human repeat sequence", "value": "TTAGGG" },
    { "label": "Enzyme", "value": "Telomerase (TERT + TERC)" },
    { "label": "Discovered", "value": "1985" },
    { "label": "Discovered by", "value": "Carol Greider, Elizabeth Blackburn" },
    { "label": "Nobel Prize", "value": "2009, Physiology or Medicine" },
    { "label": "Protective complex", "value": "Shelterin" },
    { "label": "Status as aging marker", "value": "Contested" }
  ]
}
```

**Telomeres and telomerase** are, respectively, the repetitive nucleotide sequences and associated proteins that cap the ends of linear chromosomes, and the reverse transcriptase that extends them. Telomeres shorten each time a somatic cell divides; when they become critically short the cell triggers a DNA damage response and enters [[cellular-senescence|senescence]] or apoptosis. Telomere attrition is one of the [[hallmarks-of-aging]], but its causal weight in normal human aging is smaller and more disputed than popular accounts suggest.

```figure
{"key": "telomeres-chromosome", "caption": "Telomeres cap each chromosome end. They shorten with division in most human somatic cells, which makes them a correlate of ageing and not obviously a cause."}
```

## Structure and function

Human telomeres consist of thousands of tandem TTAGGG repeats ending in a single-stranded 3′ overhang that folds back to form a lasso-like t-loop. Six proteins — TRF1, TRF2, POT1, TIN2, TPP1 and RAP1, collectively the shelterin complex — bind the repeats and hide the chromosome end from the repair machinery. Without shelterin, the cell reads a natural chromosome terminus as a double-strand break and attempts to fuse it to another chromosome, producing the end-to-end fusions and breakage cycles that characterize genomic crisis.

The capping function, not the length as such, is what matters. A telomere can be long and dysfunctional if shelterin is disrupted, and short telomeres in a cell with intact capping may not signal damage until a threshold is crossed. The relationship between average telomere length and cellular behaviour is therefore indirect: it is the shortest telomere in a cell, not the mean, that determines when arrest occurs.

## The end-replication problem and the Hayflick limit

DNA polymerase cannot replicate the extreme 5′ end of a lagging strand, so each round of replication removes a small stretch of terminal sequence. James Watson described the geometry of this in 1972, and Alexey Olovnikov independently proposed that it would impose a division counter on somatic cells and explain the replicative limit Leonard Hayflick had observed in cultured human fibroblasts.[^olovnikov1973] Human somatic cells in culture typically manage a few dozen population doublings before arresting.

Introducing the catalytic subunit of telomerase into normal human cells prevents that arrest and allows indefinite proliferation, which confirmed the causal link between telomere maintenance and the replicative limit.[^bodnar1998] This is the basis of most immortalized cell lines used in research, and telomerase reactivation is part of what makes [[induced-pluripotent-stem-cells|induced pluripotent stem cells]] proliferate without limit.

## Telomerase

```timeline
[
  { "year": "1978", "title": "Telomere sequence determined", "text": "Elizabeth Blackburn and Joseph Gall identify the repeated sequence at the chromosome ends of the ciliate Tetrahymena." },
  { "year": "1985", "title": "Telomerase identified", "text": "Carol Greider and Blackburn detect an enzyme activity in Tetrahymena extracts that adds telomeric repeats to chromosome ends." },
  { "year": "1998", "title": "Human cells immortalized", "text": "Bodnar and colleagues show that expressing hTERT in normal human cells extends their replicative lifespan indefinitely." },
  { "year": "2009", "title": "Nobel Prize", "text": "Blackburn, Greider and Jack Szostak share the prize in Physiology or Medicine for the discovery of telomeres and telomerase." },
  { "year": "2024", "title": "Telomerase inhibitor approved", "text": "Imetelstat, which targets the telomerase RNA template, is approved in the United States for anaemia in lower-risk myelodysplastic syndromes." }
]
```

Telomerase is a ribonucleoprotein: the protein subunit TERT copies a template carried within the RNA subunit TERC. It is active in germ cells, embryonic stem cells, and — at lower levels — in adult stem and progenitor compartments and activated lymphocytes. Most differentiated human somatic cells express little or none, which is why they age replicatively.

Loss-of-function mutations in telomerase components and in shelterin cause the short telomere syndromes, a spectrum that includes dyskeratosis congenita, aplastic anaemia, and a substantial fraction of familial idiopathic pulmonary fibrosis.[^armanios2012] These disorders provide the cleanest human evidence that telomere maintenance matters: patients show premature failure of exactly the high-turnover tissues that depend on it, connecting the mechanism directly to [[stem-cell-exhaustion|stem cell exhaustion]]. They do not show that ordinary aging is telomere-driven, since the affected tissues are a subset and the phenotype is not general.

## A weak biomarker

Average leukocyte telomere length is easy to sell as a measure of [[biological-age]] and difficult to defend as one. Inter-individual variation at any given chronological age is large relative to the average annual rate of loss, so a single measurement carries little information about an individual. Measurement methods disagree: quantitative PCR, the cheapest approach and the one behind most consumer tests, has poor within-sample reproducibility compared with terminal restriction fragment analysis or flow-FISH, and results are not comparable across laboratories. Longitudinal studies sometimes report telomere lengthening in individuals over follow-up intervals, which is a strong sign of measurement noise.

> [!caution] Contested
> Consumer telomere tests, sold through the same [[consumer-blood-testing|direct-to-consumer testing]]
> channels as broad blood panels, report a length and an implied "cellular age". No professional body endorses telomere length as a clinical measure of aging outside the diagnosis of telomere biology disorders, and [[epigenetic-clock|methylation-based clocks]] outperform it as mortality predictors in most head-to-head comparisons.

Associations between telomere length and lifestyle factors such as [[exercise-and-aging|physical activity]] or stress have been widely reported and are generally small, inconsistently replicated, and vulnerable to confounding. The field's own retrospective assessment is that early effect sizes were inflated by small samples and selective reporting, a pattern common across candidate [[aging-biomarkers|aging biomarkers]].

## The cancer tradeoff

The large majority of human cancers reactivate telomerase, most often through mutations in the *TERT* promoter that create new transcription factor binding sites; the remainder maintain telomeres through a recombination-based mechanism called alternative lengthening of telomeres. Unlimited division requires solving the end-replication problem, and telomere attrition is therefore a genuine tumour-suppressive barrier in long-lived, large-bodied species.

This makes telomerase therapy for aging structurally awkward. Mendelian randomization studies using genetic variants that raise telomere length find increased risk of several cancers, notably glioma and melanoma, alongside modestly reduced risk of some non-neoplastic conditions.[^tmrc2017] Germline *POT1* mutations that lengthen telomeres cause familial cancer predisposition syndromes. Any intervention that adds telomere reserve across the body would be pushing on a trait for which natural selection appears to have found a compromise.

## Therapeutic attempts

Work in mice has been more encouraging than the cancer argument would suggest, with the caveat that laboratory mice have very long telomeres and constitutive telomerase in many tissues, making them a poor model for the human situation. Systemic delivery of a *Tert* transgene using an [[aav-vectors|adeno-associated viral vector]] to adult mice extended median lifespan and improved several health measures without an observed increase in cancer incidence.[^bdj2012] Similar work in cancer-resistant mouse backgrounds also extended lifespan.

In humans the record is thinner and less reputable. Supplements marketed as telomerase activators, principally cycloastragenol derivatives, rest on small studies with weak endpoints and are sold under the [[dietary-supplements|rules that govern supplements]] rather than as drugs. Self-experimental telomerase [[gene-therapy-for-aging|gene therapy]] outside clinical trials has been publicized by at least one biotechnology executive, with self-reported telomere measurements as the only outcome and no independent verification. Meanwhile the pharmaceutical use of telomere biology has gone the opposite direction: the first approved drug in the area inhibits telomerase to treat a haematological malignancy.

## Open problems

The comparative biology does not fit a simple story. Mice have far longer telomeres than humans and live roughly two years. Some [[negligible-senescence|negligibly senescent]] species maintain telomerase in somatic tissue and control cancer by other means. Whether telomere attrition contributes materially to the trajectory of a normally aging human, as opposed to setting a limit in specific high-turnover compartments, is not resolved, and the question bears directly on how much of the [[sens-research-foundation|damage-repair]] agenda and the [[geroscience-hypothesis|geroscience]] programme should be allocated to it. The one point of agreement is that raising telomerase everywhere is not a plausible route to [[maximum-human-lifespan|extended maximum lifespan]] without a solution to the cancer problem, and no such solution exists.

## See also

- [[cellular-senescence]]
- [[hallmarks-of-aging]]
- [[epigenetic-clock]]
- [[aging-biomarkers]]
- [[gene-therapy-for-aging]]
- [[negligible-senescence]]

## References

[^olovnikov1973]: `paper` Olovnikov, A. M. "A theory of marginotomy: The incomplete copying of template margin in enzymic synthesis of polynucleotides and biological significance of the phenomenon." *Journal of Theoretical Biology*, 1973.
[^bodnar1998]: `paper` Bodnar, A. G. et al. "Extension of life-span by introduction of telomerase into normal human cells." *Science*, 1998. {Human cells in culture; extending a cell line's replicative capacity says nothing about the lifespan of an organism.}
[^armanios2012]: `paper` Armanios, M., Blackburn, E. H. "The telomere syndromes." *Nature Reviews Genetics*, 2012.
[^tmrc2017]: `paper` Telomeres Mendelian Randomization Collaboration. "Association between telomere length and risk of cancer and non-neoplastic diseases: A Mendelian randomization study." *JAMA Oncology*, 2017. {Estimates the effect of lifelong inherited differences in telomere length, which is not the same as the effect of lengthening telomeres in an adult.}
[^bdj2012]: `paper` Bernardes de Jesus, B. et al. "Telomerase gene therapy in adult and old mice delays aging and increases longevity without increasing cancer." *EMBO Molecular Medicine*, 2012. {Mice. The absence of extra cancer was an observation within a lifespan study, not the output of a dedicated carcinogenicity test.}

==============================================================================
ARTICLE: longevity-dividend
TITLE:   The longevity dividend
PORTAL:  Ethics, Society & Governance
URL:     https://futurehumanwiki.com/wiki/longevity-dividend
SOURCE:  https://futurehumanwiki.com/raw/longevity-dividend
==============================================================================

---
title: "The longevity dividend"
slug: "longevity-dividend"
type: "concept"
status: "contested"
horizon: "2030s"
categories: ["society", "longevity"]
tags: ["economics", "aging", "policy", "healthspan", "pensions", "geroscience"]
summary: "The argument that slowing biological aging would produce economic gains larger than curing any individual disease, and the disputes over how that value is calculated."
updated: "2026-07-27"
issues: ["The three-year figure for eliminating cancer mortality has no citation.", "Pension-age indexing in Denmark and the Netherlands is stated without a source."]
---

```infobox
{
  "caption": "Argument in health economics",
  "rows": [
    { "label": "Named in", "value": "2006" },
    { "label": "Original authors", "value": "Olshansky, Perry, Miller, Butler" },
    { "label": "Core claim", "value": "Delay aging, not diseases" },
    { "label": "Valuation method", "value": "Value of statistical life" },
    { "label": "Largest published estimate", "value": "Tens of trillions of dollars" },
    { "label": "Main dispute", "value": "Whether the value is fiscal" }
  ]
}
```

**The longevity dividend** is the argument that a modest slowing of biological aging would yield health and economic benefits exceeding those of eliminating any single disease, and that public research funding should be reallocated accordingly. It was named in 2006 by S. Jay Olshansky, Daniel Perry, Richard Miller, and Robert Butler, who proposed that a seven-year delay in the onset of age-related decline would deliver larger population health gains than curing cancer or heart disease.[^olshansky2006] The argument is the economic face of the [[geroscience-hypothesis]] and the principal policy case advanced by the aging-research field.

## The competing-risks argument

The technical core is simple and not much disputed. Age-related diseases compete: a person spared one dies of another. Modelling the elimination of cancer produces a smaller gain in life expectancy than intuition suggests, because the cohort that survives it remains at high risk of cardiovascular disease, dementia, and frailty. Attacking the shared upstream process — the [[hallmarks-of-aging]] — shifts the whole risk profile rather than removing one branch of it.

The corollary is that the returns to disease-specific research are subject to diminishing returns in a way that returns to geroscience are not. A US modelling study published in 2013 compared a delayed-aging scenario against scenarios in which cancer or heart disease were substantially delayed, and found the aging scenario produced both more years of life and a larger share of those years free of disability, valued at roughly seven trillion dollars in social benefit over fifty years.[^goldman2013] That paper is the origin of the dividend's quantitative form.

> [!key] Why disease-by-disease has a ceiling
> Life-table arithmetic implies that eliminating all cancer mortality would add roughly three years to US life expectancy at birth — a real gain, but far smaller than the disease's salience suggests, because the people saved remain old.

## The large valuations

The most-cited figure comes from a 2021 analysis by Andrew Scott, Martin Ellison, and David Sinclair, which valued improvements in the aging process using the standard economic method for pricing mortality risk.[^scott2021] Their headline results are that a one-year increase in life expectancy achieved by slowing aging is worth on the order of tens of trillions of dollars to the US population, that larger delays are worth proportionally more, and that the value increases as the intervention is applied to successive cohorts because each benefits from the improvements made for the last.

The paper's more interesting finding is comparative. Interventions that extend life without improving health — the Struldbrugg scenario, named for Swift's immortals — are worth much less, and can be worth negative amounts once disability costs are counted. Value depends almost entirely on whether lifespan gains are matched by [[healthspan]] gains, which is the empirical question the field has not answered.

## How the numbers are made

The valuations use the value of a statistical life, a figure derived from what people demonstrably pay to reduce small risks of death — wage premiums for dangerous work, spending on safety equipment. US regulatory agencies use values on the order of ten million dollars per statistical life. Multiplying a per-person willingness to pay by a population produces very large aggregates very quickly.

This is standard practice in regulatory cost-benefit analysis and it is frequently misread. The resulting number is not money that appears in an economy, not tax revenue, and not a budget a government could spend. It is an aggregate of what people would notionally pay for a risk reduction, a measure of welfare rather than of output. Reporting a longevity dividend "worth $38 trillion" alongside figures for national GDP, as coverage routinely does, compares quantities of different kinds.

A separate and more tractable question is the fiscal one: what happens to public budgets. Here the answer depends almost entirely on labour force participation. Extended healthy working life raises output and contributions; extended life without extended work raises pension and care outlays. The dividend argument assumes the former, which is a claim about institutions rather than biology.

## Counterarguments

**Morbidity has expanded, not compressed.** James Fries proposed in 1980 that medical progress would push the onset of disability toward a fixed lifespan limit, compressing illness into a short terminal period.[^fries1980] Demographic evidence over the following decades has mostly not supported it: gains in life expectancy have generally been accompanied by comparable or larger gains in years lived with disease, and the global gap between total life expectancy and healthy life expectancy has stayed on the order of a decade without narrowing much. [[compression-of-morbidity]] examines the evidence. If geroprotective drugs behave like the medical advances that preceded them, they would extend both curves together, which is the scenario the valuations price at close to zero.

**The dependency ratio.** Populations in most high-income and many middle-income countries are aging because fertility has fallen, not because old-age mortality has — the demographic point developed in [[overpopulation-and-longevity]]. Adding years at the end without changing retirement norms increases the ratio of non-working to working people. Several countries, Denmark and the Netherlands among them, have already indexed statutory pension ages to life expectancy, which converts longevity gains into working years automatically and is the policy the dividend argument implies.

**There is no drug to fund.** The dividend is an argument for research, not a description of an available intervention, and it should not be confused with the much stronger claim examined in [[longevity-escape-velocity]]. No pharmaceutical has been shown to slow aging in humans, and the [[glp-1-receptor-agonists]], sometimes offered as a counter-example, treat metabolic disease rather than aging. [[metformin]] has an observational signal and a trial designed to test it that has never been fully funded; [[senolytics]] have produced modest or null results in the human studies reported so far; [[rapamycin]] has strong mouse data and no completed long-term human outcome trial; unregulated practices such as [[heat-and-cold-exposure|sauna and cold-water immersion]] rest on observational cohorts and short-term physiological measures rather than outcome trials. The best-supported geroprotector remains the one described in [[exercise-and-aging]], which requires no research funding at all.

**Distribution.** A dividend accruing mostly to populations that already live longest would widen the existing gap in life expectancy between rich and poor. Nothing in the argument's structure addresses who receives the years, a problem developed in [[access-and-inequality]].

> [!debate] What the estimate is for
> Supporters treat the trillion-dollar figures as evidence that aging research is systematically underfunded relative to its expected return. Critics note that the same method applied to any large mortality reduction produces similar numbers, so the figure demonstrates that mortality is valuable rather than that geroscience will deliver.

## Policy implications

The dividend argument has had two concrete effects. It supplied the rationale for treating aging as a legitimate indication in regulatory terms, the reasoning behind [[metformin|the proposed TAME trial]] designed by [[nir-barzilai]] and colleagues, structured less to demonstrate efficacy than to establish that a regulator would accept an aging-related composite endpoint at all. And it shaped prize competitions and philanthropic funding: the endpoints chosen by [[xprize-healthspan]] are functional restoration in muscle, cognition, and immunity rather than lifespan, precisely because functional years are what the economics values.

It has not yet moved public research budgets appreciably. National aging-research institutes remain funded well below cancer and cardiovascular programmes in most countries, and the disease-specific structure of medical research funding has proved resistant to the competing-risks argument that undermines it.

## The unresolved question

Everything turns on whether an intervention that extends life also extends health, and the field has no accepted way to measure that in advance. [[aging-biomarkers]] describes the search for a surrogate endpoint that regulators would accept; without one, a trial testing the dividend's central premise would need to run for decades on hard outcomes. Until such a trial exists, the economic case rests on an assumption about biology that the biology has not yet supplied.

## See also

- [[geroscience-hypothesis]]
- [[compression-of-morbidity]]
- [[healthspan]]
- [[access-and-inequality]]
- [[overpopulation-and-longevity]]
- [[aging-biomarkers]]
- [[xprize-healthspan]]
- [[longevity-escape-velocity]]

## References

[^olshansky2006]: `paper` Olshansky, S.J., Perry, D., Miller, R.A., Butler, R.N. "In Pursuit of the Longevity Dividend." *The Scientist*, 2006. {A position piece proposing a research policy; the seven-year delay is a stipulated scenario rather than a result from any study.}
[^goldman2013]: `paper` Goldman, D.P. et al. "Substantial Health and Economic Returns from Delayed Aging May Warrant a New Focus for Medical Research." *Health Affairs*, 2013. {A microsimulation of US scenarios, not a trial; the delayed-aging case assumes an intervention that does not exist.}
[^scott2021]: `paper` Scott, A.J., Ellison, M., Sinclair, D.A. "The economic value of targeting aging." *Nature Aging*, 2021. {The trillion-dollar figures are aggregated willingness to pay for reduced mortality risk, not output, revenue or any sum a budget could contain.}
[^fries1980]: `paper` Fries, J.F. "Aging, Natural Death, and the Compression of Morbidity." *New England Journal of Medicine*, 1980.

==============================================================================
ARTICLE: teleportation-problem
TITLE:   The teleportation problem
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/teleportation-problem
SOURCE:  https://futurehumanwiki.com/raw/teleportation-problem
==============================================================================

---
title: "The teleportation problem"
slug: "teleportation-problem"
type: "concept"
status: "speculative"
horizon: "indefinite"
categories: ["minds"]
tags: ["identity", "parfit", "thought experiment", "survival", "uploading", "philosophy of mind"]
summary: "A thought experiment in which a person is scanned, destroyed, and rebuilt elsewhere, used to test whether survival requires physical continuity or only psychological continuity."
updated: "2026-07-27"
issues: ["Four-dimensionalism and non-branching accounts are described without citing a source."]
---

