ClaimsPeople

Demis Hassabis

Sir Demis Hassabis founded DeepMind, leads Isomorphic Labs, shared the 2024 Nobel Prize in Chemistry, and was knighted for services to AI; he was DeepMind's CEO until August 2026, after which he was reported as chairman of Google DeepMind and chief scientist of Alphabet.

802 claims collected, 191 researched, 86 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?

    81 76 87

    High confidence, 78 factual claims resolved. From 15 documents, 2004 to 2026; 18 documents sampled from 30 candidates found.

    Not scored: 29 unresolvable, 13 checkable only against their own records.

  • Predictions

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

    Insufficient record: 2 predictions resolved, 5 needed for a score.

    Not scored: 45 unresolvable. 16 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?

    65 45 83

    Low confidence, 6 promises resolved. From 6 documents, 2004 to 2026; 18 documents sampled from 30 candidates found.

    Not scored: 1 checkable only against their own records.

All types combined: 80 (likely 75 to 85, high confidence), from 84 claims: 78 factual claims, 6 promises. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.

Demis Hassabis's claims

37 claims

37 claims, sorted by newest said. Choose a claim to read it in full.
VerdictKindSaidClaimTopic
Resolved true Factual 9 May 2026 It has been 10 years since AlphaGo. Factual · 9 May 2026 Their own work or organisation
Resolved true Factual 29 Apr 2026 Demis Hassabis met South Korean President Lee Jae-myung in Seoul. Factual · 29 Apr 2026 Their own work or organisation
Resolved true Factual 10 Apr 2026 Over 3 million scientists are now using AlphaFold. Factual · 10 Apr 2026 Science and research methods
Resolved true Factual 10 Apr 2026 Only about 10% of drugs get through all the clinical stages. Factual · 10 Apr 2026 Science and research methods
Resolved true Factual 10 Apr 2026 SynthID uses AI to digitally watermark any generated image, and Google generative technologies such as Veo carry this watermarking. Factual · 10 Apr 2026 AI capabilities
Resolved true Factual 25 Nov 2025 In 2022, Google DeepMind released AlphaFold 2 predictions for more than 200 million protein structures. Factual · 25 Nov 2025 Science and research methods
Resolved true Prediction 23 Jul 2025 Asked whether it would be nuclear fusion, Demis Hassabis said he thinks fusion and solar are the two energy sources he would bet on. Prediction · 23 Jul 2025 Technology
Resolved true Factual 23 Jul 2025 Demis Hassabis worked on the early stages of the game Black & White in the 1990s, which he says had probably the best learning AI in a game. Factual · 23 Jul 2025 Their own work or organisation
Resolved true Factual 23 Jul 2025 AI systems are now in products used by billions of people around the world, requiring large amounts of inference compute. Factual · 23 Jul 2025 AI compute, hardware and data
Resolved true Factual 23 Jul 2025 Over the last year a new paradigm of thinking AI systems emerged, which get smarter the more inference time they are given at test time. Factual · 23 Jul 2025 AI capabilities
Resolved true Factual Dec 2024 In 2016 AlphaGo beat Lee Sedol 4-1 in a match in Seoul. Factual · Dec 2024 Other
Resolved true Factual Dec 2024 Demis Hassabis learned to play chess at age four. Factual · Dec 2024 Their own work or organisation
Resolved true Factual Dec 2024 AlphaFold2 achieved atomic accuracy at CASP14 in 2020. Factual · Dec 2024 Science and research methods
Resolved true Prediction 28 Feb 2024 Hassabis said he thought the next versions of these systems, over the next year to 18 months, might have some contextual understanding of the user's environment through a camera, phone or glasses. Prediction · 28 Feb 2024 AI timelines and forecasting
Resolved true Factual 28 Feb 2024 Hassabis said he thinks Google DeepMind helped pioneer RLHF. Factual · 28 Feb 2024 AI capabilities
Resolved true Promise 28 Feb 2024 Google DeepMind will start publishing its safety checks and capability-based safeguards, having already internal checks and balances. Promise · 28 Feb 2024 AI policy and governance
Resolved true Factual 20 Apr 2023 Zoubin Ghahramani will join the Google DeepMind research leadership team reporting to Koray Kavukcuoglu. Factual · 20 Apr 2023 Their own work or organisation
Resolved true Factual 20 Apr 2023 Google DeepMind brings together DeepMind and Brain AI talent with Google's computing power, infrastructure and resources. Factual · 20 Apr 2023 Their own work or organisation
