ClaimsPeople

Dario Amodei

Dario Amodei is co-founder and CEO of Anthropic and a former VP of Research at OpenAI, where he led work on GPT-2 and GPT-3.

1,335 claims collected, 337 researched, 97 resolved. Report of 4 November 2026, method 0.1.0+c876030c.

Scores by kind of claim

Each kind of claim is scored on its own, out of 100: the share that held up, with the range it likely falls in. Fewer resolved claims give a wider range.

  • Factual claims

    Statements about the present or the past, checked against the record: did they hold up?

    76 69 82

    High confidence, 76 factual claims resolved. From 14 documents, 2015 to 2026; 19 documents sampled from 30 candidates found.

    Not scored: 51 unresolvable, 42 checkable only against their own records.

    See the factual claims

  • Predictions

    Statements about events outside the person's control, checked once they fell due: did they come true?

    56 42 70

    Medium confidence, 17 predictions resolved. From 8 documents, 2015 to 2026; 19 documents sampled from 30 candidates found.

    Not scored: 103 unresolvable, 1 checkable only against their own records. 39 more fall due later.

    Does not yet allow for how likely each prediction was when made.

    See the predictions

  • Promises

    Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?

    Insufficient record: 4 promises resolved, 5 needed for a score.

    Not scored: 3 unresolvable, 1 not yet due.

    See the promises

All types combined: 72 (likely 66 to 78, high confidence), from 93 claims: 76 factual claims, 17 predictions. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.