```infobox
{
  "caption": "Thought experiment in philosophy of mind",
  "rows": [
    { "label": "Canonical version", "value": "Parfit, 1984" },
    { "label": "Structure", "value": "Scan, destroy, rebuild" },
    { "label": "Key variant", "value": "The branch-line case" },
    { "label": "Divides", "value": "Psychological vs. physical continuity" },
    { "label": "Applies to", "value": "Uploading, preservation, gradual replacement" },
    { "label": "Status", "value": "No consensus resolution" }
  ]
}
```

**The teleportation problem** is a thought experiment used to isolate what survival requires. A
machine records the exact state of a person's body and brain, destroys the original, transmits the
information, and assembles a molecule-for-molecule duplicate at the destination. The duplicate has
every memory, every disposition, and the vivid sense of having stepped in one door and out another.
The question is whether the person travelled or died and was replaced. It is the sharpest available
test of theories of [[personal-identity-and-continuity]], and it is the exact structure of
destructive [[mind-uploading]].

## The thought experiment

Derek Parfit gave the standard presentation.[^parfit1984] In the simple case the original is
destroyed as the scan completes, so there is only ever one claimant. Most people who imagine
themselves stepping into the booth report reluctance, but find it hard to say what is lost: the
duplicate is physically identical, psychologically continuous, and would insist it is the same
person.

Parfit then introduces the *branch-line case*. The scanner malfunctions: it copies without
destroying. The original survives on Earth with fatal cardiac damage and a few days to live, while
the replica walks around on Mars. Now the original can talk to the replica by radio. Very few people
are willing to say the original should be indifferent to their imminent death because the replica
exists — yet the replica's relation to the original is exactly the relation that, in the simple
case, was supposed to constitute survival. Nothing about the replica changed. Only the presence of a
competitor did.

> [!key] Why the variant does the work
> If identity depended only on the intrinsic relation between two people, the branch-line replica
> would be the original whether or not the original survives. That it apparently is not shows the
> criterion has an extrinsic component — whether anyone else has an equal claim — which is a strange
> feature for a fact about a person to have.

## The intuition split

Responses divide roughly into three camps, and the division is stable across audiences.

*Survivalists* hold that the duplicate is the person. What makes someone the same person is the
pattern of psychological connections, and the pattern is preserved. On this reading, ordinary
survival through dreamless sleep or general anaesthesia already involves an interruption in
experience, and nobody treats waking up as being replaced.

*Physical-continuity theorists* hold that the person died. Experience is a process running in a
particular physical system; when the system is destroyed the process ends, and starting an identical
process elsewhere begins a new life. On this view anaesthesia is disanalogous because the physical
substrate continues to run, metabolizing and maintaining itself, throughout.

*Deflationists*, following Parfit, hold that the question has no further answer. There is a
psychological relation that holds and a physical relation that does not, and once both are specified
there is nothing left to discover. Asking whether the person "really" survived is like asking
whether a rebuilt club is really the same club.

Experimental work confirms the split is genuine rather than an artefact of philosophical training.
Studies of ordinary intuitions find that people give different verdicts depending on framing, on
whether the scenario is described in first or third person, and on which mental features are said to
be preserved.[^nichols2010] Judgements about whether someone remains the same person weight moral
character heavily, more than autobiographical memory.[^strohminger2014]

## Attempts at resolution

**Identity is not what matters.** Parfit's own move is to grant that the duplicate is not
straightforwardly identical to the original while denying that identity was ever what mattered.
What matters is psychological connectedness and continuity, and these are preserved. The branch-line
intuition, on this account, is a failure of imagination rather than a discovery.

**Closest continuer.** Robert Nozick's theory makes the duplicate the original when no better
candidate exists, and not the original when the biological body survives.[^nozick1981] This delivers the
intuitive verdict in both cases at the cost of making survival depend on facts about other people —
so whether one survives can be changed by an event elsewhere, after the fact.

**Four-dimensionalism.** Treating persons as extended in time, with momentary stages as parts, lets
one say that two person-worms overlapped before the scan. This resolves the logic of branching but
does not tell anyone stepping into the booth what to expect.

**Non-branching clauses.** Some accounts define survival as psychological continuity *provided it
has not taken a branching form*. Critics regard this as writing the intuition into the definition
rather than explaining it.

## Gradual replacement as a dissolution

The most influential practical response is to change the case. Instead of scanning and rebuilding,
replace the brain's components one at a time, each substitute reproducing the function of what it
displaces, while the person remains awake. There is never a moment of destruction, never two
claimants, and never a gap in the running process — so all three camps can agree that the person
survives, provided each replacement genuinely preserves function. Weaker versions of the same move
underlie incremental framings of [[human-ai-merger]] and of the [[posthuman]] condition: a person
augmented in small steps never confronts a moment of transfer.

This is why gradual routes dominate serious discussion of substrate change, and it is the identity
half of the argument for [[substrate-independence]]. David Chalmers, who defends the functionalist
premise, is notably more confident about gradual uploading than about the destructive kind.[^chalmers2022]
The catch is engineering: the required devices would have to reproduce a single neuron's
computation, including dendritic nonlinearity and neuromodulatory response, inside living tissue —
far beyond anything achieved in [[neuroprosthetics]] or contemplated under [[medical-nanorobots]].
The philosophically comfortable route is the technically hopeless one.

## What a real teleporter would do

Physics offers no duplication machine. Quantum teleportation, proposed in 1993 and demonstrated in
laboratories since, transfers an unknown quantum state from one system to another using entanglement
and a classical channel.[^bennett1993] It necessarily destroys the state at the origin, because the
no-cloning theorem forbids copying an unknown quantum state.[^wootters1982] It is closer to Parfit's
simple case than
to the branch line, and it says nothing about identity, since the relevant question is whether brains
require quantum-state fidelity at all. Most neuroscientists think they do not: synaptic transmission
is a thermally noisy chemical process, and proposals that quantum coherence is functionally
important remain a minority position.

## Bearing on real technologies

The problem is not idle. Destructive scanning for [[whole-brain-emulation]] has the teleporter's
exact structure, minus the reassembly of a body. [[brain-preservation]] and [[cryonics]] raise a
milder version: the process is not obviously destructive, but it does stop, and whether restarting
counts as continuation depends on the same disagreement. The weaker products sold as
[[digital-immortality]] do not raise the problem at all, since a model trained on someone's messages
has no continuity claim of any kind — a distinction the marketing tends to blur.

Nothing in [[connectomics]], [[neural-correlates-of-consciousness]] research, or
[[machine-consciousness]] work settles it, and no one has described a result that could. The
physical facts in the teleporter case are stipulated and complete. What remains is a decision about which description of those facts to accept,
and the practical question of who gets to make that decision when a real procedure is on offer.

## See also

- [[personal-identity-and-continuity]]
- [[mind-uploading]]
- [[substrate-independence]]
- [[whole-brain-emulation]]
- [[brain-preservation]]
- [[cryonics]]
- [[digital-immortality]]
- [[machine-consciousness]]

## References

[^parfit1984]: `book` Parfit, D. *Reasons and Persons*. Oxford University Press, 1984.

[^nichols2010]: `paper` Nichols, S. and Bruno, M. "Intuitions about personal identity: An empirical study." *Philosophical Psychology*, 2010. {A vignette study of lay intuitions: it records what non-philosophers say about the cases, not which theory of identity is correct.}

[^strohminger2014]: `paper` Strohminger, N. and Nichols, S. "The essential moral self." *Cognition*, 2014.

[^chalmers2022]: `book` Chalmers, D. J. *Reality+: Virtual Worlds and the Problems of Philosophy*. W. W. Norton, 2022. {A general-audience philosophy book; the uploading discussion is one strand of a broader argument about virtual worlds.}

[^wootters1982]: `paper` Wootters, W. K. and Zurek, W. H. "A single quantum cannot be cloned." *Nature*, 1982.

[^bennett1993]: `paper` Bennett, C. H. et al. "Teleporting an unknown quantum state via dual classical and Einstein-Podolsky-Rosen channels." *Physical Review Letters*, 1993. {A theoretical proposal; experimental demonstrations of the protocol came several years afterwards.}

[^nozick1981]: `book` Nozick, R. *Philosophical Explanations*. Harvard University Press, 1981.

==============================================================================
ARTICLE: tissue-engineering
TITLE:   Tissue engineering
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/tissue-engineering
SOURCE:  https://futurehumanwiki.com/raw/tissue-engineering
==============================================================================

---
title: "Tissue engineering"
slug: "tissue-engineering"
type: "technology"
status: "established"
horizon: "present"
trl: 7
categories: ["bodies"]
tags: ["regenerative medicine", "scaffolds", "biomaterials", "stem cells", "transplantation", "cartilage"]
summary: "The construction of living tissue from cells, scaffolds and signalling molecules, a field with approved products for skin and cartilage but no solid organ after three decades."
updated: "2026-07-27"
humanEvidence: "Approved products treat burns, chronic wounds, focal knee cartilage defects, corneal burns and congenital athymia in patients; the only organ-like construct with years of human follow-up is a small bladder series."
access: "Skin substitutes, chondrocyte implantation, limbal grafts and cultured thymus tissue are approved and delivered through specialist centres; no engineered solid organ is available at any price."
reversibility: "difficult"
issues: ["Approval dates for the cartilage, corneal and thymus products are stated without sources."]
---

```infobox
{
  "caption": "Biomedical engineering field",
  "rows": [
    { "label": "Named", "value": "1993" },
    { "label": "Named by", "value": "Robert Langer and Joseph Vacanti" },
    { "label": "Core triad", "value": "Cells, scaffold, signals" },
    { "label": "Approved product classes", "value": "Skin, cartilage, cornea, thymus" },
    { "label": "Solid organs achieved", "value": "None" },
    { "label": "Main obstacle", "value": "Vascularization and cell supply" },
    { "label": "Status", "value": "Clinical for simple tissues" }
  ]
}
```

**Tissue engineering** is the construction of functional living tissue by combining cells, a supporting scaffold and biochemical or mechanical signals, either in a bioreactor before implantation or inside the body afterwards. The term was fixed by a 1993 *Science* article by Robert Langer and Joseph Vacanti that set out the field's programme and its founding triad.[^langer1993] Three decades later the discipline has delivered approved treatments for skin, cartilage, cornea and thymus, and none for a kidney, liver or heart. The distance between those two lists is the field's defining fact.

## The triad

The classical formulation has three components.

**Cells** provide the biological function. They may be autologous — taken from the patient, expanded in culture, and returned, which avoids rejection but takes weeks and cannot be stockpiled — or allogeneic, which allows an off-the-shelf product but usually requires immunosuppression or immune-privileged siting. Increasingly they are derived from [[induced-pluripotent-stem-cells]], which decouples supply from donor tissue at the cost of differentiation and safety work.

**Scaffold** provides shape, mechanical support and a surface for attachment. Early scaffolds were synthetic polyesters already approved as resorbable sutures: polyglycolic acid, polylactic acid, their copolymers, and polycaprolactone. These degrade by hydrolysis on a tunable timescale, ideally matched to the rate at which cells lay down their own matrix. Natural scaffolds — collagen, fibrin, silk, and stripped donor tissue as described in [[decellularized-scaffolds]] — trade manufacturing control for better biological cues.

**Signals** are the growth factors, matrix stiffness, oxygen tension and mechanical loading that tell cells what to become. A scaffold seeded with chondrocytes and left static produces poor cartilage; the same construct under cyclic compression in a bioreactor produces better cartilage. Mechanical conditioning turned out to be as important as any molecule.

## Origins

```timeline
[
  { "year": "1975", "title": "Keratinocyte culture", "text": "Howard Green and James Rheinwald establish serial cultivation of human epidermal keratinocytes, making cultured skin grafts possible and giving the field its first clinical foothold." },
  { "year": "1988", "title": "The term enters use", "text": "A National Science Foundation workshop popularises 'tissue engineering' as a name for the emerging combination of biomaterials and cell biology." },
  { "year": "1993", "title": "The programme stated", "text": "Langer and Vacanti's Science article sets out cells-plus-scaffold-plus-signals as a general strategy for replacing lost tissue, and is taken as the field's founding document." },
  { "year": "1997", "title": "The ear mouse", "text": "Charles Vacanti's group grows human-ear-shaped cartilage on a polymer scaffold implanted under the skin of an athymic mouse. The image travels far beyond the science it represents." },
  { "year": "1997–2001", "title": "First approvals", "text": "Autologous chondrocyte implantation and engineered skin substitutes reach the market, followed within a few years by the commercial collapse of several of their manufacturers." },
  { "year": "2006", "title": "Engineered bladders", "text": "Anthony Atala's group reports autologous bladder constructs implanted in seven young patients with spina bifida, the first engineered organ-like structure followed for years in humans." }
]
```

The ear mouse deserves separate mention because it shaped public expectation more than any result in the field. The construct was cartilage — avascular, mechanically simple, immunologically quiet — grown in an immunodeficient animal that could not reject it. It demonstrated that a polymer scaffold could hold a complex shape while cells filled it in. It demonstrated nothing about vascularized organs, and it was read as though it had.

## What reached the clinic

Engineered tissues in routine or approved use share a profile: thin, avascular or thinly vascularized, and small enough that diffusion suffices.

Cultured epidermal autografts have been used for massive burns since the 1980s. Bilayered skin substitutes combining allogeneic fibroblasts and keratinocytes on a collagen or polymer matrix are approved for chronic wounds; they act largely as living dressings that release growth factors rather than as permanent grafts. Autologous chondrocyte implantation, first approved in the United States in 1997 and superseded by a matrix-associated version in 2016, treats focal cartilage defects in the knee. Cultured limbal stem cell grafts for corneal burns received European approval in 2015. Allogeneic cultured thymus tissue for children born without a thymus was approved in the United States in 2021, and is one of the few products that reconstitutes an organ's function rather than patching a surface.

Atala's bladder work remains the most ambitious result with long human follow-up: autologous urothelial and muscle cells seeded on a biodegradable scaffold and implanted, with functional improvement sustained over years in a small cohort.[^atala2006] The attempt to commercialize it failed, as did several contemporaneous ventures. The commercial failures were not primarily scientific. Autologous products are closer to a service than a manufactured good, with a batch size of one, a cold chain, and a reimbursement code that often does not exist. This is the gap between laboratory validation and operational use that [[technology-readiness-level]] describes, and it determines pricing and therefore the questions raised in [[access-and-inequality]].

> [!key] Why the simple tissues worked
> Every clinically successful engineered tissue is one where oxygen can diffuse from surrounding host tissue to every cell. That single physical constraint, not regulatory caution or funding, separates the products that exist from the ones that do not.

## Why organs did not follow

Four obstacles have proved durable.

**Vascularization.** A construct thicker than roughly a millimetre needs its own blood supply. Building a capillary bed remains unsolved, whether by printing it, as in [[organ-bioprinting]], by reusing a donor organ's vascular tree, or by relying on host angiogenesis, which grows in at roughly a fraction of a millimetre per day — too slow for a thick construct.

**Cell number and phenotype.** A liver contains on the order of a hundred billion cells, and primary hepatocytes dedifferentiate within days outside the body. Producing enough cells of the right type, in the right state, remains a manufacturing problem as much as a biological one.

**Maturation.** Stem-cell-derived tissue is fetal-like. Cardiomyocytes contract weakly, neurons are electrically immature, and no reliable protocol drives them to adult phenotype on a useful timescale. The same ceiling limits [[organoids]]. Cells taken from an elderly recipient bring a second problem, the functional decline catalogued under [[stem-cell-exhaustion]], which is part of why several groups have proposed combining autologous cell manufacture with [[epigenetic-reprogramming]] to restore proliferative capacity before use.

**Integration.** An implanted construct must connect to host vasculature, innervation, and in some cases a duct or lumen, and must survive the host's foreign-body response. Work by Jennifer Elisseeff's group showed, in a mouse muscle-injury model, that biomaterial scaffolds elicit a type 2 immune response involving T cells that determines whether the outcome is regeneration or fibrosis, reframing the immune system as an active participant in the repair rather than an obstacle to be suppressed. The same dependence has not been demonstrated in humans.[^sadtler2016]

## Current directions

The field's centre of gravity has shifted in three ways. First, away from the pre-fabricated construct and toward *in vivo* tissue engineering, where an acellular material recruits the patient's own cells and the body serves as the bioreactor; most commercially successful regenerative products now work this way. Second, toward immunomodulatory design, in which the scaffold's job is to steer macrophage and T-cell behaviour rather than merely to provide shape. Third, toward biofabrication methods that give spatial control, including printing, moulding of tissue building blocks, and assembly of organoid units into larger structures.

Two adjacent products show how the boundary of the field has moved. Engineered cardiac patches are being developed as an adjunct to, not a replacement for, the mechanical circulatory support described in [[artificial-heart]]. Cultured red cells, discussed in [[artificial-blood]], are an engineered tissue in everything but name, and face a manufacturing problem of scale rather than architecture.

Engineered tissue has also found a market that does not require implantation at all. Liver, cardiac and tumour tissue built on chips or in multiwell formats is used for toxicity screening, and demand grew as regulators moved away from mandatory animal testing for some drug classes — the same shift that names the computational models described in [[human-digital-twins]] among the acceptable alternatives.

## Outlook

Tissue engineering is a mature field with a modest clinical footprint and an unmet founding promise. Its practitioners generally no longer forecast a printed kidney; they forecast incremental gains in islet constructs, vascular grafts, nerve conduits, corneal tissue and cardiac patches, each of which is a real clinical need served by a few cubic centimetres of tissue. Whether the discipline ever produces one of the [[lab-grown-organs]] it was founded to deliver depends on the vascular problem, and the strongest current argument is that any answer would come from combining a biologically constructed capillary bed with an engineered large-vessel framework rather than from fabricating either alone. The alternative is to stop building and start inducing, which is why the biology of animals that rebuild structures on their own, covered in [[limb-regeneration]], has drawn renewed attention from engineers. Meanwhile the [[organ-shortage]] is being addressed by [[xenotransplantation]], machine perfusion and donation policy, all of which were considered less likely than engineered organs when the field was named. The same ordering holds in reproductive medicine, where [[uterus-transplantation]] has produced live births while recellularized uterine scaffolds remain in animal work.

## See also

- [[organ-bioprinting]]
- [[decellularized-scaffolds]]
- [[organoids]]
- [[lab-grown-organs]]
- [[induced-pluripotent-stem-cells]]
- [[limb-regeneration]]
- [[organ-shortage]]
- [[stem-cell-exhaustion]]

## References

[^langer1993]: `paper` Langer, R. and Vacanti, J. P. "Tissue Engineering." *Science*, 1993.
[^atala2006]: `paper` Atala, A. et al. "Tissue-engineered autologous bladders for patients needing cystoplasty." *The Lancet*, 2006. {A small series of young patients, and the construct augmented an existing bladder rather than replacing a whole organ.}
[^sadtler2016]: `paper` Sadtler, K. et al. "Developing a pro-regenerative biomaterial scaffold microenvironment requires T helper 2 cells." *Science*, 2016. {The immune requirement was established in a mouse muscle-injury model; no human study has shown the same dependence.}

==============================================================================
ARTICLE: transhumanism
TITLE:   Transhumanism
PORTAL:  People & Movements
URL:     https://futurehumanwiki.com/wiki/transhumanism
SOURCE:  https://futurehumanwiki.com/raw/transhumanism
==============================================================================

---
title: "Transhumanism"
slug: "transhumanism"
type: "concept"
status: "established"
horizon: "present"
categories: ["people", "foundations"]
tags: ["transhumanism", "movements", "history", "enhancement", "ideology", "bioethics"]
summary: "The intellectual and cultural movement holding that human beings can and should use technology to overcome the biological limits of the human condition."
updated: "2026-07-27"
issues: ["Membership figures are characterised qualitatively and carry no source."]
---

```infobox
{
  "caption": "Intellectual and cultural movement",
  "rows": [
    { "label": "Term popularised", "value": "1957, Julian Huxley" },
    { "label": "Modern movement", "value": "From c. 1990" },
    { "label": "Key organisations", "value": "Extropy Institute, Humanity+", "link": "/wiki/humanity-plus" },
    { "label": "Core commitments", "value": "Enhancement, longevity, autonomy" },
    { "label": "Principal critics", "value": "Fukuyama, Habermas, Kass" },
    { "label": "Status", "value": "Diffuse; institutionally weak, culturally influential" }
  ]
}
```

**Transhumanism** is the position that the human condition is not fixed, that its biological limits are legitimate targets of technological intervention, and that people should be free to use such interventions on themselves. It is a philosophical stance rather than a research programme, and its adherents disagree with each other about nearly everything except that premise. Its influence is easiest to see not in its own organisations, which have always been small, but in the assumptions that circulate in longevity biotech, neurotechnology and artificial-intelligence circles.

## Precursors

The word is older than the movement and older than its twentieth-century users. Dante's *Paradiso* coins *trasumanar* for the passage beyond the human, and nineteenth-century English translators rendered it as "transhumanise". Modern usage does not descend from that line, and the coincidence has produced a good deal of retrospective genealogy.

Julian Huxley used the word in a 1957 essay to name the idea of humanity consciously fulfilling its own potential, "man remaining man, but transcending himself".[^huxley1957] Huxley was a biologist and a eugenicist, and the movement's later critics have made much of that lineage; its defenders point out that Huxley's meaning was cultural and educational rather than genetic.

The substantive antecedents are older and mostly literary or speculative. J. B. S. Haldane's *Daedalus, or Science and the Future* (1923) predicted gestation outside the body and argued that biological invention would prove more disruptive than physical invention. J. D. Bernal's *The World, the Flesh and the Devil* (1929) sketched surgical replacement of the body, colonisation of space, and eventually the dissolution of individual minds into networked ones.[^bernal1929] Nikolai Fyodorov's Russian cosmism, half a century earlier, had made resurrection of the dead a religious duty of science. Robert Ettinger's *The Prospect of Immortality* (1964) founded [[cryonics]] and, in *Man into Superman* (1972), extended the argument to bodily redesign.

