Ray Kurzweil: revision 1 to 2 — Future Human Wiki

Revision 1

Revision 2

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."

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"

updated: "2026-08-23"

issues: ["Hofstadter's often-quoted verdict is paraphrased without a source", "The self-assessment scorecard is described without citing Kurzweil's own review"]

---

```infobox

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,

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:

his own biography now gives his title there as principal researcher and AI visionary.[^rksite]

*The Age of Intelligent Machines* (1990), *The Age of Spiritual Machines* (1999), *The Singularity

His forecasting work runs in parallel: *The Age of Intelligent Machines* (1990), *The Age of

Is Near* (2005), *How to Create a Mind* (2012) and *The Singularity Is Nearer* (2024).

Spiritual Machines* (1999), *The Singularity Is Near* (2005), *How to Create a Mind* (2012) and

*The Singularity Is Nearer* (2024).[^tsin2024]

## The law of accelerating returns

## Predictions and their scorecard

Kurzweil has published self-assessments of his own forecasts, scoring most of his 1999 predictions

Kurzweil has published self-assessments of his own forecasts, scoring the large majority of the

for 2009 as correct or "essentially correct". Independent reviewers have been harsher, noting that

147 predictions he made in 1999 for 2009 as correct or "essentially correct".[^rk2010] 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.

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

often-quoted verdict, from a 2007 interview, is that the good and the indefensible ideas are mixed

that separating them is the reader's problem.

together so thoroughly that separating them is the reader's problem.[^hofstadter2007]

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

capability was always asserted rather than derived. Kurzweil's own dates have not moved: *The

Singularity Is Nearer* restates 2029 and 2045 unchanged.[^tsin2024] See

[[artificial-general-intelligence]] for the definitional fight this exposes.

## Reception and legacy

[^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.}

[^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.}

[^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.

[^rk2010]: `statement` Kurzweil, R. "How My Predictions Are Faring." Essay, October 2010. {Kurzweil's own review of the 147 predictions in The Age of Spiritual Machines; he grades 86% of them entirely or essentially correct, a figure independent reviewers dispute.}

[^hofstadter2007]: `news` "An Interview with Douglas R. Hofstadter." *American Scientist*, 2007.

[^tsin2024]: `book` Kurzweil, R. *The Singularity Is Nearer: When We Merge with AI*. Viking, 2024.

[^rksite]: `statement` "About Ray Kurzweil." thesingularityisnearer.com, the author biography for the 2024 book. {A self-description; it is the source for his current Google title, not an independent report.}