Dario Amodei's claims

40 claims

40 claims, sorted by newest said. Choose a claim to read it in full.
VerdictKindSaidClaimTopic
Not yet due Prediction 5 Jun 2025 Amodei said he believes that if the measures Anthropic has advocated are executed well, they would slow China's progress enough to widen America's lead significantly over the next 3-5 years. Prediction · 5 Jun 2025 AI policy and governance
Not yet due Prediction 5 Jun 2025 Amodei worries that within 6-12 months a swarm with greater capabilities but a similar level of misalignment to the one in the OAI-HF incident could be capable of taking over the entire internet with a persistent botnet. Prediction · 5 Jun 2025 AI safety and alignment
Not yet due Prediction 5 Jun 2025 AI could cure most major diseases within 5-10 years, greatly accelerate economic growth, and usher in a renaissance of democracy and freedom. Prediction · 5 Jun 2025 AI timelines and forecasting
Not yet due Prediction 10 Mar 2025 China's chip gap with the US will eventually close over roughly ten to twenty years. Prediction · 10 Mar 2025 AI compute, hardware and data
Not yet due Prediction 10 Mar 2025 Amodei said the same exponential trends and arguments predict that, if progress continues for another two, three or maybe four years, AI models will be as intelligent as Nobel Prize winners across a whole bunch of fields. Prediction · 10 Mar 2025 AI timelines and forecasting
Not yet due Prediction 10 Mar 2025 Amodei said the AI industry as a whole probably needs maybe fifty gigawatts of additional energy by 2027 to fully power AI with the properties he had been describing. Prediction · 10 Mar 2025 AI compute, hardware and data
Not yet due Prediction 28 Feb 2025 Over the next two years, people will wake up to both the risks and benefits of AI to a much more extreme extent than before. Prediction · 28 Feb 2025 Philosophy, ethics and society
Not yet due Prediction 28 Feb 2025 In the longer run (perhaps 18 to 24 months), AI may replace coders, particularly at lower levels. Prediction · 28 Feb 2025 AI, the economy and work
Not yet due Prediction 28 Feb 2025 Amodei put a chance of maybe 70–80% on a very large number of AI systems much smarter than humans at almost everything arriving before the end of the decade, and said his guess was 2026–2027. Prediction · 28 Feb 2025 AI timelines and forecasting
Not yet due Promise 18 Feb 2025 Anthropic has a goal of reaching the point where interpretability can reliably detect most model problems by 2027. Promise · 18 Feb 2025 Their own work or organisation
Not yet due Prediction 18 Feb 2025 Amodei said that, on its current trajectory, he would bet strongly in favor of interpretability becoming a sophisticated and reliable way to diagnose problems in even very advanced AI within 5-10 years. Prediction · 18 Feb 2025 AI timelines and forecasting
Not yet due Prediction 11 Nov 2024 Amodei said that eyeballing the rate at which capabilities are increasing, which he called totally unscientific, does make him think 'we'll get there' by 2026 or 2027, and he added that lots of things could derail it. Prediction · 11 Nov 2024 AI timelines and forecasting
Not yet due Prediction 11 Nov 2024 In the near term, and maybe even in the medium term of 2, 3 or 4 years, Amodei expects humans to keep a huge role and programming as a job not to change, although the nature of programming will change, with less line-by-line writing and more macroscopic work. Prediction · 11 Nov 2024 AI, the economy and work
Not yet due Prediction 11 Nov 2024 Amodei said that probably by 2027 there will be ambitions to build hundred-billion-dollar compute clusters, and that he thinks this will actually happen. Prediction · 11 Nov 2024 AI compute, hardware and data
Not yet due Prediction 11 Nov 2024 Powerful AI's transformative deployment in the world will take more likely five or 10 years than 50 or 100 years. Prediction · 11 Nov 2024 AI timelines and forecasting
Not yet due Prediction 11 Nov 2024 Over the next year or two, even without better models, there would be enormous opportunity to boost productivity by catching mistakes and doing grunt work with current AI models. Prediction · 11 Nov 2024 AI, the economy and work
Not yet due Prediction 11 Nov 2024 Eyeballing the rate at which AI capabilities are increasing makes Amodei think "we'll get there" by 2026 or 2027, though he says there are still worlds where it does not happen in 100 years. Prediction · 11 Nov 2024 AI timelines and forecasting
Not yet due Prediction 11 Nov 2024 Within two to three years, whether or not super powerful AI exists, compute clusters will be large enough to deploy millions of AI model instances. Prediction · 11 Nov 2024 AI compute, hardware and data
Not yet due Prediction 11 Nov 2024 Within two to three years, compute clusters will be large enough to deploy millions of instances of AI models. Prediction · 11 Nov 2024 AI compute, hardware and data
Not yet due Prediction 11 Nov 2024 Within two to three years, the millions of deployable AI model instances will run faster than humans. Prediction · 11 Nov 2024 AI capabilities
Not yet due Prediction 12 Apr 2024 By 2026 to 2028, the compute needed for AI will reach levels the semiconductor industry struggles to supply. Prediction · 12 Apr 2024 AI compute, hardware and data
Not yet due Prediction 12 Apr 2024 If a pause in AI development of one year were needed in 2027, it would be feasible. Prediction · 12 Apr 2024 AI policy and governance
Not yet due Prediction 12 Apr 2024 AI Safety Level ASL-3 capabilities may arrive this year or next (2024 or 2025), and ASL-4 between 2025 and 2028. Prediction · 12 Apr 2024 AI timelines and forecasting
Not yet due Prediction 12 Apr 2024 Amodei said he thinks the cost of training AI models may reach $5 billion to $10 billion in 2025 to 2026. Prediction · 12 Apr 2024 AI timelines and forecasting
Not yet due Prediction 8 Aug 2023 Control of powerful AI models five years from now will involve substantial involvement of a government or assembly of government bodies. Prediction · 8 Aug 2023 AI policy and governance
Not yet due Prediction 8 Aug 2023 AI models that act like a generally well-educated human in an hour-long conversation could exist within two or three years of August 2023. Prediction · 8 Aug 2023 AI timelines and forecasting
Not yet due Prediction 8 Aug 2023 Amodei said that the trend from earlier models to state-of-the-art models shows every sign that, two or three years after August 2023, there will be a real problem. Prediction · 8 Aug 2023 Biosecurity
Not yet due Prediction 2023 AI misuse risks such as biological or chemical weapons uplift might become real risks in 2025 or 2026. Prediction · 2023 Biosecurity
Not yet due Prediction 13 Oct 2022 AI smarter than almost all humans at almost all things will require millions of chips and at least tens of billions of dollars, and is most likely to arrive in 2026-2027. Prediction · 13 Oct 2022 AI timelines and forecasting
Not yet due Prediction 17 Aug 2015 Powerful AI could be built as early as 2026, though it could take much longer. Prediction · 17 Aug 2015 AI timelines and forecasting
Not yet due Prediction 17 Aug 2015 Most mental illness can probably be cured with AI-accelerated neuroscience. Prediction · 17 Aug 2015 Science and research methods
Not yet due Prediction 17 Aug 2015 Amodei thinks AI will similarly accelerate the rate of neuroscience advances, so that the framework of '100 years of progress in 5-10 years' applies to neuroscience in the same way it does to biology. Prediction · 17 Aug 2015 AI capabilities
Not yet due Prediction 17 Aug 2015 The human economy may continue to make sense a little past the point where AI reaches a 'country of geniuses in a datacenter'. Prediction · 17 Aug 2015 AI, the economy and work
Not yet due Prediction 17 Aug 2015 Amodei guessed that powerful AI could at least 10x the rate of these biological discoveries, giving the next 50-100 years of biological progress in 5-10 years. Prediction · 17 Aug 2015 Science and research methods
Not yet due Prediction 17 Aug 2015 Amodei guessed that diseases like PTSD, depression, schizophrenia and addiction can be figured out and very effectively treated. Prediction · 17 Aug 2015 Science and research methods
Not yet due Prediction 17 Aug 2015 While AI is better at only 90% of a given job, the remaining 10% will make humans highly leveraged, raise compensation and create new complementary human jobs. Prediction · 17 Aug 2015 AI, the economy and work
Not yet due Prediction 17 Aug 2015 AI-accelerated neuroscience is likely to vastly improve treatments for, or cure, most mental illness and expand cognitive and mental freedom. Prediction · 17 Aug 2015 AI capabilities
Not yet due Prediction 17 Aug 2015 Amodei thinks 5-10 years after powerful AI is a reasonable timeline for a good fraction (maybe 50%) of AI-driven health benefits to reach even the poorest countries in the world. Prediction · 17 Aug 2015 AI capabilities
Not yet due Prediction 17 Aug 2015 Mind uploading is almost certainly possible in principle but likely will not be achieved within the 5-10 year window of powerful AI. Prediction · 17 Aug 2015 Science and research methods
Not yet due Prediction 17 Aug 2015 Because cancer death rates have fallen about 2% per year for decades, humanity is on track to eliminate most cancer in the 21st century at the current pace of science. Prediction · 17 Aug 2015 Science and research methods
Back to the list