```timeline
[
  { "year": "1923", "title": "Daedalus", "text": "J. B. S. Haldane argues that biological invention will outrun physical invention and predicts gestation outside the body." },
  { "year": "1957", "title": "Huxley's essay", "text": "Julian Huxley publishes 'Transhumanism', giving the word its modern currency without founding a movement." },
  { "year": "1964", "title": "The Prospect of Immortality", "text": "Robert Ettinger's book launches cryonics and supplies the movement with its first practical undertaking." },
  { "year": "1989", "title": "Are You a Transhuman?", "text": "FM-2030 publishes a book treating the transhuman as a transitional being and offers a questionnaire for measuring how far along one is." },
  { "year": "1988–1990", "title": "The Extropians", "text": "Max More and colleagues launch Extropy magazine and publish the Principles of Extropy, giving the movement its first organised form." },
  { "year": "1998", "title": "World Transhumanist Association", "text": "Nick Bostrom and David Pearce found the WTA and draft the Transhumanist Declaration, aiming for academic respectability rather than Californian libertarianism." },
  { "year": "2004", "title": "The dangerous idea", "text": "Francis Fukuyama names transhumanism the world's most dangerous idea in Foreign Policy, marking its arrival as a target of mainstream political theory." },
  { "year": "2008", "title": "Humanity+", "text": "The WTA rebrands as Humanity+; the movement's centre of gravity shifts toward well-funded technology firms and away from membership organisations." }
]
```

## Formation of the modern movement

[[fm-2030]], born Fereidoun M. Esfandiary, taught at the New School in New York and used "transhuman" to mean a transitional human being whose values and habits already anticipate a post-biological future. His 1989 book *Are You a Transhuman?* is the first popular use of the term in something like its current sense.

The organised movement began in California around 1988, when [[max-more]] and Tom Bell founded *Extropy* magazine; the Extropy Institute followed in 1992 and closed in 2006. More's Principles of Extropy — perpetual progress, self-transformation, practical optimism, intelligent technology, open society, self-direction, rational thinking — supplied the movement's first explicit creed and its libertarian inflection.[^more1990] The extropians mailing list served through the 1990s as a common seedbed for cryonics, molecular nanotechnology, mind uploading and early artificial-intelligence risk arguments; the details are in [[extropianism]].

In 1998 [[nick-bostrom]] and the philosopher David Pearce founded the World Transhumanist Association, deliberately positioned as a broader and more academically presentable body. Its Transhumanist Declaration, revised in 2009, is the nearest thing the movement has to a consensus text: it asserts the possibility of redesigning the human condition, insists on serious study of risks including [[existential-risk|existential ones]], advocates wide access rather than access for the wealthy, and defends personal choice over enhancement. The WTA became [[humanity-plus]] in 2008. [[natasha-vita-more]] contributed the movement's principal artistic strand, including the Primo Posthuman design study and a Transhumanist Arts Statement.

The movement has always been small in headcount and disproportionately present in adjacent institutions. Its participants founded or staffed cryonics providers, prize bodies such as the [[methuselah-foundation]], and research organisations including the Future of Humanity Institute at Oxford and the Machine Intelligence Research Institute in California. That pattern — few members, many institutions — accounts for both its reach and the difficulty of saying how many transhumanists there are. Membership figures for the organisations that kept them ran in the thousands rather than the hundreds of thousands, and no survey has established how many people hold the position without joining anything.

## Core commitments

Four claims recur across otherwise incompatible transhumanist positions.

**The human condition is contingent.** Ageing, cognitive limits, mood set-points and the biological substrate of the mind are historical accidents rather than necessary features, and can in principle be engineered. This underwrites interest in [[hallmarks-of-aging|the biology of ageing]] and [[longevity-escape-velocity]].

**Enhancement is legitimate.** The therapy–enhancement line does not mark a moral boundary; see [[human-enhancement]]. What matters is whether an intervention is safe, chosen, and good for the person.

**Morphological freedom.** Individuals have a strong claim to modify their own bodies and minds, and a correspondingly strong claim not to be modified. [[morphological-freedom]] traces the argument.

**Substrate independence, at least as a live possibility.** Many transhumanists hold that mind is a matter of organisation rather than material, which makes [[whole-brain-emulation]] and [[mind-uploading]] coherent projects rather than category errors. This is a philosophical commitment, not an empirical finding; [[substrate-independence]] sets out the objections.

Bostrom's own summary, in his history of the movement, is that transhumanism is best understood as an extension of Enlightenment humanism — the same commitment to reason, science and individual liberty, applied to human biology rather than only to institutions.[^bostrom2005]

## Internal disagreements

The movement's public image as a single ideology is misleading.

The oldest split is political. Extropian transhumanism is libertarian, sceptical of regulation, and confident that markets will diffuse enhancement. Democratic transhumanism, argued by James Hughes in *Citizen Cyborg*, holds that without redistribution and public provision, enhancement will entrench class advantage, and that the movement's libertarian wing has been its own worst enemy on this point.[^hughes2004] The dispute is the internal version of [[access-and-inequality]].

The second split concerns priorities. Longevity-focused transhumanists treat ageing as the central problem and view artificial intelligence as instrumental; see [[aubrey-de-grey]]. Singularitarians treat machine intelligence as the dominant variable, on the grounds that everything else follows from it; see [[singularitarianism]]. The two camps have coexisted uneasily since the 1990s.

The third is about risk. One current, associated with Bostrom's later work and with [[differential-technological-development]], holds that some technologies should be deliberately slowed relative to others. The opposing current, most visible in [[effective-accelerationism]], treats deceleration as itself the primary danger. Both claim the same intellectual ancestry.

A fourth concerns suffering. David Pearce's abolitionist position, set out in *The Hedonistic Imperative* (1995), holds that the elimination of involuntary suffering, not longevity or intelligence, is the movement's proper goal, and that gradients of well-being should replace the pain–pleasure axis entirely. It depends on control of hedonic set-points that no present technique approaches, and it divides the movement over whether the object of enhancement is capability or well-being.

> [!debate] Is transhumanism a religion
> Critics from both religious and secular positions have described transhumanism as a technological eschatology: resurrection by cryopreservation, transcendence by uploading, salvation by superintelligence. Transhumanists generally reply that their claims are empirical and falsifiable in a way that religious ones are not. The comparison is nonetheless doing real work, since the movement's central promises are unfalsifiable on any near timescale.

## Criticism

Francis Fukuyama's 2004 essay naming transhumanism "the world's most dangerous idea" argued that liberal democracy rests on a shared human essence, and that a society whose members had different capacities by design could not sustain equal rights.[^fukuyama2004] Habermas made a parallel argument about the asymmetry between a designer and a designed person. [[leon-kass]] offered the deepest version of the objection: that finitude is a condition of meaning rather than an obstacle to it, and that a project organised around escaping death misunderstands what makes a life good. These positions are developed in [[bioconservatism]].

A second line of criticism comes from science rather than philosophy. Most of the movement's flagship expectations have slipped badly. Molecular assemblers of the kind described in [[molecular-assembler]] were not built; cryopreservation has produced no revival of any complex organism; emulation remains many orders of magnitude from feasibility on both scanning and compute; and clinical [[gene-therapy-for-aging|genetic rejuvenation]] has not been demonstrated in humans. Transhumanist forecasting has a poor calibration record, which its more careful members acknowledge.

A third line, from science and technology studies and from critical theory, holds that transhumanism universalises a narrow set of preferences — individual autonomy, cognitive performance, control over the body — and treats them as human nature. Critics in disability studies argue that the movement's framing of impairment as a defect to be engineered away is precisely what the social model of disability rejects; see [[disability-rights-and-enhancement]]. The critical-posthumanist tradition, distinct from transhumanism despite the overlapping vocabulary, rejects the liberal individual subject that transhumanism takes for granted; [[posthuman]] separates the two.

The most recent critique treats transhumanism as one component of a larger ideological bundle. Timnit Gebru and Émile Torres coined the acronym TESCREAL to link transhumanism with extropianism, singularitarianism, cosmism, rationalism, effective altruism and longtermism, arguing that the cluster carries eugenic assumptions and that its influence on artificial-intelligence development deserves scrutiny.[^gebru2024] Transhumanists dispute both the coherence of the bundle and the eugenics attribution; the debate is unresolved and has become unusually acrimonious.

## Influence and current standing

As a membership movement, transhumanism is weaker than it was in 2005. Extropy Institute closed, Humanity+ has been intermittently active, and the Future of Humanity Institute, which housed much of the movement's most rigorous work, shut in 2024.

As a set of assumptions, it is stronger. Companies pursuing [[epigenetic-reprogramming]], [[brain-computer-interface|neural interfaces]] and machine intelligence are funded by people who read the movement's literature, and several of its founding ideas — that ageing is a treatable condition, that the brain is an information-processing system, that [[artificial-general-intelligence]] is a near-term engineering target — are now mainstream positions in their respective industries whether or not anyone using them accepts the label.

A useful test of influence is vocabulary. Terms coined or popularised inside the movement — existential risk, superintelligence, escape velocity, morphological freedom, substrate independence — now appear in policy documents, corporate research agendas and legislative hearings, generally without attribution and often without the arguments that originally supported them. That is what makes the movement's current position hard to assess. The words travel; the reasoning that constrained them does not.

The open question is whether that diffusion counts as success. A movement whose premises have been absorbed while its organisations dissolved has won the argument in one sense and lost the ability to shape what follows in another, at exactly the point where the technologies it anticipated are becoming decisions rather than predictions.

## See also

- [[extropianism]]
- [[posthuman]]
- [[human-enhancement]]
- [[bioconservatism]]
- [[singularitarianism]]
- [[morphological-freedom]]
- [[humanity-plus]]
- [[future-of-humanity]]

## References

[^huxley1957]: `book` Huxley, J. "Transhumanism." In *New Bottles for New Wine*. Chatto & Windus, 1957.
[^bernal1929]: `book` Bernal, J. D. *The World, the Flesh and the Devil: An Enquiry into the Future of the Three Enemies of the Rational Soul*. Kegan Paul, 1929.
[^more1990]: `paper` More, M. "Transhumanism: Toward a Futurist Philosophy." *Extropy*, 1990. {Published in the magazine More co-founded; a statement of the extropian position rather than an independent account of it.}
[^bostrom2005]: `paper` Bostrom, N. "A History of Transhumanist Thought." *Journal of Evolution and Technology*, 2005. {A history written by a co-founder of the World Transhumanist Association, in a journal aligned with the movement.}
[^hughes2004]: `book` Hughes, J. *Citizen Cyborg: Why Democratic Societies Must Respond to the Redesigned Human of the Future*. Westview Press, 2004.
[^fukuyama2004]: `paper` Fukuyama, F. "Transhumanism." In "The World's Most Dangerous Ideas," *Foreign Policy*, 2004.
[^gebru2024]: `paper` Gebru, T. and Torres, É. P. "The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence." *First Monday*, 2024. {Written by critics of the movement; it argues a thesis about an ideological cluster rather than surveying who holds which view.}

==============================================================================
ARTICLE: utah-array
TITLE:   Utah array
PORTAL:  Cybernetics & Neurotech
URL:     https://futurehumanwiki.com/wiki/utah-array
SOURCE:  https://futurehumanwiki.com/raw/utah-array
==============================================================================

---
title: "Utah array"
slug: "utah-array"
type: "technology"
status: "established"
horizon: "present"
trl: 7
categories: ["cybernetics"]
tags: ["bci", "microelectrodes", "neural recording", "braingate", "implants", "neurotechnology"]
summary: "A silicon microelectrode array of 100 needles on a 10-by-10 grid, the standard sensor for chronic single-neuron recording in human brain–computer interface research."
updated: "2026-07-27"
humanEvidence: "People with paralysis have used implanted arrays under investigational device exemptions to control cursors, robotic arms and speech decoders; some implants recorded usable signals beyond a thousand days."
access: "Not a therapy: the US clearance covers recording for up to thirty days, and every long-term implant is placed in a research study under an investigational device exemption."
reversibility: "difficult"
issues: ["The 30-day US clearance appears in the infobox and prose without a source."]
---

```infobox
{
  "caption": "Intracortical microelectrode array",
  "rows": [
    { "label": "Developed at", "value": "University of Utah" },
    { "label": "First described", "value": "1991" },
    { "label": "Principal developer", "value": "Richard Normann" },
    { "label": "Layout", "value": "10 × 10 needles, 400 µm pitch" },
    { "label": "Typical channels", "value": "96" },
    { "label": "Manufacturer", "value": "Blackrock Neurotech" },
    { "label": "Regulatory status", "value": "Cleared for ≤30-day use; chronic use investigational" }
  ]
}
```

**Utah array** is a silicon microelectrode array consisting of a grid of sharpened needles etched
from a single block of silicon, inserted into the cortex so that each tip records the extracellular
voltage near a small number of neurons. Introduced at the University of Utah around 1990, it became
and remains the standard sensor for chronic single-neuron recording in humans, and nearly every
major human [[brain-computer-interface]] result involving individual action potentials has used
one.

Its persistence is notable given how little it has changed. A device designed before the field had
a single human participant still sets the performance benchmark that newer, higher-channel-count
designs are measured against.

```keyfacts
[
  { "value": "1991", "label": "First manufacturing description", "note": "Campbell et al., IEEE Transactions on Biomedical Engineering" },
  { "value": "96", "label": "Channels typically wired", "note": "of 100 electrode sites on the grid" },
  { "value": "~4 mm", "label": "Footprint edge length", "note": "shanks usually 1.0–1.5 mm long" }
]
```

