ClaimsPeople

Ray Kurzweil

Ray Kurzweil is an author of books on artificial intelligence and the technological singularity, including The Singularity Is Near, The Singularity Is Nearer, The Age of Spiritual Machines and How to Create a Mind; he received the 1999 National Medal of Technology and Innovation, became Director of Engineering at Google and co-founded Singularity University.

894 claims collected, 397 researched, 78 resolved. Report of 5 November 2026, method 0.1.0+28da144b.

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?

    68 61 75

    High confidence, 69 factual claims resolved. From 13 documents, 2000 to 2026; 15 documents sampled from 30 candidates found.

    Not scored: 61 unresolvable, 14 checkable only against their own records.

  • Predictions

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

    42 25 60

    Low confidence, 9 predictions resolved. From 4 documents, 2001 to 2020; 15 documents sampled from 30 candidates found.

    Not scored: 140 unresolvable, 2 checkable only against their own records. 98 more fall due later.

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

  • Promises

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

    Insufficient record: 0 promises resolved, 5 needed for a score.

    Not scored: 1 not yet due.

All types combined: 65 (likely 58 to 71, high confidence), from 78 claims: 69 factual claims, 9 predictions. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.

Ray Kurzweil's claims

9 claims

9 claims, sorted by newest said. Choose a claim to read it in full.
VerdictKindSaidClaimTopic
Mixed Factual Mar 2025 AI hallucinations persist but are becoming much less frequent. Factual · Mar 2025 AI capabilities
Mixed Factual Mar 2025 Kurzweil said that at about 10 billion connections neural networks started to do things. Factual · Mar 2025 AI compute, hardware and data
Mixed Factual Mar 2025 More and more people from more backgrounds, including third world countries in Africa and South America, now participate in technological advances than previously. Factual · Mar 2025 Economics and markets
Mixed Factual 9 Jul 2024 Among the forecasts he was describing, which averaged a hundred years, nobody said it would happen within 30 years, but now everybody is saying that. Factual · 9 Jul 2024 AI timelines and forecasting
Mixed Factual 9 Jul 2024 Cloud computing is expanding exponentially, doubling in speed every year for the same money. Factual · 9 Jul 2024 AI compute, hardware and data
Mixed Factual May 2020 The current method for testing the few treatments that are tried organizes a few hundred human subjects and tests them over about a year and a half, at a cost of hundreds of millions of dollars. Factual · May 2020 Science and research methods
Mixed Factual Mar 2017 In recent years Minsky lamented the 'success' of his book and had become respectful of recent gains in neural net technology. Factual · Mar 2017 AI capabilities
Mixed Factual Mar 2017 For every job that was eliminated, several new ones were invented, with the work force growing from 24 million people (31 percent of the population) in 1900 to 142 million (44 percent of the population) in 2015. Factual · Mar 2017 AI, the economy and work
Mixed Factual 23 Oct 2000 The 'immune system' that evolved in response to computer viruses has been largely effective. Factual · 23 Oct 2000 Technology
Back to the list

Ray Kurzweil

Open its page

Factual AI capabilities

“AI hallucinations are still there, but are becoming much less frequent.”

Ray Kurzweil, speaking at MWC 25 March 2025 Ray Kurzweil: 2022-2025 Updates (transcripts) (LifeArchitect.ai)

AI hallucinations persist but are becoming much less frequent.

Mixed Partly true and partly false in roughly equal measure.

What would count as true or false

The claim: as of March 2025, AI (LLM) hallucinations still occur but are becoming much less frequent. True: benchmark or evaluation data available by about March 2025 (e.g. Vectara hallucination leaderboard, SimpleQA, vendor system cards, independent studies) show hallucinations still happen in current models, and newer models have substantially lower rates than earlier ones (e.g. roughly halving or better over 1–2 years). Mostly true: hallucinations persist and are declining, but the improvement is moderate or uneven, not 'much' less frequent. Mixed: some benchmarks show clear declines while others show flat or higher rates (e.g. reasoning models hallucinating more). Mostly false: hallucination rates are broadly flat, and only a narrow benchmark shows a decline. False: hallucinations have stopped occurring, or evaluations show they are generally becoming more frequent.

Evidence

  1. OpenAI o3 and o4-mini System Card - OpenAI Deployment Safety Hub OpenAI
    “Dataset Metric o3 o4-mini o1 SimpleQA accuracy (higher is better) 0.49 0.20 0.47 hallucination rate (lower is better) 0.51 0.79 0.44 PersonQA accuracy (higher is better) 0.59 0.36 0.47 hallucination rate (lower is better) 0.33 0.48 0.16”
  2. DeepSeek-R1 hallucinates more than DeepSeek-V3 Vectara · 30 January 2025
    “The results were surprising: Deepseek-R1’s hallucination rate was 14.3% - much higher than its non-reasoning predecessor: Deepseek-V3.”
  3. DeepSeek-R1 hallucinates more than DeepSeek-V3 Vectara · 30 January 2025
    “According to Table 3, while HHEM 2.1 shows that the reasoning LLM (GPT-o1) has a higher hallucination rate than the generic LLM (GPT-4o), FACTS shows otherwise.”
  4. Measuring short-form factuality in large language models OpenAI
    “Current frontier models sometimes produce false outputs or answers that are not substantiated by evidence, a problem known as “hallucinations.””

Reasoning

E4 confirms that frontier models still hallucinate, so the first half of the claim holds. The evidence does not show hallucinations becoming much less frequent. Vectara's January 2025 test found DeepSeek-R1 at a 14.3% hallucination rate, much higher than its predecessor DeepSeek-V3 (E2). Vectara also found that comparing o1 with GPT-4o gives opposite results depending on the measure: HHEM 2.1 shows o1 hallucinating more, while FACTS shows otherwise (E3). OpenAI's undated o3/o4-mini system card shows higher hallucination rates for o3 and o4-mini than for o1 on both PersonQA and SimpleQA (E1). With one measure showing a decline and others showing higher rates for newer reasoning models, this matches the criteria's description of a mixed result.

Cite this claim

Ray Kurzweil, “AI hallucinations persist but are becoming much less frequent.” (said March 2025). Verdict: mixed. Reliability API, claim c_4006e11392ae, method 0.1.0+28da144b. https://epistacklabs.com/claims/c_4006e11392ae

Claims without a verdict

Is a verdict here wrong, or do you speak for Ray Kurzweil? 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.