Dario Amodei

Open its page

Prediction Science and research methods

“it’s my guess that powerful AI could at least 10x the rate of these discoveries, giving us the next 50-100 years of biological progress in 5-10 years.”

Dario Amodei, Anthropic CEO, essay author 17 August 2015 Machines of Loving Grace (darioamodei.com)

Amodei guessed that powerful AI could at least 10x the rate of these biological discoveries, giving the next 50-100 years of biological progress in 5-10 years.

Not yet due The date it refers to has not come.

The second judge would not resolve this claim; the verdict shown is the first judge's.

What would count as true or false

This is a conditional prediction about the period after 'powerful AI' arrives. It is judged 5–10 years after powerful AI (as Amodei defines it) is widely recognised to exist. TRUE: within 10 years of that point, the rate of major biological discoveries is at least 10 times the pre-AI baseline. This should show as roughly 50–100 years' worth of progress at the historical pace, judged by expert assessments and counts of landmark discoveries or approved therapies. MOSTLY TRUE: the acceleration is close to 10 times (for example 7–10 times), or 50+ years of progress arrives slightly beyond 10 years. MIXED: there is substantial acceleration of about 3–7 times. MOSTLY FALSE: acceleration is only modest (under about 3 times). FALSE: there is no meaningful acceleration of biological discovery attributable to AI. If powerful AI has not arrived, or 5 years have not passed since it did, the claim cannot be resolved.

Evidence

  1. Are Amodei’s Medical Predictions on Track for 2028-2033? ETC Journal · 22 June 2026
    “The field is moving faster than it did before deep learning entered biology, but it is not yet moving fast enough to make those outcomes likely on that schedule.”
  2. Decade of AI in Drug Discovery: Hype or Reality? International Journal of Pharmaceutical Sciences and Drug Research (IJPSDR) · 30 May 2026
    “The data reveal a bifurcated picture: AI demonstrably outperforms historical industry norms in Phase I (80–90% vs. ~52% success), reflecting genuine superiority in molecular safety engineering and ADMET prediction, while Phase II success (~40%) remains statistically indistinguishable from conventional drug development (~37%).”
  3. Decade of AI in Drug Discovery: Hype or Reality? International Journal of Pharmaceutical Sciences and Drug Research (IJPSDR) · 30 May 2026
    “Yet as of early 2026, no AI-discovered drug has received regulatory approval.”
  4. Artificial intelligence tools expand scientists’ impact but contract science’s focus - Nature Nature · 14 January 2026
    “Scientists who engage in AI-augmented research publish 3.02 times more papers, receive 4.84 times more citations and become research project leaders 1.37 years earlier than those who do not.”

Reasoning

The criteria say the claim can only be judged 5–10 years after 'powerful AI' is widely recognised to exist. No evidence shows that has happened, so the 5-year window has not been shown to have passed. E1 (June 2026) puts Amodei's timeline at 2028–2033 and says biology is moving faster than before deep learning, but not yet fast enough. E3 says no AI-discovered drug had regulatory approval as of early 2026. The outcome is not settled yet, so the verdict is not_yet_due.

Cite this claim

Dario Amodei, “Amodei guessed that powerful AI could at least 10x the rate of these biological discoveries, giving the next 50-100 years of biological progress in 5-10 years.” (said 17 August 2015). Verdict: not yet due. Reliability API, claim c_6f7fcc91f12b, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_6f7fcc91f12b

Claims without a verdict

Is a verdict here wrong, or do you speak for Dario Amodei? Reply on the claim's own page, with the form at its foot (how replies work). Verdicts are made automatically; how verdicts and scores are made.