```figure
{"key": "utah-array-patent", "caption": "The array as filed. Each shank records from a small population of neurons; the tissue's response to all of them is the limiting problem."}
```

## How it works

The array is machined from a monolithic block of doped silicon. Dicing saws cut a grid of columns,
which are then acid-etched into tapered needles on a 400-micrometre pitch, giving a device roughly
four millimetres on a side with 100 electrode sites. The shanks, typically one to one and a half
millimetres long, reach the cortical layers where pyramidal-cell somata are dense. The whole
structure except the tips is insulated with parylene-C; the exposed tips are coated with platinum or
iridium oxide to lower impedance.[^campbell1991]

Insertion is the tricky part. Pressed slowly, the array dimples the cortical surface rather than
penetrating it, damaging tissue and pial vessels. A pneumatic inserter fires the array in at several
metres per second, which produces cleaner penetration.[^maynard1997] Wires from each shank run
through a bundle to a connector; in the standard research configuration this is a percutaneous
titanium pedestal fixed to the skull, through which amplifiers are plugged in during sessions.

What the array measures is the extracellular field near each tip. Filtered above roughly 250 Hz and
thresholded, this yields action potentials from one to a few nearby neurons per channel; filtered
below about 300 Hz, it yields local field potentials reflecting activity over a larger volume. Both
feed the decoders described under [[neural-decoding]].

## Development history

Richard Normann's group designed the array in the late 1980s with a cortical visual prosthesis in
mind — a grid of stimulating electrodes in visual cortex intended to evoke a pattern of phosphenes,
the same goal that continues to motivate work described under [[retinal-implant]]. The
manufacturing process was published in 1991 and the recording characterization followed through the
1990s.

Commercialization passed through several hands: Bionic Technologies, then Cyberkinetics
Neurotechnology Systems, then Blackrock Microsystems, now Blackrock Neurotech. Cyberkinetics ran
the BrainGate pilot trial, in which Matthew Nagle, paralysed by a knife wound to the spinal cord,
received an array in his motor cortex in 2004 and used it to move a cursor, open email, and operate
a prosthetic hand.[^hochberg2006] The BrainGate2 consortium — Brown, Massachusetts General Hospital,
Stanford, and later UC Davis — has run the successor studies since.

In the United States the array holds a clearance for temporary recording of up to thirty days,
which supports intraoperative and epilepsy-monitoring use. Every long-term human implant is
conducted under an investigational device exemption, not as approved therapy.

## Role in human BCI research

Because the array resolves individual neurons, it supports the highest-dimensional control yet
demonstrated. Robotic-arm control with multiple degrees of freedom, high-rate cursor typing, and
several of the leading [[speech-neuroprosthesis]] results rest on Utah arrays, usually two to
four implanted across motor and premotor areas; the fastest reported speech rates have come from
surface grids rather than penetrating arrays, and the two approaches sit close together. It is also the sensor behind most attempts at
writing information back in through intracortical microstimulation, the route by which touch
percepts have been delivered to users of [[neuroprosthetics]].

A variant with shanks of graded length, the Utah slanted electrode array, is designed for
peripheral nerve rather than cortex, and has been used to give sensation to users of
[[myoelectric-prosthetics]] — an early instance of the closed-loop principle that
[[sensory-augmentation]] research generalizes to senses the body never had.

Depth electrodes of other designs, rather than the Utah array, carry most of the human hippocampal
work behind the [[memory-prosthesis]] literature, because the relevant structures lie centimetres
below the cortical surface that the array is built to sample.

> [!note] What "96 channels" does and does not mean
> Each channel may record several neurons whose spikes must be sorted, or none at all. The number
> of channels yielding usable single units is generally well below the number wired, falls over
> time, and varies by implant site and participant. Reported channel counts are a property of the
> hardware, not of the signal obtained.

## Failure modes and longevity

Arrays degrade. Some human implants have produced usable neural control for more than a thousand
days, and a few for several years, but signal yield generally declines over months to
years.[^simeral2011]

The causes are mixed. Insertion injury and chronic micromotion provoke a foreign-body response:
microglia activate, astrocytes form an encapsulating sheath, and neurons retreat from the tip.
Independently, the device itself deteriorates — parylene insulation cracks and delaminates, tip
metallization corrodes, wire bonds and connectors fail. A systematic analysis in non-human primates
found that material and mechanical failures accounted for a large share of losses, which matters
because it implies that improving biocompatibility alone will not fix
longevity.[^barrese2013]

Percutaneous pedestals add infection risk and prevent the user from living with the system
untethered. This is the specific limitation that motivates fully implanted wireless designs,
including [[neuralink]]'s sealed package and the surface arrays discussed under
[[ecog-interfaces]].

## Alternatives

Three lines of work aim to displace the array. Higher-density silicon probes, most prominently
Neuropixels, put hundreds to thousands of recording sites along a single thin shank and have been
used acutely in human cortex during neurosurgery.[^jun2017][^paulk2022] Flexible polymer threads,
as used by Neuralink, trade rigidity for reduced chronic tissue reaction but require robotic
insertion. Non-penetrating approaches — subdural films and the endovascular [[stentrode]] — give up
single-neuron resolution entirely in exchange for stability and lower surgical risk.

A fourth approach avoids metal electrodes altogether: [[optogenetics]] can read and write neural
activity optically, but requires genetically modifying the target neurons, which places it far from
routine human use.

None has yet accumulated a comparable human track record. That is the array's main remaining
advantage: two decades of implants across multiple sites, with known failure statistics, against
competitors whose chronic behaviour in humans is largely unmeasured.

## Outlook

The realistic future for penetrating arrays is not that they win or lose outright but that they
specialize. Applications demanding many independent control dimensions — dexterous limb control,
fluent speech, and any bidirectional system delivering fine sensory feedback — need single-neuron
resolution that surface and vascular electrodes cannot supply. Applications needing a reliable
switch do not.

Whether that specialization survives depends on an unresolved empirical question: whether a
penetrating device can be built that still records well after ten years in a human brain. The
best-established implanted neurotechnologies, among them the [[cochlear-implant]] and
[[deep-brain-stimulation]], stimulate rather than record and sit in tissue that tolerates
them for decades. No recording array has demonstrated equivalent durability, and the failure
analyses suggest the answer depends as much on materials engineering as on neuroscience.

## See also

- [[brain-computer-interface]]
- [[neural-decoding]]
- [[speech-neuroprosthesis]]
- [[ecog-interfaces]]
- [[stentrode]]
- [[neuralink]]
- [[neuroprosthetics]]
- [[memory-prosthesis]]

## References

[^campbell1991]: `paper` Campbell, P. K., Jones, K. E., Huber, R. J., Horch, K. W. and Normann, R. A. "A silicon-based, three-dimensional neural interface: manufacturing processes for an intracortical electrode array." *IEEE Transactions on Biomedical Engineering*, 1991.

[^maynard1997]: `paper` Maynard, E. M., Nordhausen, C. T. and Normann, R. A. "The Utah intracortical electrode array: a recording structure for potential brain-computer interfaces." *Electroencephalography and Clinical Neurophysiology*, 1997.

[^hochberg2006]: `paper` Hochberg, L. R. et al. "Neuronal ensemble control of prosthetic devices by a human with tetraplegia." *Nature*, 2006. {Centres on the first BrainGate participant, performing tasks in supervised laboratory sessions rather than in independent daily use.}

[^simeral2011]: `paper` Simeral, J. D., Kim, S.-P., Black, M. J., Donoghue, J. P. and Hochberg, L. R. "Neural control of cursor trajectory and click by a human with tetraplegia 1000 days after implant of an intracortical microelectrode array." *Journal of Neural Engineering*, 2011. {One participant and one long-lived implant; it establishes that the duration is possible, not that it is typical.}

[^barrese2013]: `paper` Barrese, J. C. et al. "Failure mode analysis of silicon-based intracortical microelectrode arrays in non-human primates." *Journal of Neural Engineering*, 2013.

[^jun2017]: `paper` Jun, J. J. et al. "Fully integrated silicon probes for high-density recording of neural activity." *Nature*, 2017.

[^paulk2022]: `paper` Paulk, A. C. et al. "Large-scale neural recordings with single neuron resolution using Neuropixels probes in human cortex." *Nature Neuroscience*, 2022. {Recordings were made acutely during neurosurgery; these probes have no chronic human track record to compare with the array's.}

==============================================================================
ARTICLE: uterus-transplantation
TITLE:   Uterus transplantation
PORTAL:  Reproduction & Development
URL:     https://futurehumanwiki.com/wiki/uterus-transplantation
SOURCE:  https://futurehumanwiki.com/raw/uterus-transplantation
==============================================================================

---
title: "Uterus transplantation"
slug: "uterus-transplantation"
type: "intervention"
status: "emerging"
horizon: "present"
trl: 8
categories: ["reproduction", "bodies"]
tags: ["transplantation", "reproduction", "infertility", "pregnancy", "immunosuppression", "ivf", "surgery"]
summary: "A temporary transplant of a donated uterus that lets a woman without a working one carry a pregnancy, then is removed so that immunosuppression can stop."
updated: "2026-07-27"
humanEvidence: "Dozens of live births have been reported since the first in Sweden in 2014, from both living and deceased donors; graft loss and preterm delivery are common."
access: "Performed at a small number of specialist centres, largely under research protocols, rarely reimbursed, and only after the recipient has completed IVF."
reversibility: "difficult"
issues: ["Worldwide birth and graft-failure counts are stated qualitatively; a registry citation would sharpen them.", "Outcomes for the children get one sentence; the follow-up literature is not surveyed."]
---

```infobox
{
  "caption": "Reproductive transplant procedure",
  "rows": [
    { "label": "Indication", "value": "Absolute uterine factor infertility" },
    { "label": "First live birth", "value": "Sweden, 2014" },
    { "label": "Donor", "value": "Living or deceased" },
    { "label": "Conception", "value": "IVF embryo transfer" },
    { "label": "Delivery", "value": "Caesarean, always" },
    { "label": "Graft retained for", "value": "One or two pregnancies" },
    { "label": "Births reported", "value": "Dozens worldwide" },
    { "label": "Readiness", "value": "TRL 8" }
  ]
}
```

**Uterus transplantation** is the transfer of a donated uterus into a woman who has none, or whose
own cannot carry a pregnancy, so that she can gestate and deliver a child. Almost alone among
transplants, it is intended from the outset to be temporary: the graft is removed after one or two
births so that lifelong immunosuppression is never required. Since the first live birth in Sweden in
2014, dozens of children have been born this way — a record no engineered substitute for the uterus,
the [[artificial-womb]] included, can claim.

```keyfacts
[
  { "value": "2014", "label": "First live birth", "note": "Gothenburg, Sweden; living donor" },
  { "value": "2017", "label": "First US birth and first deceased-donor birth", "note": "Dallas and São Paulo" },
  { "value": "<100", "label": "Births reported worldwide", "note": "as of 2026, from programmes on several continents" }
]
```

## Who it is for

The indication is absolute uterine factor infertility: the uterus is absent, or present but incapable
of sustaining a pregnancy. The largest congenital cause is Mayer–Rokitansky–Küster–Hauser
syndrome, a form of Müllerian agenesis in which the uterus and upper vagina fail to develop while the
ovaries, hormones, and karyotype are typically normal; it occurs in something on the order of one in
4,000 to 5,000 female births. The acquired causes are more varied: hysterectomy for postpartum
haemorrhage, cervical cancer, or fibroids; severe intrauterine adhesions; radiation injury.

The detail that makes the procedure attractive to candidates is that the ovaries usually work. A woman
with Müllerian agenesis produces her own oocytes, so a child gestated in a transplanted uterus is
genetically hers. The alternatives — adoption and gestational surrogacy — either sever that link or
place the pregnancy in another woman's body, and surrogacy is prohibited or legally unenforceable in
much of Europe, including Sweden, where the first sustained programme was built.

## How it works

Conception cannot happen the ordinary way. The graft is anastomosed to blood vessels but not to the
recipient's fallopian tubes, so sperm and egg never meet inside her. Every candidate therefore
completes in vitro fertilisation, and most programmes bank frozen embryos before the transplant is
attempted; this is the same laboratory pipeline that supports [[embryo-selection]], used here simply
to produce embryos before there is a uterus to carry them.

The donor operation removes the uterus with long vascular pedicles, usually including segments of the
internal iliac vessels. Venous drainage is the technical crux: uterine veins are thin-walled,
variable, and awkward to sew, and some teams take utero-ovarian veins instead. In the recipient, the
graft is joined to the external iliac vessels, suspended from the pelvic ligaments, and connected to
a vaginal cuff. Menstruation returning within weeks to months is the first sign that the graft is
alive.

Nothing reconnects the nerves. A transplanted uterus is denervated, which means the recipient does
not feel contractions and cannot labour normally; every reported delivery has been by caesarean
section. Immunosuppression is typically tacrolimus-based and adjusted for pregnancy — mycophenolate
is teratogenic and is exchanged for an alternative before any embryo transfer. Rejection is monitored
by cervical biopsy rather than by symptoms, because episodes are common and frequently silent, and
most are reversed with corticosteroids.

Embryo transfer is generally deferred for several months to a year after transplantation. After one
or two live births, the graft is removed in a second major operation and the drugs stop.

> [!key] The transplant meant to end
> A solid-organ transplant is normally kept for as long as it lasts, and the recipient accepts
> immunosuppression, infection risk, and raised cancer risk for the rest of their life as the price of
> staying alive. A uterus is not life-sustaining, so the calculation inverts: the graft is worth
> keeping only while it is being used, and the standard of success includes taking it out. That is why
> its reversibility differs from the implanted devices and edited cells covered elsewhere on this wiki
> — undoing it requires surgery, but the undoing is part of the plan.

## Development history

```timeline
[
  { "year": "2000", "title": "First human attempt", "text": "A team in Saudi Arabia transplants a uterus from a living donor. The graft is removed after about three months when its blood supply thromboses." },
  { "year": "2011", "title": "First deceased-donor transplant", "text": "A team at Akdeniz University in Turkey transplants a uterus from a multi-organ donor. The first pregnancies are lost; a live birth from the same graft is reported only years afterwards." },
  { "year": "2012", "title": "The Montreal Criteria", "text": "Lefkowitz, Edwards and Balayla publish conditions for ethically permissible uterine transplantation, covering recipient, donor and institutional requirements." },
  { "year": "2012–2013", "title": "The Gothenburg trial", "text": "Mats Brännström's group at the University of Gothenburg performs nine living-donor transplants, most from mothers and other close relatives, after more than a decade of animal work." },
  { "year": "2014", "title": "First live birth", "text": "A boy is born in Sweden to a recipient with Müllerian agenesis, delivered preterm by caesarean after she develops pre-eclampsia." },
  { "year": "2017", "title": "United States and deceased-donor births", "text": "Baylor University Medical Center in Dallas reports the first birth in the United States; a team in São Paulo reports the first anywhere from a deceased donor." },
  { "year": "2016–2025", "title": "Diffusion", "text": "Programmes open in more than a dozen countries, robotic donor hysterectomy is introduced to shorten the donor operation, and the United Kingdom reports its first birth in 2025." }
]
```

The first attempt, in Saudi Arabia in 2000, established that the graft could survive at all and that
vascular thrombosis was the thing most likely to kill it.[^fageeh2002] The Turkish case a decade later
showed that a uterus from a deceased donor could be transplanted and could conceive.[^ozkan2013] The
first pregnancies there were lost, and a live birth from that graft came only years afterwards. What
separated the Swedish programme from both was preparation: more than a decade of work in rodents,
sheep, and non-human primates on vascular technique and on whether a transplanted
uterus could support gestation in any species before it was tried in a person. The 2014 birth was
reported with the recipient, the donor, the immunosuppressive regimen, and the pregnancy course
described in full.[^brannstrom2015] The United States programme at Baylor followed a similar
structured-trial model,[^testa2018] and the Brazilian case removed the assumption that a living donor
was necessary.[^ejzenberg2019]

## Outcomes so far

Fewer than a hundred births have been reported worldwide, from programmes on several continents,
and the count is compiled through an international registry maintained by the transplant societies
rather than through any regulator. On the [[technology-readiness-level|standard readiness scale]] the
procedure sits near the top — a complete system, qualified through repeated use — while remaining
concentrated in a small number of centres with the surgical, reproductive-medicine, and transplant
expertise it requires in one building.

The failures are as informative as the births. A substantial minority of grafts have been lost, most
of them in the first weeks and most to thrombosis at the vascular anastomoses, which forces an
unplanned hysterectomy and ends the attempt. Among pregnancies that continue, preterm delivery and
hypertensive disorders including pre-eclampsia are considerably more frequent than in the general
obstetric population, and caesarean delivery is universal by design. The children have generally been
of appropriate size for their gestational age, and no consistent pattern of malformation has been
reported, but the oldest of them are barely into their second decade and systematic long-term
follow-up is thin.

> [!caution] What the birth count does not show
> Announcements report births; they report graft losses, failed embryo transfers, and recipients who
> never became pregnant far less consistently. The published denominator — how many women were
> screened, transplanted, and left without a child — is incomplete, and no randomised or matched
> comparison against surrogacy or adoption exists or is likely to. A live birth demonstrates that the
> procedure can work. It does not establish how often it does.

## Donors, risk, and consent

Living donors have most often been the recipient's mother or sister, and postmenopausal uteri have
proved usable — the donor in the Swedish case that produced the first birth was a family friend who
had been through menopause years earlier. The donation is not a small thing. Removing a uterus with vascular pedicles long enough
to graft is a longer and more dissection-heavy operation than a standard hysterectomy, with ureteral
injury as the characteristic complication; robotic assistance was introduced partly to reduce that
burden. For the donor the procedure is permanent in a way it is not for the recipient.

Deceased donation removes that objection entirely and expands the pool, at the cost of less workup,
less scheduling control, and a graft whose history is known only from records. Which source is
preferable is an open practical question rather than a settled one, and programmes have split.

> [!debate] A non-lifesaving transplant with three parties at risk
> Critics argue that subjecting a healthy woman to a long operation, a second woman to
> immunosuppression, and a fetus to exposure to those drugs — for an outcome that is not survival but
> gestation — sits outside what transplantation has previously justified, and that surrogacy or
> adoption achieves parenthood without any of it. Supporters answer that the same reasoning would
> forbid living kidney donation to a patient who could dialyse, that surrogacy is unavailable or
> illegal for many candidates, and that carrying a pregnancy is what these women are asking for and
> is not interchangeable with obtaining a child. The Montreal Criteria were an early attempt to fix
> the boundary,[^montreal2012] and professional bodies have generally held that the procedure is
> defensible under research protocols with independent donor advocacy.[^asrm2018]

## Limitations and open questions

Cost and concentration are the immediate limits. The procedure requires a transplant team, a
reproductive endocrinology unit, a high-risk obstetric service, and years of follow-up; it is rarely
reimbursed, and the pattern described in [[access-and-inequality]] applies with unusual force to an
intervention that is expensive, elective, and available in a handful of cities.

Whether transgender women could be recipients is discussed in the literature; no such transplant has
been reported. The obstacles are anatomical and endocrine rather than ideological — a male pelvis has
different vascular geometry and no vaginal canal to attach the graft
to without prior surgery, and the hormonal support of a pregnancy would have to be supplied entirely
by drugs — and the question is a recurring test case for arguments about
[[morphological-freedom]] and for the framing of medical need set out in
[[disability-rights-and-enhancement]]. It is also the version of the procedure most likely to be
argued about long before it is attempted, which is the pattern [[precautionary-principle|precautionary
reasoning]] tends to produce.

## Alternatives and outlook

The engineered alternative would be a uterus that is not donated at all. Groups including Brännström's
have recellularized [[decellularized-scaffolds|decellularized uterine scaffolds]] and reported
pregnancies in rats, and endometrial [[organoids]] reproduce part of the tissue's cyclical behaviour
in culture. Neither approaches a transplantable human organ, and the difficulty is the one that limits
[[lab-grown-organs]] and [[organ-bioprinting]] generally: building a vascular tree that can be plumbed
into a circulation. A uterus built from a patient's own [[induced-pluripotent-stem-cells]] would remove
both the donor and the immunosuppression, which is why it appears in the field's own roadmaps as a
long-term goal rather than a near-term prospect.

The instructive comparison is with the technologies that have not produced a
child. [[ectogenesis]] in its complete form — gestation from fertilisation to term outside a body — has no
demonstration in any mammal, and the devices built so far take over a pregnancy already underway, in
sheep. [[in-vitro-gametogenesis]] has made functional gametes in mice and not in humans. Uterus
transplantation reached live births by using an organ that already works, transplanted by methods
adapted from other transplants, and accepting the costs of doing so. It is a reminder that
biological problems are more often solved by moving working tissue than by manufacturing it, and that
the reproductive technology furthest from the futurist imagination is the one with children to show
for itself. Whether it becomes standard care for absolute uterine factor infertility or remains the
speciality of a few centres depends less on surgery than on whether health systems decide that
carrying a pregnancy, as distinct from having a child, is a medical need they are obliged to meet.

## See also

- [[artificial-womb]]
- [[ectogenesis]]
- [[in-vitro-gametogenesis]]
- [[reproductive-longevity]]
- [[organ-shortage]]
- [[lab-grown-organs]]
- [[tissue-engineering]]
- [[bioethics-of-enhancement]]

## References

[^fageeh2002]: `paper` Fageeh, W. et al. "Transplantation of the human uterus." *International Journal of Gynecology & Obstetrics*, 2002. {The first reported human attempt; the graft was removed after about three months when its vessels thrombosed.}
[^ozkan2013]: `paper` Ozkan, O. et al. "Preliminary results of the first human uterus transplantation from a multiorgan donor." *Fertility and Sterility*, 2013.
[^brannstrom2015]: `paper` Brännström, M. et al. "Livebirth after uterus transplantation." *The Lancet*, 2015. {The recipient had Müllerian agenesis; the uterus came from a living donor who had passed menopause years before.}
[^testa2018]: `paper` Testa, G. et al. "First live birth after uterus transplantation in the United States." *American Journal of Transplantation*, 2018.
[^ejzenberg2019]: `paper` Ejzenberg, D. et al. "Livebirth after uterus transplantation from a deceased donor in a recipient with uterine infertility." *The Lancet*, 2019. {The birth occurred in December 2017 in São Paulo; the report established that a living donor is not required.}
[^montreal2012]: `paper` Lefkowitz, A., Edwards, M. and Balayla, J. "The Montreal Criteria for the Ethical Feasibility of Uterine Transplantation." *Transplant International*, 2012.
[^asrm2018]: `statement` American Society for Reproductive Medicine, Ethics Committee. "American Society for Reproductive Medicine position statement on uterus transplantation: a committee opinion." *Fertility and Sterility*, 2018.

==============================================================================
ARTICLE: wearable-health-sensors
TITLE:   Wearable health sensors
PORTAL:  Longevity & Aging
URL:     https://futurehumanwiki.com/wiki/wearable-health-sensors
SOURCE:  https://futurehumanwiki.com/raw/wearable-health-sensors
==============================================================================

---
title: "Wearable health sensors"
slug: "wearable-health-sensors"
type: "technology"
status: "established"
horizon: "present"
trl: 9
categories: ["longevity", "cybernetics"]
tags: ["wearables", "heart rhythm", "sleep", "screening", "consumer health", "measurement"]
summary: "Consumer devices worn on the wrist or finger that continuously record pulse waveforms, movement, and skin temperature, and infer heart rhythm, sleep, and activity from them."
updated: "2026-07-27"
humanEvidence: "Prospective studies in hundreds of thousands of people show smartwatch algorithms can flag atrial fibrillation, and a randomized trial in older adults raised diagnosis rates; no trial has shown better clinical outcomes."
access: "Sold over the counter worldwide from roughly a hundred to several hundred dollars, some features behind subscriptions, several available only where a regulator has cleared them."
reversibility: "reversible"
issues: ["Non-US regulatory treatment of cleared features is covered only in passing", "Runs past the 1,600-word target unless the article is treated as a flagship"]
---

```infobox
{
  "caption": "Consumer biosensing hardware",
  "rows": [
    { "label": "Form factors", "value": "Watch, ring, strap, patch" },
    { "label": "Core sensors", "value": "Optical pulse, accelerometer, single-lead ECG" },
    { "label": "Built-in smartwatch ECG cleared", "value": "2018 (US)" },
    { "label": "Best-supported use", "value": "Irregular rhythm notification" },
    { "label": "Not measurable", "value": "Blood glucose, non-invasively" },
    { "label": "Main makers", "value": "Apple, Google, Samsung, Garmin, Oura, Whoop" },
    { "label": "Readiness", "value": "TRL 9" }
  ]
}
```

**Wearable health sensors** are consumer devices worn on the wrist, finger, or torso that record optical pulse waveforms, movement, and skin temperature continuously, and infer heart rhythm, sleep, and activity from those signals. They are among the most widely deployed health-measurement devices ever made, worn by hundreds of millions of people who are mostly not patients. Almost nothing they report has been tested against a clinical outcome; the exception is detection of atrial fibrillation, which has been studied in cohorts of several hundred thousand and cleared by regulators as a notification feature rather than a diagnosis.

```keyfacts
[
  { "value": "0.52%", "label": "Apple Heart Study participants notified", "note": "of roughly 419,000 enrolled, over a median of about four months" },
  { "value": "~1/3", "label": "Notified participants with confirmed AF", "note": "among the minority who returned usable ECG patch recordings" },
  { "value": "0", "label": "Consumer devices measuring glucose non-invasively", "note": "FDA safety communication, 2024" }
]
```

## What the sensors measure

Three sensor types do most of the work. Photoplethysmography shines green light into the skin and measures how much reaches a photodiode; capillary blood volume rises and falls with each beat, so the returning signal yields interbeat intervals, and from those come heart rate, heart-rate variability, and a measure of rhythm irregularity. A triaxial accelerometer records movement, supplying step counts, fall detection, and the raw material for sleep and wake classification. Several watches add a single-lead electrocardiogram: touching the crown, a button, or the metal frame closes a circuit across the body and produces a roughly thirty-second tracing comparable to lead I of a clinical ECG.

Rings use infrared light on the finger, where perfusion is stronger and motion artefact lower than at the wrist. Some devices add skin temperature relative to a personal baseline rather than core temperature, bioimpedance, and red-and-infrared pulse oximetry.

The hardware measures light, acceleration, and voltage. Sleep stages, stress, recovery, and readiness are inferences layered on top by proprietary models, and each has to be validated separately. A device that captures the pulse waveform faithfully has established nothing about the sleep classifier built from it.

## Development history

```timeline
[
  { "year": "1965", "title": "The ten-thousand-step target", "text": "A Japanese manufacturer markets a pedometer named manpo-kei, or ten-thousand-step meter. The number was a marketing choice and survives as a default goal in software sixty years later." },
  { "year": "1980s", "title": "Chest-strap heart-rate monitors", "text": "Wireless electrode straps reach endurance athletes, establishing continuous cardiac measurement outside a clinic." },
  { "year": "2009", "title": "Fitbit ships its first tracker", "text": "A clip-on accelerometer sold direct to consumers, framed around steps and sleep rather than sport." },
  { "year": "2015", "title": "Wrist and finger optical sensing", "text": "The Apple Watch ships with green-light photoplethysmography; Oura's first finger-worn ring reaches backers around the same time." },
  { "year": "2018", "title": "First cleared consumer ECG", "text": "US regulators authorize an electrocardiogram app and an irregular rhythm notification on a general-purpose smartwatch through the De Novo pathway." },
  { "year": "2019", "title": "Apple Heart Study published", "text": "A single-arm study of roughly 419,000 participants reports how often the notification fires and how often a mailed ECG patch confirms it." },
  { "year": "2021", "title": "LOOP trial reports", "text": "Implanted monitors roughly triple atrial fibrillation diagnoses in older adults at risk without a statistically significant reduction in stroke." },
  { "year": "2022", "title": "Screening judged unproven", "text": "The Fitbit Heart Study is published, and the US Preventive Services Task Force finds the evidence insufficient to recommend for or against screening asymptomatic adults for atrial fibrillation." },
  { "year": "2024", "title": "Glucose warning and apnea clearances", "text": "The FDA warns against devices claiming non-invasive blood glucose measurement; sleep apnea notification features are authorized for Samsung and Apple watches." },
  { "year": "2026", "title": "First randomized consumer-wearable screening trial", "text": "The EQUAL trial in the Netherlands reports that six months of smartwatch monitoring finds more new atrial fibrillation than usual care in adults over 65 at elevated stroke risk. The endpoint was diagnosis, not stroke." }
]
```

The industry consolidated as the features became regulated. Google completed its acquisition of Fitbit in 2021; Apple, Samsung, and Garmin build sensing into general-purpose watches; Oura and Whoop couple hardware to subscriptions, which shifts the incentive from a single sale toward continuous engagement with the data.

## Heart rhythm, the best-supported use

The Apple Heart Study enrolled roughly 419,000 US participants and monitored them for a median of about four months. Just over half a percent received an irregular pulse notification. Those who did were mailed an ECG patch, and among the minority who returned usable recordings, atrial fibrillation was present in about a third during the patch period.[^perez2019] Two readings of that fraction are both fair. It understates the algorithm, because atrial fibrillation is often paroxysmal and the patch arrived days after the alert. It also flatters the algorithm, because only notified participants were investigated, so the study could not estimate how many cases were missed.

The Fitbit Heart Study used a similar design in a comparable number of people and reported a much higher positive predictive value, defined against atrial fibrillation present during the window the algorithm had flagged rather than across a later monitoring period.[^lubitz2022] The two headline figures are not comparable, and the difference lies in the definition rather than in the hardware.

Clearances for rhythm features are narrow by design. Labelling restricts them to adults without a prior atrial fibrillation diagnosis and states that they are not a substitute for clinical evaluation; what the device produces is a prompt to seek one.

## Sleep and breathing

Sleep tracking is heavily marketed and weakly validated. A comparison of seven consumer devices against polysomnography found that they estimated total sleep time reasonably well in healthy sleepers and classified sleep stages poorly, with performance varying by device and by stage.[^chinoy2021] The reason is structural. Polysomnography scores stages from EEG, eye movement, and muscle tone; a wrist device infers them from movement and cardiac features, and no wrist-accessible signal maps cleanly onto the stage definitions.

A second failure mode matters more. Devices that infer sleep from stillness tend to score motionless wakefulness as sleep, so they overestimate sleep in precisely the people with insomnia who consult the data most closely. That also constrains how the epidemiology summarized under [[sleep-and-longevity]] can be translated into a nightly score.

> [!note] Orthosomnia
> Sleep clinicians coined the term for patients whose insomnia is driven or worsened by the pursuit of a perfect tracker score, and who discount a normal sleep study in favour of the device.[^baron2017] The report is a small case series rather than a prevalence estimate, and it remains the clearest documented instance of a consumer health metric aggravating the condition it claims to monitor.

Sleep apnea notification is among the most recent cleared functions on these devices. Samsung received US authorization in early 2024 and Apple later the same year; both detect breathing-related disturbance across multiple nights and are framed as a reason to seek testing. Confirmation still requires a sleep study.

## What these devices do not measure

No consumer wearable measures blood glucose. The FDA issued a safety communication in 2024 warning against smartwatches and smart rings sold with claims to measure blood glucose without piercing the skin, and stating that it had authorized no such device.[^fda2024] A watch that displays glucose is receiving the value by radio from a [[continuous-glucose-monitoring|continuous glucose monitor]] whose filament sits in interstitial fluid under the skin.

Cuffless blood pressure from optical sensors is likewise unvalidated, and cardiology bodies do not recommend the current devices for diagnosis or treatment decisions. Samsung's estimate must be recalibrated against an upper-arm cuff roughly monthly, and it reached users in the United States in 2026 as a general-wellness feature rather than a cleared medical one. A hypertension notification cleared for Apple's watch in 2025 sidesteps the accuracy problem by reporting a pattern consistent with chronically raised pressure over about a month while deliberately producing no number. Hydration status, blood alcohol, and circulating cortisol have all been claimed for optical sensors without published validation. Stress and recovery scores are repackaged heart-rate variability, which shifts with posture, breathing rate, caffeine, alcohol, and illness.

> [!caution] The glucose claim is the field's clearest falsehood
> Non-invasive optical glucose measurement has been announced repeatedly for decades and demonstrated by nobody. Products making the claim are sold anyway, and a person with diabetes acting on a fabricated number can be harmed within hours. This is the sharpest example of a general pattern in [[consumer-blood-testing|consumer diagnostics]], where the marketed capability runs ahead of the physics.

## The screening problem

The central critique of consumer health sensing is arithmetic rather than technical. A test with imperfect specificity applied to a population in which the condition is rare produces mostly false positives, however good it looks in a clinic. Atrial fibrillation is uncommon before fifty, and the cohorts wearing these devices skew young. Each false alert generates a clinic visit, often a monitor, sometimes an echocardiogram, and reliably some anxiety, plus incidental findings that start cascades of their own.

Whether finding more atrial fibrillation helps is a separate question, and the strongest evidence is discouraging. The LOOP trial randomized older adults with stroke risk factors to an implanted loop recorder or usual care. The monitor roughly tripled diagnoses and increased anticoagulation, and the reduction in stroke or systemic embolism was not statistically significant.[^svendsen2021] The EQUAL trial in the Netherlands, the first randomized test of a consumer device, reported in 2026 that six months of smartwatch monitoring produced roughly four times as many new atrial fibrillation diagnoses as usual care in adults over 65 at elevated stroke risk. It counted diagnoses, not strokes.[^vansteijn2026] In 2022 the US Preventive Services Task Force concluded that evidence was insufficient to assess the balance of benefits and harms of screening asymptomatic adults for atrial fibrillation.[^uspstf2022]

> [!debate] Detection is not benefit
> Device makers cite cases where a notification preceded a diagnosis, which is real and unsystematic. Cardiologists point out that the randomized tests of intensive rhythm monitoring have raised diagnosis and treatment rates without demonstrating fewer strokes. Both positions are consistent with the evidence, because the trials run so far, on implants and on watches alike, have counted diagnoses rather than events.

## Activity, longevity, and behaviour change

Step counts are the oldest wearable metric and the one with the most epidemiology behind it. Pooled cohort data associate higher daily step counts with lower all-cause mortality, with the association flattening well below the familiar ten-thousand target, and at lower counts in older adults.[^paluch2022] These are observational findings and carry the obvious confound in both directions: illness reduces walking as well as following from it.

Evidence that wearing a device changes behaviour durably is weaker than the market implies. In a randomized trial of a behavioural weight-loss programme in young adults, the group given a wearable activity monitor had lost less weight at two years than the group self-monitoring without one.[^jakicic2016] Nothing about that result argues against [[exercise-and-aging|physical activity]], which remains the best-supported intervention for the functional outcomes the longevity field cares about. It argues that measurement is not motivation.

Wearable streams increasingly feed other consumer health products: [[quantified-self]] tracking, [[biological-age]] estimates sold alongside [[epigenetic-clock|methylation tests]], the personalized models discussed under [[human-digital-twins]], and the [[dietary-supplements|supplement]] regimens of the [[biohacking]] community. Weight-management programmes built around [[glp-1-receptor-agonists]] have an obvious use for activity and lean-mass tracking, whose contribution to results has not been established. None of these composites has cleared the validation bar set out under [[aging-biomarkers]], and a watch-estimated maximal oxygen uptake is a model output rather than the treadmill measurement whose association with mortality gave the metric its reputation.

## Outlook

The technology is mature and the evidence is not. Sensors are cheap and accurate enough at the physical layer, and body-worn measures have begun to serve as endpoints in regulated drug trials. What is missing is the study that would settle the consumer question: randomizing device use against usual care in a defined population and counting strokes and deaths rather than diagnoses.

Governance lags further. Health data from a consumer device sits largely outside medical privacy law in the United States, in the hands of advertising companies and insurers, and life insurers have offered premium discounts for sharing activity data. The concerns are those raised under [[genetic-discrimination]] and, for neural signals, under [[mental-privacy]] and [[neurorights]], transposed to a stream that is continuous, behavioural, and voluntarily surrendered. How voluntary it is depends on whether the device arrived through a purchase or through an employer or insurer programme, a distinction treated under [[quantified-self]]. The devices also cost money and reach the people least likely to need screening, which makes them a case for [[access-and-inequality]].

The contrast with implants clarifies what these devices are. A [[brain-computer-interface]] accepts surgical risk for signal quality that surface sensing cannot approach, and [[non-invasive-neuromodulation]] accepts weaker effects to avoid the surgery. Wearables sit at the far end of that trade, buying scale and reversibility at the cost of resolution. The open question is whether a low-resolution signal collected continuously from millions of people is worth more than a high-resolution signal collected from patients occasionally, and whether the diagnoses it produces amount to anything like the [[compression-of-morbidity]] the marketing invokes.

## See also

- [[continuous-glucose-monitoring]]
- [[quantified-self]]
- [[consumer-blood-testing]]
- [[human-digital-twins]]
- [[sleep-and-longevity]]
- [[aging-biomarkers]]
- [[exercise-and-aging]]
- [[healthspan]]

## References

[^perez2019]: `paper` Perez, M.V. et al. "Large-Scale Assessment of a Smartwatch to Identify Atrial Fibrillation." *New England Journal of Medicine*, 2019. {The cohort was self-selected and skewed young, unlike the older populations in which atrial fibrillation screening is usually considered.}
[^lubitz2022]: `paper` Lubitz, S.A. et al. "Detection of Atrial Fibrillation in a Large Population Using Wearable Devices: The Fitbit Heart Study." *Circulation*, 2022.
[^svendsen2021]: `paper` Svendsen, J.H. et al. "Implantable loop recorder detection of atrial fibrillation to prevent stroke (The LOOP Study): a randomised controlled trial." *The Lancet*, 2021. {The cleanest available test of detection against benefit, in a higher-risk population than the one wearing consumer devices.}
[^vansteijn2026]: `paper` van Steijn, N. et al. "Enhanced Detection and Prompt Diagnosis of Atrial Fibrillation Using Apple Watch: A Randomized Controlled Trial." *Journal of the American College of Cardiology*, 2026. {A few hundred participants over 65 with elevated stroke risk; the endpoint was new atrial fibrillation diagnosis over six months, not stroke.}
[^uspstf2022]: `report` US Preventive Services Task Force. "Screening for Atrial Fibrillation: US Preventive Services Task Force Recommendation Statement." *JAMA*, 2022.
[^fda2024]: `regulator` US Food and Drug Administration. "Do Not Use Smartwatches or Smart Rings to Measure Blood Glucose Levels: FDA Safety Communication," 2024.
[^chinoy2021]: `paper` Chinoy, E.D. et al. "Performance of seven consumer sleep-tracking devices compared with polysomnography." *Sleep*, 2021. {Conducted in healthy sleepers, which is the easy case; accuracy in disordered sleep is generally worse.}
[^baron2017]: `paper` Baron, K.G. et al. "Orthosomnia: Are Some Patients Taking the Quantified Self Too Far?" *Journal of Clinical Sleep Medicine*, 2017.
[^paluch2022]: `paper` Paluch, A.E. et al. "Daily steps and all-cause mortality: a meta-analysis of 15 international cohorts." *The Lancet Public Health*, 2022.
[^jakicic2016]: `paper` Jakicic, J.M. et al. "Effect of Wearable Technology Combined With a Lifestyle Intervention on Long-term Weight Loss: The IDEA Randomized Clinical Trial." *JAMA*, 2016. {Participants were adults aged 18 to 35; the comparison group self-monitored diet and activity through a website rather than a worn device.}

==============================================================================
ARTICLE: whole-brain-emulation
TITLE:   Whole brain emulation
PORTAL:  Minds & Consciousness
URL:     https://futurehumanwiki.com/wiki/whole-brain-emulation
SOURCE:  https://futurehumanwiki.com/raw/whole-brain-emulation
==============================================================================

---
title: "Whole brain emulation"
slug: "whole-brain-emulation"
type: "concept"
status: "speculative"
horizon: "2050s+"
categories: ["minds", "foundations"]
tags: ["emulation", "connectomics", "simulation", "consciousness", "computational neuroscience", "uploading"]
summary: "A proposed procedure for reproducing a particular mind by scanning one brain's structure in detail, building a computational model from the scan, and running it."
updated: "2026-07-27"
issues: ["The claim that groups are building dynamical models from the fly connectome carries no source."]
---

```infobox
{
  "caption": "Hypothetical technology in computational neuroscience",
  "rows": [
    { "label": "Also called", "value": "Brain emulation, substrate transfer" },
    { "label": "Pipeline", "value": "Scan → translate → run" },
    { "label": "Key roadmap", "value": "Sandberg & Bostrom, 2008" },
    { "label": "Largest full connectome", "value": "Adult fruit fly (2024)" },
    { "label": "Emulated nervous systems", "value": "None, at any scale" },
    { "label": "Status", "value": "No working demonstration" }
  ]
}
```

**Whole brain emulation** is a proposed procedure for reproducing a particular mind on a computer:
scan the structure of one specific brain at sufficient resolution, translate that scan into a
computational model of its components, and run the model fast enough and faithfully enough that it
produces the same behaviour the original brain would have produced. It is distinct from building
an artificial mind from scratch, and distinct from simulating a generic brain — the object is a
copy of *this* brain, retaining its memories and dispositions. No nervous system of any size has
been emulated in this sense, including the 302-neuron nervous system of the nematode *C. elegans*,
whose wiring diagram has been published since 1986.

## Overview

The idea is usually presented as an engineering problem rather than a scientific one. Its
proponents argue that no new physics is required, only the extension of three existing capabilities:
volume microscopy, computational neuroscience, and computing hardware. That framing carries a
substantive assumption — that there exists some level of biological description below which further
detail can be replaced by statistical noise without changing the emulated system's behaviour. The
assumption is called *scale separation*, and whether it holds for brains is unresolved.

The scale of the object is worth stating plainly. A human brain contains roughly 86 billion neurons
and a comparable number of non-neuronal cells, with synapse counts estimated in the hundreds of
trillions.[^azevedo2009] Every one of those synapses is a structure whose strength, receptor
composition, and history would have to be captured or inferred. No proposal treats this as anything
other than the largest measurement problem ever attempted on a single object.

Whole brain emulation is the technical substrate of [[mind-uploading]]; the two are often conflated
but answer different questions. Emulation asks whether a functionally equivalent process can be run
on other hardware. Uploading asks whether the person goes with it — a question about
[[personal-identity-and-continuity]] that no amount of engineering settles. The philosophical
premise both share, that mental states depend on organization rather than material, is
[[substrate-independence]].

## The pipeline

### Scanning

The scan must capture whatever the model needs. At minimum that means the complete synaptic wiring
of the brain — the connectome — including which neuron contacts which, where, and through what kind
of synapse. Electron microscopy at nanometre resolution can resolve this, but only destructively and
only on tissue that has been chemically fixed, stained with heavy metals, embedded in resin, and cut
into sections tens of nanometres thick. Every published mammalian connectome fragment has been
produced this way. The scale gap is severe: the largest human cortical volume reconstructed to date
is about one cubic millimetre, roughly a millionth of a whole brain. See [[connectomics]].

### Translating

A wiring diagram alone specifies almost nothing about dynamics. The translation step must assign to
each element a model and a parameter set: membrane properties, channel densities, synaptic strengths
and their short-term dynamics, neuromodulatory sensitivities. Some of these can in principle be
inferred from structure — synapse size correlates with strength — but many cannot, and current
electron microscopy does not read out molecular composition. Correlative approaches that image the
same tissue functionally before slicing it, as the MICrONS mouse cortex project did, are the
existing partial answer, but they work only on a living animal, not on a preserved brain.

### Running

The resulting model must be simulated in something close to real time if the emulation is to
interact with anything, and must be embedded in a body or a simulated environment, since a cortex
with no sensory input and no motor output has no obvious behaviour to validate against. The roadmap
by [[anders-sandberg]] and [[nick-bostrom]] treats the environment and body model as a required
component rather than an afterthought.

## Levels of detail

The 2008 roadmap organizes the problem as a ladder of modelling levels, from coarse-grained
connectivity between brain regions at the top, through analogue population models, spiking networks,
detailed compartmental electrophysiology, and metabolic and proteomic detail, down to stochastic
molecular dynamics and, at the bottom, quantum-level simulation.[^sb2008] Each rung down multiplies
the storage and computation required by orders of magnitude. The roadmap's central point is not any
particular estimate but the structure of the bet: emulation is feasible if and only if the necessary
level is high on the ladder. Most computational neuroscientists who engage with the proposal think
the spiking-network level is too coarse; almost nobody thinks the quantum level is required.

> [!key] Why the level matters
> The difference between the coarsest and finest proposed levels is not a factor of ten. It is the
> difference between a problem that large computing facilities might approach this century and one
> that exceeds any physically plausible machine. The roadmap does not resolve which applies.

## Origins

The idea has two lineages. One runs through science fiction, where stored and reinstantiated minds
appear from the 1950s onward. The other runs through the practical demonstration that a nervous
system can be mapped exhaustively: the completion of the *C. elegans* wiring diagram in 1986 showed
that "every connection in an animal" was a finite quantity of work rather than a
figure of speech.[^white1986] Hans Moravec joined the two in 1988, arguing that a brain is a
finite-state physical system, that computing capacity was growing fast enough to matter, and that
the transfer could be done either by scanning or by piecewise replacement.[^moravec1988] The
proposal acquired a technical framework two decades later.

```timeline
[
  { "year": "1956", "title": "Fictional precursor", "text": "Arthur C. Clarke's The City and the Stars describes citizens stored as patterns in a central computer and reissued into bodies, an early literary statement of the idea." },
  { "year": "1986", "title": "The first complete connectome", "text": "White, Southgate, Thomson and Brenner publish the full wiring diagram of the C. elegans hermaphrodite nervous system from electron micrographs, a project that took over a decade." },
  { "year": "1988", "title": "Mind Children", "text": "Hans Moravec sets out both the destructive-scan and the gradual neuron-by-neuron replacement routes, framing emulation as a foreseeable consequence of computing trends." },
  { "year": "2005", "title": "Blue Brain begins", "text": "Henry Markram launches a project at EPFL to reconstruct cortical microcircuitry in biophysical detail, the first sustained attempt at bottom-up brain simulation at scale." },
  { "year": "2008", "title": "The roadmap", "text": "Anders Sandberg and Nick Bostrom publish a technical report at the Future of Humanity Institute defining the scan-translate-run framework and its levels of detail." },
  { "year": "2013–2023", "title": "The Human Brain Project", "text": "An EU flagship funded at around a billion euros; after an open letter of protest from hundreds of neuroscientists in 2014 it was restructured away from whole-brain simulation and ended having delivered research infrastructure instead." },
  { "year": "2024", "title": "A whole adult brain mapped", "text": "The FlyWire consortium publishes the complete synaptic connectome of an adult fruit fly, roughly 140,000 neurons — the first whole brain of an animal with complex behaviour to be reconstructed." }
]
```

## What has actually been built

Nothing that qualifies. The largest published biophysical reconstructions are of cortical
microcircuits, not brains: the Blue Brain Project's 2015 model of a rat somatosensory column
comprised on the order of thirty thousand neurons and tens of millions of synapses, built from
statistical rules rather than from a scan of one animal.[^markram2015] It reproduced some
electrophysiological phenomena and was not an emulation of any individual rat. The Blue Brain
Project wound down at the end of 2024.

Large-scale spiking simulations at the level of whole rodent or human cortex have been run on
supercomputers, but with randomized connectivity and generic neuron models. They demonstrate that
hardware can carry the neuron count; they say nothing about whether the resulting activity
corresponds to a mind.

The nearest thing to a substrate for a real emulation arrived in 2024, when the complete synaptic
wiring of an adult fruit fly was published — roughly 140,000 neurons with cell-type annotation and
predicted neurotransmitter identity.[^dorkenwald2024] Groups have since begun building
whole-brain-scale dynamical models from it. These are properly called simulations rather than
emulations: they use the anatomy of one fly with parameters fitted or assumed, and their outputs are
compared against population statistics rather than against the behaviour of the individual animal
that was scanned. That comparison is the test the field needs and has not yet passed.

The most informative negative result comes from the nematode. *C. elegans* has an invariant nervous
system of 302 neurons, a connectome published four decades ago and refined since with sex-specific
and developmental detail. The OpenWorm project and related efforts have not produced a simulation
that reproduces the animal's behavioural repertoire. The obstacles are exactly the ones the roadmap
identifies as the translation step: the connectome does not specify synaptic polarity, strength, or
the extrasynaptic neuropeptide signalling that reconfigures the circuit.

```keyfacts
[
  { "value": "302", "label": "Neurons in C. elegans", "note": "Connectome published 1986; still not emulated" },
  { "value": "~86 billion", "label": "Neurons in a human brain", "note": "Azevedo et al., 2009, isotropic fractionator count" },
  { "value": "~1 mm³", "label": "Largest human cortex volume reconstructed", "note": "About one millionth of a brain" }
]
```

## Objections

**The single-neuron model is not settled.** Emulation assumes a validated model of the components.
Cortical pyramidal neurons perform nonlinear computation in their dendrites, and the mapping from
input to output is not captured by a threshold unit. Attempts to fit a detailed biophysical neuron
with an artificial network have required networks of substantial depth, which is a measure of how
much is being abstracted away at the spiking level.

**Chemistry is not wiring.** Neuromodulators diffuse and act on receptors at a distance from any
synapse. Eve Marder's work on the crustacean stomatogastric ganglion — a circuit of about thirty
neurons whose anatomy has been fully known for decades — shows that the same wiring produces
qualitatively different rhythms under different neuromodulatory conditions, and that widely
different parameter sets produce identical output.[^marder2012] A structural scan cannot distinguish
those parameter sets. Glia, which outnumber neurons in some regions and modulate synaptic
transmission, are absent from most emulation proposals entirely.

**Validation has no obvious method.** There is no test that distinguishes a correct emulation from a
plausible-sounding one. A behavioural comparison requires an original to compare against, and the
scan is destructive. The standard proposal is to validate on progressively larger animals first,
which is sensible and has not begun. Validation also raises a problem with no precedent in software
engineering: a partially correct emulation of a person would be a system that might be conscious and
might be suffering, and there is no way to test it that does not consist of running it. Proposals
for staged validation on animals inherit the same difficulty in milder form.

**The compute estimates are unreliable.** Forecasts of the hardware required span many orders of
magnitude and depend entirely on the assumed level of detail, which is the unknown. Citing a
particular figure alongside a projection of computing growth, as arguments connected to
[[accelerating-change]] and the [[technological-singularity]] often do, imports the conclusion into
the premise.

> [!debate] How far off
> Estimates from within the emulation community typically fall in the second half of this century.
> Kenneth Miller, a computational neuroscientist at Columbia, argued in 2015 that even mapping the
> relevant structure of a human brain plausibly requires centuries at any realistic rate of
> improvement.[^miller2015] The disagreement is not about the physics; it is about how many
> unknown parameters the translation step hides.

## Relationship to other technologies

Emulation is the destination that gives [[brain-preservation]] and, on some readings,
[[cryonics]] their rationale: if a preserved brain retains the information an emulation would need,
preservation converts a deadline into a storage problem. Whether aldehyde-stabilized or vitrified
tissue retains that information is disputed and depends, again, on which level of detail matters.

It is sometimes proposed that a [[brain-computer-interface]] could provide the read-out instead of a
destructive scan. Current implants sample on the order of a thousand channels from a cortex
containing billions of synapses, and record activity rather than structure; the gap is not one that
incremental electrode counts close, and the methods of [[neural-decoding]] recover motor plans and
attempted speech rather than the standing structure that an emulation would need. Person-specific
computational models of the body, described under [[human-digital-twins]], share the goal of modelling
one individual rather than a generic one, at a resolution many orders of magnitude coarser. Emulation
is also frequently invoked as a route to
[[artificial-general-intelligence]] that sidesteps the need to understand intelligence, and as the
mechanism behind speculative economies of copied minds.

## Outlook

The near-term work that would actually bear on the question is unglamorous: automated volume
microscopy at higher throughput, better segmentation, molecular labelling of synapse types, and
above all a successful emulation of a small animal whose behaviour can be checked. A fly, whose
complete connectome now exists and whose behaviour is well characterized, is the obvious test case
and the field's clearest near-term falsifier. If a fly connectome plus reasonable biophysical
assumptions cannot be made to fly, the scale-separation assumption is in trouble; if it can, the
argument moves from principle to engineering. As of 2026 nobody has claimed to have done it.

## See also

- [[mind-uploading]]
- [[connectomics]]
- [[substrate-independence]]
- [[brain-preservation]]
- [[personal-identity-and-continuity]]
- [[machine-consciousness]]
- [[neural-correlates-of-consciousness]]
- [[digital-immortality]]

## References

[^sb2008]: `report` Sandberg, A. and Bostrom, N. "Whole Brain Emulation: A Roadmap." Technical Report 2008-3, Future of Humanity Institute, University of Oxford, 2008. {An institute technical report, not a peer-reviewed paper; its requirement estimates are conditional on an assumed level of detail.}

[^white1986]: `paper` White, J. G., Southgate, E., Thomson, J. N. and Brenner, S. "The structure of the nervous system of the nematode *Caenorhabditis elegans*." *Philosophical Transactions of the Royal Society B*, 1986.

[^markram2015]: `paper` Markram, H. et al. "Reconstruction and simulation of neocortical microcircuitry." *Cell*, 2015.

[^marder2012]: `paper` Marder, E. "Neuromodulation of neuronal circuits: back to the future." *Neuron*, 2012.

[^miller2015]: `news` Miller, K. D. "Will You Ever Be Able to Upload Your Brain?" *The New York Times*, 2015. {A newspaper opinion piece by a computational neuroscientist, arguing a position rather than reporting a result.}

[^azevedo2009]: `paper` Azevedo, F. A. C. et al. "Equal numbers of neuronal and nonneuronal cells make the human brain an isometrically scaled-up primate brain." *Journal of Comparative Neurology*, 2009.

[^dorkenwald2024]: `paper` Dorkenwald, S. et al. "Neuronal wiring diagram of an adult brain." *Nature*, 2024. {A structural reconstruction of one fly brain with predicted neurotransmitters; it carries no physiological recordings from the animal that was scanned.}

[^moravec1988]: `book` Moravec, H. *Mind Children: The Future of Robot and Human Intelligence*. Harvard University Press, 1988.

==============================================================================
ARTICLE: xenobots
TITLE:   Xenobots
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/xenobots
SOURCE:  https://futurehumanwiki.com/raw/xenobots
==============================================================================

---
title: "Xenobots"
slug: "xenobots"
type: "technology"
status: "experimental"
horizon: "2040s"
trl: 3
categories: ["bodies"]
tags: ["synthetic biology", "morphogenesis", "biorobotics", "regeneration", "self-assembly", "michael levin"]
summary: "Sub-millimetre motile constructs assembled from frog embryonic cells and shaped by an evolutionary algorithm, classed as neither robots nor organisms in any settled sense."
updated: "2026-07-27"
humanEvidence: "No xenobot has been placed in an animal or a person; the human-cell version, anthrobots, has been tested only in culture dishes, including on sheets of cultured human neurons."
access: "Not a product. Built in a small number of academic laboratories from frog embryos or donor-derived human airway cells; none are sold, licensed, or approved for any use."
reversibility: "reversible"
issues: ["Anthrobot lifespan and culture conditions are not covered.", "The bioelectric-signalling literature behind Levin's framing is summarized rather than covered."]
---

```infobox
{
  "caption": "Engineered living construct",
  "rows": [
    { "label": "Cell source", "value": "Xenopus laevis embryo" },
    { "label": "Size", "value": "Under 1 mm" },
    { "label": "Genome", "value": "Unedited" },
    { "label": "Motility", "value": "Cilia or cardiac contraction" },
    { "label": "Design method", "value": "Evolutionary algorithm" },
    { "label": "Energy source", "value": "Maternal yolk reserves" },
    { "label": "Lifespan", "value": "About 10 days unfed" },
    { "label": "Human-cell version", "value": "Anthrobots, 2023" },
    { "label": "Readiness", "value": "TRL 3 (proof of concept)" }
  ]
}
```

**Xenobots** are motile constructs, under a millimetre across, made by sculpting clumps of
embryonic cells from the African clawed frog *Xenopus laevis* into shapes selected by an
evolutionary algorithm running in a physics simulator. They carry no engineered hardware, no
circuitry, and no edited DNA; motion comes from beating cilia or from contracting embryonic heart
muscle, and the energy comes from yolk the cells carried out of the egg. The first were reported in
2020 by Sam Kriegman, Douglas Blackiston, Michael Levin and Josh Bongard, working between Tufts
University and the University of Vermont.[^kriegman2020] The name joins the genus *Xenopus* to the
back half of *robot*, and that second half has caused most of the trouble since.

## How they are made

The pipeline has two halves that never touch. In the first, an evolutionary algorithm searches a
space of soft-body designs built from voxels of two kinds — passive tissue and contractile tissue —
scoring each candidate in simulation on a task such as travelling in a straight line or pushing a
particle, then filtering the survivors for robustness, since a design that collapses under small
perturbations will not survive transfer to tissue.

In the second half a person builds the winning shape by hand. Animal cap tissue is dissected from
blastula-stage frog embryos, dissociated, and left to reaggregate into a coherent ball.
Microsurgical forceps and a cautery electrode then carve that ball to match the design, and in the
original constructs a graft of cardiac progenitor tissue supplies the contractions that push the
object along. Nothing is loaded into the finished construct afterwards. The design step delivers a
body plan and stops, which is the first respect in which the word "robot" misleads.

A later generation dispensed with the surgery and the heart muscle. Animal cap cells left to
themselves form spheroids that place their cilia on the outside — the same cilia that would
otherwise sweep mucus across frog skin — and swim. These constructs repair themselves within
minutes of being lacerated, mill through fields of loose particles, and can be made to record an
exposure to light by expressing a photoconvertible fluorescent protein, injected into the embryo as
messenger RNA rather than written into the genome.[^blackiston2021] Nothing here involves
[[crispr-cas9|genome editing]]: the cells carry the frog's ordinary genome in an unfamiliar setting.

Lifespan is set by the maternally loaded yolk platelets the cells inherited. Unfed, the ciliated
constructs move for roughly ten days and then break down; in frog culture medium they have been kept
alive for months, since the cells retain ordinary metabolic machinery. None can find its own food.

## Development history

```timeline
[
  { "year": "1955", "title": "Dissociated amphibian cells reaggregate", "text": "Philip Townes and Johannes Holtfreter show that embryonic amphibian cells broken apart in culture reassemble and sort themselves by tissue type, establishing that architecture can arise from cells with no embryo around them." },
  { "year": "1966", "title": "Kinematic self-reproduction formalized", "text": "John von Neumann's posthumously edited work distinguishes a machine that copies a description of itself from one that merely assembles copies out of parts lying around it." },
  { "year": "2020", "title": "First reconfigurable organisms", "text": "Kriegman, Blackiston, Levin and Bongard report simulated designs built from Xenopus laevis skin and cardiac progenitor cells that move, push objects, and heal after being cut." },
  { "year": "2021", "title": "Ciliated, self-assembling version", "text": "Blackiston and colleagues report constructs that self-organize from skin cells alone and swim on their own cilia, removing the microsurgery step." },
  { "year": "2021", "title": "Kinematic self-replication reported", "text": "Parent constructs are shown to sweep loose stem cells into piles that mature into a further generation of swimmers, in a dish continuously supplied with fresh cells." },
  { "year": "2023", "title": "Anthrobots", "text": "Gumuskaya and colleagues in Levin's laboratory report motile spheroids that self-assemble from adult human tracheal epithelial cells." }
]
```

## Kinematic self-replication

The 2021 report that xenobots "self-replicate" is the most misread result in the
field.[^kriegman2021] What was observed is mechanical. A parent construct swimming
through a dish strewn with dissociated stem cells pushes them into heaps; a heap that reaches a
threshold size compacts, grows cilia, and swims off as a new construct. An evolutionary search found
a C-shaped parent, widely described in coverage as AI-designed, that gathers cells more efficiently
than a sphere and sustains the cycle for more rounds.

Nothing is copied. No genome is read or written, no template is used, and the offspring are
spheroids that do not inherit the parent's shape, so the very feature that made a good replicator is
lost in the first generation. The dish must be restocked with fresh cells by an experimenter; take
that away and the process stops immediately. This is replication in von Neumann's kinematic sense —
assembly of a copy from parts already present — rather than in the biological sense of heredity with
variation.[^vonneumann1966]

> [!caution] Not replication in the biological sense
> Constructs that genuinely propagate a design through a population, such as a [[gene-drive]], do it
> by copying DNA — the mechanism absent here. Without heredity there is no variation for selection to
> act on, so the process cannot evolve, and the runaway scenarios of the [[grey-goo]] literature
> require a closure over materials and energy that nothing in this work approaches.

## Anthrobots and human cells

The same laboratory has reported an equivalent construct made from adult human cells. Anthrobots
self-assemble from donated tracheal epithelial cells, which under the right culture conditions turn
their cilia outward and become motile spheroids ranging from tens to a few hundred micrometres
across.[^gumuskaya2023] They are made without reprogramming, unlike constructs built on
[[induced-pluripotent-stem-cells]], and without an external scaffold, unlike [[organ-bioprinting]]
or [[decellularized-scaffolds]]. Nothing is imposed on the cells except the medium they sit in.

In the reported experiment, clusters of anthrobots placed across a scratch in a monolayer of
cultured human neurons were followed by regrowth across the gap that untreated controls did not
show. That is a result in a dish. No mechanism has been established and no animal has been treated,
and the distance between closing a scratch in culture and repairing tissue in a person is the one
most of regenerative medicine has never crossed.

## The morphogenesis claim

The scientifically interesting argument attached to this work is not about robotics at all. Levin
holds that the genome specifies cellular hardware rather than anatomy directly, that cells and
tissues coordinate through bioelectric signalling toward target morphologies, and that a group of
cells removed from the signalling context of an embryo will settle into a different but stable body
plan.[^levin2021] On that reading a xenobot is not a machine made of tissue but a frog cell
collective exercising an option its genome always contained. Levin and Daniel Dennett have pressed a
stronger version of the framing, in which goal-directedness is a property found at every scale of
biology rather than only in nervous systems.[^dennett2020] Supporting work outside the xenobot
programme comes from planarian flatworms: after a brief disruption of gap-junction signalling a
minority regenerate two heads, and those animals go on yielding two-headed worms through later
amputations in plain water, with no change to their DNA.[^durant2017] The result bears on
[[limb-regeneration]] and on what a genome does and does not determine.

> [!debate] How much does the construct actually show
> Critics accept the observations and dispute the interpretation. Dissociated amphibian cells were
> known to reaggregate and sort by tissue type in the 1950s, and ciliated epithelium swims because
> that is what ciliated epithelium does.[^townes1955] On the sceptical reading, a xenobot is a motile
> explant with a haircut, and the language of body plans, goals and reconfigurable organisms is
> doing work the data do not require. The reply from Levin's side is that a stable, reproducible,
> functional morphology absent from the frog's life cycle is precisely what needs explaining. The
> disagreement is about the explanatory frame, not the images.

## Limitations

Two misreadings are near-universal in coverage. These are sub-millimetre objects, orders of
magnitude larger than the constructions in the [[medical-nanorobots]] tradition and comparable to
the larger devices in [[microrobots-in-medicine]]; and they are not machines, containing no sensor,
no actuator, no controller, and no program. The evolutionary search runs on a computer before any
cells are touched, and the resulting object cannot be reprogrammed, updated, or told what to do.

The transfer from simulation to tissue is loose. The simulator treats cells as voxels of passive or
contractile material, omitting adhesion, signalling, and the tendency of tissue to remodel itself,
so built constructs approximate their designs rather than instantiate them. Batch variation is
substantial, in the manner familiar from [[organoids]], and nothing can be steered: motion is
directed only in the statistical sense that a shape biases where a swimmer tends to go.

No therapeutic application exists. Proposals in circulation — clearing microplastics from water,
scraping plaque from vessels, carrying a payload in the manner of [[targeted-drug-delivery]] — are
speculative in the strict sense that none has been attempted in an animal. A construct that runs out
of fuel in days, cannot be located once released, and consists of foreign frog cells has no obvious
route into a body.

## Risk and governance

The mainstream assessment of biosafety risk is that it is low, for a specific reason rather than a
reassuring one: xenobots cannot feed themselves, cannot reproduce without cells supplied to them,
and are built from vertebrate cells that die outside a controlled medium. That places them below the
containment concerns motivating the safeguards in [[recoded-organisms]], and far below those about
[[mirror-life]]. Invoking the [[precautionary-principle]] here is an argument about a research
programme whose future forms cannot be specified, not about the constructs already made.

The harder questions are definitional. Whether these things are organisms, and what would be owed to
them if they were, meets the same absence of an agreed test that troubles
[[machine-consciousness]] — with less urgency than for cerebral organoids, since a xenobot contains
no neurons at all. Regulatory categories fit badly: a construct of unmodified animal cells is not a
genetically modified organism, not a medical device, and not a laboratory animal, the same gap
[[synthetic-embryos]] have opened from the other direction. The technique is also cheap and needs no
gene synthesis, so it sits outside most of the machinery built for [[dual-use-research]].

## Outlook

The near-term value of the work is as an experimental system rather than a product. It offers a way
to ask what a group of cells builds when the embryo's instructions are removed, on a timescale of
days and at a cost that permits many replicates, a question upstream of [[tissue-engineering]] and
of every attempt at [[lab-grown-organs]]. Whether the constructs themselves ever do useful work is a
separate and longer bet: the missing pieces are a food supply, control, and a reason to prefer a
living object to an engineered one.

The claim that would matter most if it held is the one hardest to test. If anatomy is a
reconfigurable target rather than a fixed genomic output, then instructing cells at the level of
pattern, instead of editing the genes beneath it, becomes a legitimate strategy for regeneration.
Frog cells arranged into something the frog never makes demonstrate that the option exists. They are
not yet evidence that anyone knows how to specify which option is taken.

## See also

- [[organoids]]
- [[synthetic-embryos]]
- [[limb-regeneration]]
- [[tissue-engineering]]
- [[microrobots-in-medicine]]
- [[medical-nanorobots]]
- [[machine-consciousness]]
- [[mirror-life]]

## References

[^kriegman2020]: `paper` Kriegman, S., Blackiston, D., Levin, M., Bongard, J. "A scalable pipeline for designing reconfigurable organisms." *PNAS*, 2020. {The design and build stages are decoupled: the evolutionary search runs entirely in simulation and the constructs are assembled by hand.}
[^blackiston2021]: `paper` Blackiston, D. et al. "A cellular platform for the development of synthetic living machines." *Science Robotics*, 2021. {Reports the ciliated, self-assembling constructs, their roughly ten-day unfed lifespan, and the injected-mRNA light reporter.}
[^kriegman2021]: `paper` Kriegman, S. et al. "Kinematic self-replication in reconfigurable organisms." *PNAS*, 2021. {Replication here is mechanical piling of loose cells supplied by the experimenter; no genome is copied and offspring do not inherit the parent shape.}
[^vonneumann1966]: `book` von Neumann, J. *Theory of Self-Reproducing Automata*, edited and completed by Arthur W. Burks, 1966.
[^gumuskaya2023]: `paper` Gumuskaya, G. et al. "Motile Living Biobots Self-Construct from Adult Human Somatic Progenitor Seed Cells." *Advanced Science*, 2023. {The neural result is a scratch assay in cultured human cells; the paper does not test anthrobots in an animal.}
[^levin2021]: `paper` Levin, M. "Bioelectric signaling: Reprogrammable circuits underlying embryogenesis, regeneration, and cancer." *Cell*, 2021.
[^dennett2020]: `statement` Levin, M. and Dennett, D. "Cognition all the way down." *Aeon*, 2020. {An argued position by two of the framework's proponents, not a report of experimental results.}
[^townes1955]: `paper` Townes, P. L. and Holtfreter, J. "Directed movements and selective adhesion of embryonic amphibian cells." *Journal of Experimental Zoology*, 1955.
[^durant2017]: `paper` Durant, F. et al. "Long-Term, Stochastic Editing of Regenerative Anatomy via Targeting Endogenous Bioelectric Gradients." *Biophysical Journal*, 2017. {A minority of treated planarian fragments became two-headed; the altered form then persisted through further amputations.}

==============================================================================
ARTICLE: xenotransplantation
TITLE:   Xenotransplantation
PORTAL:  Regenerative Bodies
URL:     https://futurehumanwiki.com/wiki/xenotransplantation
SOURCE:  https://futurehumanwiki.com/raw/xenotransplantation
==============================================================================

---
title: "Xenotransplantation"
slug: "xenotransplantation"
type: "technology"
status: "experimental"
horizon: "2030s"
trl: 6
categories: ["bodies", "genetics"]
tags: ["transplantation", "organ shortage", "genome editing", "immunology", "pigs", "clinical trials"]
summary: "The transplantation of organs between species, now pursued with genetically engineered pigs whose antigens and clotting proteins have been altered to survive in a human recipient."
updated: "2026-07-27"
humanEvidence: "Two patients received gene-edited pig hearts and several others pig kidneys under compassionate use; every graft has failed or been removed within months, and formal kidney trials were cleared in 2025."
access: "No marketed product and no price: recipients so far have been patients ineligible for a human organ, treated under compassionate-use authorisation, with the first kidney trials cleared in 2025."
reversibility: "difficult"
issues: ["Waiting-list, daily-deaths and 130-day graft figures are stated without citations."]
---

```infobox
{
  "caption": "Transplantation technology",
  "rows": [
    { "label": "Donor species", "value": "Domestic pig" },
    { "label": "Core knockouts", "value": "GGTA1, CMAH, B4GALNT2" },
    { "label": "Human transgenes", "value": "Complement and clotting regulators" },
    { "label": "Edits per animal", "value": "10 to 69" },
    { "label": "First gene-edited pig heart", "value": "David Bennett, 2022" },
    { "label": "Longest graft to date", "value": "About 130 days" },
    { "label": "Status", "value": "Early clinical trials" }
  ]
}
```

**Xenotransplantation** is the transplantation of living cells, tissues or organs between species. In its modern form it means moving a genetically engineered pig organ into a human, using edits that remove the sugars human antibodies attack and add human proteins that control complement activation and blood clotting. Between 2021 and 2025 the field moved from brain-dead research recipients to living patients under compassionate-use authorisation, and then to formally cleared clinical trials. Every graft so far has failed or been removed within months.

## Why pigs

The driver is arithmetic. More than a hundred thousand people are on the United States transplant waiting list at any time, the great majority waiting for a kidney, and the US Health Resources and Services Administration puts the number dying each day while waiting at around seventeen. Waiting lists understate demand, because patients too sick or too old to qualify are never listed. The structural causes are set out in [[organ-shortage]], and no combination of donation reform, [[artificial-heart|mechanical support]] and improved organ preservation closes the gap.

Non-human primates are immunologically closer but are slow-breeding, expensive, ethically fraught and carry viruses that cross to humans readily. Pigs breed fast, reach human organ size in months, tolerate genetic modification, and are already farmed at scale. Their organs are anatomically and physiologically similar enough that the remaining mismatches are addressable one gene at a time, which is the premise of the whole enterprise.

## The immunological barrier

Pig organs fail in a human in a defined sequence, and each stage has a corresponding engineering answer.

**Hyperacute rejection.** Humans have high-titre preformed antibodies against galactose-α-1,3-galactose, a sugar humans and other Old World primates do not make. Within minutes these antibodies fix complement and destroy the graft. Knocking out the GGTA1 gene that installs the sugar eliminates this, and was the first modification made, in the early 2000s. Two further glycans, produced by CMAH and B4GALNT2, are removed for the same reason, giving the "triple-knockout" pig.

**Acute humoral rejection.** Non-glycan antibodies and residual complement activity damage the graft over days to weeks. Human complement regulatory proteins CD46 and CD55 are added as transgenes to hold this in check, alongside CD47 to reduce macrophage attack and heme oxygenase-1 to limit oxidative injury.

**Coagulation dysregulation.** Pig thrombomodulin activates human protein C poorly, so the graft endothelium fails to restrain clotting and thrombotic microangiopathy develops. Human thrombomodulin and endothelial protein C receptor are added to correct the molecular mismatch. This is the barrier least amenable to conventional immunosuppression and the one that most distinguishes xenografts from allografts.

**Physiological mismatch.** Pig organs grow to pig proportions. Knocking out the growth hormone receptor limits this. Other mismatches, in erythropoietin signalling and in the renin-angiotensin system, are managed pharmacologically rather than genetically.

Two engineering strategies have emerged. Revivicor's animals carry ten modifications, four knockouts and six human transgenes. eGenesis, founded out of [[george-church]]'s laboratory, adds inactivation of the porcine endogenous retrovirus copies scattered through the pig genome, giving sixty-nine total edits. The PERV work grew out of demonstrations that dozens of retroviral copies could be knocked out at once in pig cells and then in live piglets, an early proof that [[crispr-cas9]] could be used at the edit counts previously seen only in the [[recoded-organisms|genome-recoding]] work on bacteria.[^yang2015][^niu2017] The same multiplex capability underlies the proxy-species programmes described in [[de-extinction]], and pigs are the animal in which it has been pushed furthest toward a product.

## Development history

```timeline
[
  { "year": "1963–1964", "title": "Primate era", "text": "Keith Reemtsma transplants chimpanzee kidneys, one recipient surviving nine months, and James Hardy places a chimpanzee heart in a dying patient. Immunology of the day cannot sustain the grafts." },
  { "year": "1984", "title": "Baby Fae", "text": "An infant with hypoplastic left heart syndrome receives a baboon heart at Loma Linda and lives 21 days. The case prompts lasting debate about consent and about experimental surgery in neonates." },
  { "year": "2003", "title": "Alpha-gal knockout pigs", "text": "Pigs lacking GGTA1 are produced, removing the antigen responsible for hyperacute rejection and making the modern programme possible." },
  { "year": "2018", "title": "Preclinical durability", "text": "A Munich group reports baboons surviving for months on life-supporting pig hearts using non-ischaemic preservation and blood-pressure control, establishing that the model can work." },
  { "year": "2021–2023", "title": "Decedent studies", "text": "Teams at NYU and Alabama transplant pig kidneys into brain-dead recipients maintained on ventilators, one for 61 days, to study early function without risk to a patient." },
  { "year": "2022–2025", "title": "Living recipients", "text": "Two pig hearts and several pig kidneys are transplanted under compassionate-use rules, with survival measured in weeks to months, and US regulators clear the first formal kidney trials in 2025." }
]
```

## Clinical experience

The preclinical foundation is stronger than the clinical record. Pig hearts have supported baboons for periods measured in months rather than days once preservation and perioperative management were optimised.[^langin2018] Humanised pig kidneys have supported cynomolgus monkeys for many months, with individual animals surviving beyond a year.[^anand2023]

In humans the record is short. David Bennett Sr. received a ten-gene pig heart at the University of Maryland in January 2022 and survived about two months; the graft showed diffuse damage, and porcine cytomegalovirus DNA was detected in his circulation, raising the possibility that an incompletely screened donor virus contributed.[^griffith2022] A second Maryland heart recipient in 2023 survived roughly six weeks. Kidney recipients have done somewhat better: a Massachusetts General Hospital patient received a sixty-nine-edit kidney in March 2024, and an Alabama patient's kidney at NYU functioned about four months before it was removed in 2025, the longest such graft reported. Earlier work in brain-dead recipients had shown that a pig kidney makes urine and clears creatinine in a human circulation.[^montgomery2022]

Every recipient so far has been ineligible for a human organ, which is what made the risk-benefit calculation defensible and also what makes the results hard to interpret. These are severely ill patients on experimental immunosuppression, often including costimulation blockade agents not approved for transplantation.

```compare
{
  "columns": ["Pig xenograft", "Deceased-donor allograft", "Bioprinted organ"],
  "rows": [
    { "label": "Supply", "values": ["In principle unlimited", "Severely constrained", "None yet"] },
    { "label": "Immunosuppression", "values": ["Heavier, experimental", "Standard regimens", "Potentially none if autologous"] },
    { "label": "Longest human graft", "values": ["About 130 days", "Decades", "No solid organ implanted"] },
    { "label": "Cross-species infection risk", "values": ["Present, monitored", "None", "None"] },
    { "label": "Readiness", "values": ["Early trials", "Routine", "Laboratory"] }
  ]
}
```

## Infection, welfare and consent

Cross-species infection is the risk that extends beyond the patient. Porcine endogenous retroviruses are integrated in the pig genome and cannot be bred out, which is why eGenesis inactivates them; a long-term follow-up of patients exposed to living pig tissue in extracorporeal treatments found no evidence of transmission, which is reassuring but not decisive for whole-organ grafts under immunosuppression.[^paradis1999] Exogenous viruses have proved the more immediate problem, as the porcine cytomegalovirus finding in the first heart recipient showed. Donor animals are therefore raised in designated pathogen-free facilities with barrier housing and repeated screening, and recipients are enrolled in indefinite surveillance.

That surveillance carries an unusual consent structure. A xenograft recipient accepts lifelong monitoring, restrictions on blood and tissue donation, and in some protocols obligations extending to close contacts, on grounds of public health rather than personal benefit. Regulators have generally required this, and it is one of the few areas of medicine where a patient's autonomy is curtailed to protect third parties.

Animal welfare is the standing objection. Donor animals are produced by the nuclear-transfer methods described in [[human-cloning]], raised in isolation, and killed on a surgical schedule. Defenders point out that pigs are farmed for food in vastly greater numbers under less controlled conditions; critics reply that instrumental use for organs invites a scale of purpose-bred confinement that food production does not justify by analogy. Religious authorities in traditions that restrict pork have generally permitted xenotransplantation where life is at stake.

> [!debate] Bridge or destination
> One reading treats pig organs as a bridge, keeping patients alive until [[lab-grown-organs]], [[organ-bioprinting]] or the reseeded [[decellularized-scaffolds|matrix scaffolds]] of [[tissue-engineering]] mature. Another treats them as the destination, on the argument that growing a vascularised human kidney from [[induced-pluripotent-stem-cells]] remains far harder than editing a pig, and that a manufacturable animal organ is the realistic supply solution for this century.

## Outlook

The decisive data should come from the cleared kidney trials, where recipients are less critically ill than the compassionate-use cohort and where survival can be compared against dialysis rather than against imminent death. The endpoint that matters is graft function at one year. Nothing in the human record yet approaches it, and the preclinical primate data suggest it is attainable.

Three questions remain open. Whether the coagulation mismatch is fully corrected by the current transgene set, or merely delayed, will only be answered by grafts that survive long enough to develop chronic injury. Whether immunosuppression can be reduced to allograft-like levels, rather than the intensive experimental regimens used so far, determines whether the treatment is tolerable for patients who have a dialysis alternative. And whether a manufactured organ produced by a small number of firms is distributed any more equitably than donated ones is a question that the technology does not answer and that [[access-and-inequality]] raises for every expensive intervention.

## See also

- [[organ-shortage]]
- [[lab-grown-organs]]
- [[organ-bioprinting]]
- [[decellularized-scaffolds]]
- [[crispr-cas9]]
- [[de-extinction]]
- [[george-church]]
- [[artificial-heart]]

## References

[^yang2015]: `paper` Yang, L. et al. "Genome-wide inactivation of porcine endogenous retroviruses (PERVs)." *Science*, 2015.
[^niu2017]: `paper` Niu, D. et al. "Inactivation of porcine endogenous retrovirus in pigs using CRISPR-Cas9." *Science*, 2017.
[^langin2018]: `paper` Längin, M. et al. "Consistent success in life-supporting porcine cardiac xenotransplantation." *Nature*, 2018.
[^anand2023]: `paper` Anand, R. P. et al. "Design and testing of a humanized porcine donor for xenotransplantation." *Nature*, 2023.
[^griffith2022]: `paper` Griffith, B. P. et al. "Genetically Modified Porcine-to-Human Cardiac Xenotransplantation." *New England Journal of Medicine*, 2022. {A single-patient case report; the porcine cytomegalovirus finding emerged afterwards and its contribution to the graft failure is unestablished.}
[^montgomery2022]: `paper` Montgomery, R. A. et al. "Results of Two Cases of Pig-to-Human Kidney Xenotransplantation." *New England Journal of Medicine*, 2022. {Both recipients were brain-dead and observed on ventilators for a short period, so the result covers immediate function only.}
[^paradis1999]: `paper` Paradis, K. et al. "Search for cross-species transmission of porcine endogenous retrovirus in patients treated with living pig tissue." *Science*, 1999.

==============================================================================
ARTICLE: xprize-healthspan
TITLE:   XPRIZE Healthspan
PORTAL:  Organizations
URL:     https://futurehumanwiki.com/wiki/xprize-healthspan
SOURCE:  https://futurehumanwiki.com/raw/xprize-healthspan
==============================================================================

---
title: "XPRIZE Healthspan"
slug: "xprize-healthspan"
type: "organization"
status: "emerging"
horizon: "late 2020s"
categories: ["organizations", "longevity"]
tags: ["healthspan", "prizes", "clinical trials", "aging", "functional endpoints", "geroscience"]
summary: "A $101 million incentive competition launched in 2023 that rewards a therapy restoring at least a decade of muscle, cognitive and immune function in older adults."
updated: "2026-07-28"
issues: ["Registration and semifinalist counts in the timeline carry no source."]
---

```infobox
{
  "caption": "Incentive prize competition",
  "rows": [
    { "label": "Launched", "value": "November 2023" },
    { "label": "Purse", "value": "$101 million" },
    { "label": "Operator", "value": "XPRIZE Foundation" },
    { "label": "Principal sponsors", "value": "Hevolution Foundation; Chip Wilson" },
    { "label": "Target population", "value": "Adults aged 50–80" },
    { "label": "Domains scored", "value": "Muscle, cognition, immunity" },
    { "label": "Duration", "value": "Seven years" }
  ]
}
```

**XPRIZE Healthspan** is a seven-year, $101 million incentive competition launched in November 2023 that offers its grand prize to a team demonstrating a therapy that restores at least a decade of function in three domains — muscle, cognition and immunity — in adults aged 50 to 80, using a treatment course of no more than a year. Its significance to the field lies less in the money than in the endpoint: it is the most detailed public attempt to define what a successful [[healthspan]] intervention would have to show.

```keyfacts
[
  { "value": "$101m", "label": "Total purse", "note": "grand and tiered awards plus milestone payments" },
  { "value": "20 years", "label": "Function restored for the top award", "note": "with lower tiers for 15 and 10 years" },
  { "value": "3", "label": "Functional domains scored", "note": "muscle, cognition and immune function, measured together" }
]
```

## Overview

Incentive prizes are used where a goal is well defined, a solution is plausible, and no market pays for the intermediate steps. The XPRIZE Foundation, which ran the Ansari prize for suborbital spaceflight and later competitions in genomics and carbon removal, applies the model by specifying an outcome and letting teams choose the route.

The Healthspan competition's design responds to a specific problem. No regulator recognizes aging as an indication, so no company can run a registrational trial against it, and the field has no qualified [[aging-biomarkers|surrogate endpoint]] that would let a shorter trial substitute for decades of follow-up. The prize sidesteps both by scoring measured function rather than mortality or biomarkers: if a 70-year-old performs across three domains as a 60-year-old would, the intervention has restored ten years of function by the competition's definition, whatever it did to any clock.

## Structure

The competition was announced with a $101 million purse funded principally by the Hevolution Foundation, a Saudi-backed nonprofit that has become one of the largest funders of aging biology, and by Chip Wilson, the founder of Lululemon, who has facioscapulohumeral muscular dystrophy and funds research on it. The published competition guidelines specify the population, the treatment window and the scoring of each domain.[^xprize2023]

The prize is tiered by effect size. The largest award goes to a team restoring twenty years of function; smaller awards go to fifteen and ten years. Milestone payments are made to semifinalist teams during the competition to fund the work, which addresses the standard criticism of prizes — that they reward only the winner and therefore attract only the already-funded.

The judging design constrains what can win. Treatment must be delivered within a defined window of up to one year, so a lifelong regimen does not qualify. Function must improve in all three domains, which excludes single-tissue interventions: a drug that rebuilds muscle without touching cognition or immunity scores zero on two thirds of the assessment. Participants are drawn from a general older population rather than from a disease cohort.

```timeline
[
  { "year": "2023", "title": "Launch", "text": "XPRIZE announces a $101 million, seven-year competition with Hevolution Foundation and Chip Wilson as principal funders." },
  { "year": "2024", "title": "Registration and qualifying", "text": "Several hundred teams from dozens of countries register, spanning drug repurposing, exercise and nutrition protocols, cell therapies and devices." },
  { "year": "2025", "title": "Semifinalists named", "text": "A cohort of semifinalist teams is selected and receives milestone funding to generate preliminary human data." },
  { "year": "Late 2020s", "title": "Finalist trials", "text": "A smaller group of finalists is to run the year-long intervention trials against which the grand prize will be judged." }
]
```

## Why the endpoints matter

The competition's most useful output may be its measurement framework rather than its winner.

**Muscle.** Skeletal muscle mass, strength and power decline steeply with age, and the resulting sarcopenia predicts falls, hospitalization and mortality. Grip strength, gait speed and chair-rise time are validated, cheap, and correlate strongly with outcomes; gait speed alone predicts survival in older adults across large pooled cohorts,[^studenski2011] which makes muscle the least contentious domain. It is also the domain where [[exercise-and-aging|exercise]] produces large effects, setting a demanding comparator for any drug, and where [[myostatin-inhibition]] has repeatedly added mass without adding proportionate strength.
Training adjuncts can subtract as well as add: repeated cold-water immersion immediately after
resistance training reduces the muscle gains that training produces, among the better-replicated
findings on [[heat-and-cold-exposure|heat and cold exposure]].

**Cognition.** Age-related cognitive change is heterogeneous and slow, and the available instruments have practice effects that complicate repeated testing over a year. Distinguishing a genuine restoration of processing speed or executive function from familiarity with the test is a real methodological problem, and the competition's handling of it is watched closely.

**Immunity.** Immune aging encompasses thymic involution, narrowing of the naive T-cell repertoire, accumulation of senescent lymphocytes, and the chronic [[inflammaging|sterile inflammation]] that accompanies them. Vaccine response is the standard functional readout. This is the domain with the weakest consensus on what to measure, and the one where a positive result would be most informative.

Requiring all three at once is the design's sharpest feature. It operationalizes the [[geroscience-hypothesis]] directly: if aging is a shared upstream driver of chronic disease, as the field's founding statement of the position argues,[^kennedy2014] an intervention on that driver should move multiple systems together. An entrant that improves one domain has demonstrated a useful drug; an entrant that improves all three has demonstrated something about aging.

> [!caution] What a win would and would not prove
> Restoring measured function in three domains over one year would be the strongest human evidence any geroprotective intervention has produced. It would not show that lifespan is extended, that mortality falls, or that the effect persists after treatment stops. Function and survival are correlated but not the same, and the competition's timeframe cannot address the second.

## Field of entrants

Registered teams span a wide range of approaches: repurposed drugs including [[rapamycin]] and metformin analogues, [[senolytics]], plasma-derived and [[parabiosis-and-young-blood|blood-factor]] therapies, cell and stem-cell treatments, [[partial-reprogramming|reprogramming]] approaches, [[nad-precursors|NAD-raising]] compounds, hormonal interventions, and structured training and nutrition protocols. Several entrants are academic groups; several are companies for whom the prize functions as non-dilutive funding and validation.

The breadth is itself informative. A competition scored on function rather than mechanism attracts entrants with no shared theory of aging, which means the results will be a comparative test of approaches that have never been run against a common endpoint.

## Criticism

Three objections have been raised.

The first concerns feasibility. No intervention has ever been shown to restore a decade of function in any of the three domains in humans, let alone all three, within a year. Skeptics argue the target is set so far beyond current capability that the prize will go unclaimed and the tiered lower awards will do the real work.

The second concerns measurement. Composite functional scores can be gamed, and a team that optimizes training and testing conditions may produce apparent gains without underlying biological change. Practice effects in cognitive testing and learning effects in physical assessments are well documented, and the competition's protocols have to control for both.

The third concerns funding provenance. Hevolution's association with the Saudi state has drawn objections from researchers who decline funding on those grounds, and the competition's dependence on it is a governance question the foundation has addressed only briefly.

## Outlook

The prize's contribution to the field is likely to be the trial infrastructure and the endpoint definitions rather than a winning therapy. If several teams run year-long, function-scored interventions in older adults under a common protocol, the resulting data would be the closest thing the field has produced to a head-to-head comparison, and it would give regulators something concrete to evaluate when the question of aging as an indication next arises — the same question [[nir-barzilai]]'s [[metformin|TAME trial]] was designed to force.

The unresolved issue is what happens if nothing works. A well-designed, well-funded competition producing no qualifying entrant across seven years would be a substantial negative result, and one the field has so far had little practice interpreting. Its advocates would attribute the failure to the difficulty of the target; its critics would read it as evidence that the [[longevity-escape-velocity|near-term optimism]] surrounding geroscience has not been earned.

## See also

- [[healthspan]]
- [[geroscience-hypothesis]]
- [[aging-biomarkers]]
- [[senolytics]]
- [[exercise-and-aging]]
- [[metformin]]
- [[buck-institute]]
- [[biological-age]]

## References

[^xprize2023]: `statement` XPRIZE Foundation. "XPRIZE Healthspan: Competition Guidelines." XPRIZE Foundation, 2023. {The organiser's own rulebook: it establishes what the competition asks for, and is not evidence about any entrant or result.}
[^kennedy2014]: `paper` Kennedy, B.K., Berger, S.L., Brunet, A. et al. "Geroscience: Linking Aging to Chronic Disease." *Cell*, 2014.
[^studenski2011]: `paper` Studenski, S. et al. "Gait Speed and Survival in Older Adults." *JAMA*, 2011. {A pooled analysis of observational cohorts, so gait speed predicts survival without any evidence that raising it raises survival.}

==============================================================================
ARTICLE: yamanaka-factors
TITLE:   Yamanaka factors
PORTAL:  Genetic Engineering
URL:     https://futurehumanwiki.com/wiki/yamanaka-factors
SOURCE:  https://futurehumanwiki.com/raw/yamanaka-factors
==============================================================================

---
title: "Yamanaka factors"
slug: "yamanaka-factors"
type: "technology"
status: "established"
horizon: "present"
trl: 7
categories: ["genetics", "longevity"]
tags: ["reprogramming", "stem cells", "transcription factors", "epigenetics", "pluripotency"]
summary: "The four transcription factors Oct4, Sox2, Klf4 and c-Myc, which together convert a differentiated cell into an induced pluripotent stem cell."
updated: "2026-07-27"
humanEvidence: "Human cells have been reprogrammed to iPSCs since 2007 and clinical-grade lines are made routinely; no partial-reprogramming therapy using these factors has entered human testing with published results."
access: "A research reagent: the factors are used daily in laboratories and in clinical-grade iPSC manufacturing, and are not available as a treatment in any country."
reversibility: "irreversible"
issues: ["The statement that Sendai virus and episomal plasmids dominate iPSC derivation carries no source."]
---

```infobox
{
  "caption": "Set of reprogramming transcription factors",
  "rows": [
    { "label": "Members", "value": "Oct4, Sox2, Klf4, c-Myc" },
    { "label": "Abbreviation", "value": "OSKM" },
    { "label": "Identified", "value": "2006 (mouse), 2007 (human)" },
    { "label": "Identified by", "value": "Kazutoshi Takahashi and Shinya Yamanaka", "link": "/wiki/shinya-yamanaka" },
    { "label": "Nobel Prize", "value": "2012, with John Gurdon" },
    { "label": "Principal use", "value": "iPSC derivation", "link": "/wiki/induced-pluripotent-stem-cells" },
    { "label": "Readiness", "value": "TRL 7" }
  ]
}
```

**Yamanaka factors** are the four transcription factors, Oct4, Sox2, Klf4 and c-Myc, whose forced expression converts an ordinary differentiated cell into a pluripotent stem cell. Their identification in 2006 collapsed a problem that had required eggs, embryos and nuclear transfer into a defined set of four genes that any laboratory could introduce into skin cells. The same four are now the starting point for [[epigenetic-reprogramming]] as a rejuvenation strategy, where the goal is to apply them briefly enough that the cell is reset without being converted.

```keyfacts
[
  { "value": "4", "label": "Factors in the set", "note": "narrowed from 24 candidate genes by elimination" },
  { "value": "2006", "label": "Mouse iPSCs reported", "note": "Takahashi and Yamanaka, Cell" },
  { "value": "2012", "label": "Nobel Prize in Physiology or Medicine", "note": "shared with John Gurdon for nuclear reprogramming" }
]
```

## What each factor does

Oct4, encoded by *POU5F1*, is the core pluripotency regulator; it is expressed in the early embryo and in embryonic stem cells, and no reprogramming cocktail has reliably dispensed with it or a close functional substitute. Sox2 binds cooperatively with Oct4 at composite motifs and helps activate the pluripotency network. Klf4 supports self-renewal, suppresses apoptosis in cells being pushed out of their identity, and cooperates with the other two at enhancers. c-Myc is not a pluripotency factor at all: it is a broadly acting driver of proliferation and chromatin opening, and it raises reprogramming efficiency largely by making the genome more permissive.

Two of the four, Klf4 and c-Myc, are proto-oncogenes. That fact shapes everything downstream, from tumour incidence in early chimeric mice to the design of every proposed therapeutic use.

## How reprogramming proceeds

Introducing the factors does not flip a switch. Most treated cells never reprogramme; success rates in the original protocols were a small fraction of one per cent, and even optimised modern methods convert a minority of cells. The cells that do succeed pass through a stochastic early phase, including a mesenchymal-to-epithelial transition in fibroblasts, followed by a more deterministic late phase in which the endogenous pluripotency network activates and the exogenous factors become dispensable.

Along the way the epigenome is rewritten. DNA methylation patterns that encoded the somatic identity are erased and replaced, X-chromosome inactivation is reversed in female cells, and telomeres are re-extended by reactivated telomerase, a point of contact with [[telomeres-and-telomerase|telomere biology]]. Age-associated methylation is erased too: a fully reprogrammed cell from an eighty-year-old donor reads as embryonic on an [[epigenetic-clock|epigenetic clock]], and its estimated [[biological-age]] by that measure is effectively zero, an observation [[steve-horvath]] reported when he introduced the first multi-tissue clock. That observation is the origin of the rejuvenation programme, and also its central puzzle, because the same process destroys the cell's usefulness as a liver or retinal cell. Cells caught partway also shed markers of [[cellular-senescence]], which is one reason the effect is hard to attribute to any single mechanism.

## Development history

The precondition was John Gurdon's demonstration in the early 1960s that a nucleus from a differentiated frog intestinal cell, transplanted into an enucleated egg, could support development to a tadpole. Differentiation therefore did not delete information; something in the egg cytoplasm could reset it. Somatic cell nuclear transfer, later used to produce Dolly the sheep and discussed under [[human-cloning]], worked but required eggs and gave no account of the responsible molecules.

Kazutoshi Takahashi and [[shinya-yamanaka]] approached the problem by assembling a list of two dozen genes enriched in embryonic stem cells, introducing all of them into mouse fibroblasts by retrovirus, and then removing them one at a time to find the minimal sufficient set. Four remained.[^takahashi2006] The following year the group repeated the result with human fibroblasts, and James Thomson's laboratory independently reported human iPSCs using a partly different quartet, OCT4, SOX2, NANOG and LIN28, showing that the specific four were sufficient rather than uniquely necessary.[^takahashi2007][^yu2007] The 2012 Nobel Prize in Physiology or Medicine went to Yamanaka and Gurdon.

## Variants and reduced sets

Because c-Myc is the most oncogenic member, its removal was pursued immediately. Reprogramming with Oct4, Sox2 and Klf4 alone works at lower efficiency and produced chimeric mice with fewer tumours.[^nakagawa2008] The three-factor combination, written OSK, is now the default in rejuvenation research, including the optic nerve work central to [[partial-reprogramming]].

Other substitutions have been mapped extensively. Nanog, Lin28, Glis1, Esrrb and various small molecules can replace or supplement individual factors. Chemical reprogramming dispenses with transgenes entirely: cocktails of small molecules produced mouse pluripotent cells in 2013 and human ones in 2022, at the cost of long, finicky protocols.[^guan2022] A parallel line of work skips pluripotency altogether and uses lineage-specific transcription factors to convert one somatic cell type directly into another, a strategy that avoids the tumour risk by never passing through a pluripotent state. Companies such as [[newlimit]] apply the same logic to aging, screening transcription-factor combinations for youthful gene expression within a fixed cell identity rather than for pluripotency. [[retro-biosciences]] has instead worked on the canonical factors themselves, reporting computationally designed variants of SOX2 and KLF4 intended to raise reprogramming efficiency; that claim was announced through press coverage rather than a peer-reviewed paper and has not been independently replicated. A third approach, [[epigenome-editing]], targets specific loci with catalytically dead Cas9 fused to methylation-writing or -erasing domains, trading the breadth of transcription-factor reprogramming for site-level control.

## Delivery

The original retroviral vectors integrate into the genome, which makes derived lines unsuitable for clinical use. The field moved through adenovirus, piggyBac transposons, episomal plasmids, Sendai virus, and synthetic modified mRNA, the last of which leaves no genetic footprint at all.[^warren2010] Sendai virus and episomal plasmids dominate routine iPSC derivation as of 2026.

For in vivo use the constraints invert. Reaching cells inside a living animal requires [[aav-vectors|AAV]] or [[lipid-nanoparticles|lipid nanoparticles]], and the requirement shifts from "leave no trace in a cultured line" to "stop expressing on command". Doxycycline-inducible transgenes provide that control in laboratory mice and have no clinical equivalent. This is the practical reason that no partial-reprogramming therapy has entered human testing with published results as of 2026, and it is a specific instance of the delivery problem that constrains [[somatic-gene-therapy]] as a whole.

> [!note] Naming
> "Yamanaka factors" refers to the specific four-gene set, not to reprogramming in general. Papers that use OSK, or that substitute Nanog or Glis1, are often described loosely with the same term. The distinction matters when reading claims about safety, since most of the tumour data concerns constructs containing c-Myc.

## Limitations

Reprogramming is inefficient, slow, and variable between donors and cell types. Derived lines accumulate copy-number changes and point mutations, some pre-existing in the donor cell and clonally amplified, some acquired in culture; screening for them is a standard and expensive part of any clinical-grade [[induced-pluripotent-stem-cells|iPSC]] workflow.

The factors also do not restore everything that aging removes. Reprogramming resets methylation and telomere length but does nothing about somatic mutations, which are carried forward into every daughter cell. An iPSC line from an old donor is epigenetically young and genomically old. Whether that matters for [[gene-therapy-for-aging|rejuvenation applications]] depends on how much of aging is genomic, an unsettled question at the core of the [[hallmarks-of-aging|hallmarks framework]].

## Outlook

As a laboratory reagent the four factors are settled technology, used daily to make patient-derived cells for disease modelling, drug screening, [[organoids]], [[tissue-engineering]] and cell therapy. Well-funded programmes at [[altos-labs]] and elsewhere are betting that the same molecules can be turned into medicines rather than reagents.

As a therapeutic input they remain unproven. The open question is not whether the factors can reset a cell but whether the reset can be stopped at a chosen point, in a chosen tissue, in a living person, without producing a tumour or a cell that has forgotten its job.

## See also

- [[epigenetic-reprogramming]]
- [[partial-reprogramming]]
- [[induced-pluripotent-stem-cells]]
- [[shinya-yamanaka]]
- [[epigenetic-clock]]
- [[human-cloning]]
- [[newlimit]]

## References

[^takahashi2006]: `paper` Takahashi, K. & Yamanaka, S. "Induction of Pluripotent Stem Cells from Mouse Embryonic and Adult Fibroblast Cultures by Defined Factors." *Cell*, 2006.
[^takahashi2007]: `paper` Takahashi, K. et al. "Induction of Pluripotent Stem Cells from Adult Human Fibroblasts by Defined Factors." *Cell*, 2007.
[^yu2007]: `paper` Yu, J. et al. "Induced Pluripotent Stem Cell Lines Derived from Human Somatic Cells." *Science*, 2007.
[^nakagawa2008]: `paper` Nakagawa, M. et al. "Generation of induced pluripotent stem cells without Myc from mouse and human fibroblasts." *Nature Biotechnology*, 2008. {The tumour comparison was made in chimeric mice, and dropping c-Myc lowered reprogramming efficiency substantially.}
[^warren2010]: `paper` Warren, L. et al. "Highly efficient reprogramming to pluripotency and directed differentiation of human cells with synthetic modified mRNA." *Cell Stem Cell*, 2010. {Human cells in culture, using repeated daily transfections, a protocol burden that limited routine adoption.}
[^guan2022]: `paper` Guan, J. et al. "Chemical reprogramming of human somatic cells to pluripotent stem cells." *Nature*, 2022.