Resolved true Factual 20 Apr 2023 On 20 April 2023 Sundar Pichai announced that DeepMind and Google Research's Brain team would join as a single unit called Google DeepMind. Factual · 20 Apr 2023 Their own work or organisation
Resolved true Factual 9 Mar 2023 Clement (@clmt) joined DeepMind. Factual · 9 Mar 2023 Field building and careers
Resolved true Factual 23 Feb 2023 Jeff Clune joined DeepMind. Factual · 23 Feb 2023 Field building and careers
Resolved true Factual 9 Jan 2023 Nobel laureates Jennifer Doudna, David MacMillan, Paul Nurse and Venki Ramakrishnan joined the Isomorphic Labs Scientific Advisory Board. Factual · 9 Jan 2023 Their own work or organisation
Resolved true Factual 10 Dec 2021 DeepMind has published work applying AI to quantum chemistry in Science magazine. Factual · 10 Dec 2021 Science and research methods
Resolved true Factual 10 Dec 2021 The properties of materials largely depend on their electrons. Factual · 10 Dec 2021 Science and research methods
Delivered on time Promise 9 Dec 2021 DeepMind will release more than 100 million protein structures over the next year, covering almost every sequenced protein known to science. Promise · 9 Dec 2021 Science and research methods
Resolved true Factual 22 Jul 2021 DeepMind and EMBL-EBI launched the AlphaFold Protein Structure Database, the most complete and accurate picture of the human proteome to date, more than doubling known high-accuracy human protein structures. Factual · 22 Jul 2021 Science and research methods
Resolved true Factual 22 Jul 2021 Through early collaborations DeepMind has seen promising signals from researchers using AlphaFold, including DNDi and the University of Portsmouth's Centre for Enzyme Innovation. Factual · 22 Jul 2021 Science and research methods
Resolved true Factual 22 Jul 2021 Besides the ~20,000 human proteins, DeepMind provides open access to proteomes of 20 other organisms, totalling over 350,000 protein structures. Factual · 22 Jul 2021 Science and research methods
Resolved true Factual 12 Feb 2020 Demis Hassabis received the 2020 Dan David Prize Future Award alongside Prof. Amnon Shashua. Factual · 12 Feb 2020 Their own work or organisation
Resolved true Factual 15 Jan 2020 The DeepMind Nature paper takes inspiration from its recent reinforcement learning work to shed new light on how dopamine works in the brain. Factual · 15 Jan 2020 Science and research methods
Resolved true Factual 15 Jan 2020 A DeepMind paper on AI and neuroscience was published in Nature on 2020-01-15. Factual · 15 Jan 2020 Science and research methods
Resolved true Factual 15 Jan 2020 DeepMind's AlphaFold work, using deep neural networks to predict protein structures, was published in Nature on 2020-01-15. Factual · 15 Jan 2020 Science and research methods
Resolved true Factual 1 Jan 2020 The DeepMind, Google Health and Cancer Research UK breast cancer screening study was published in Nature on 2020-01-01. Factual · 1 Jan 2020 Science and research methods
Delivered on time Promise 2004 Hassabis said Elixir was confident that players would get their hands on Evil Genius in fall 2004. Promise · 2004 Other
Resolved true Factual 2004 In Evil Genius, more gloating increases the chance that a captured agent turns the tables and escapes. Factual · 2004 Other
Resolved true Factual 2004 Evil Genius was influenced by non-spy classics such as Enter the Dragon and Flash Gordon. Factual · 2004 Other
Resolved true Factual 2004 Many journalists came away from previews of Evil Genius amused. Factual · 2004 Other
Back to the list

Factual AI compute, hardware and data

“there’s also because now AI systems are in products and being used by billions of people around the world, you need a ton of inference compute now.”

Demis Hassabis, CEO of Google DeepMind, interviewee 23 July 2025 Transcript for Demis Hassabis: Future of AI, Simulating Reality, Physics and Video Games (#475) (Lex Fridman)

AI systems are now in products used by billions of people around the world, requiring large amounts of inference compute.

Resolved true The record shows it happened or was so.

Said for an organisation they lead or work for; it counts on both records.

What would count as true or false

Judged as of 2025-07-23. True if, at that date, AI systems were embedded in products used by billions of people (e.g. Google Search AI Overviews, Gemini features or other AI-enabled products with user counts in the billions) and inference compute demand was large and growing. Mostly true if user counts reach the billions only by counting broad AI-enabled products (e.g. ranking or recommendation systems) rather than generative AI, or if the figures are somewhat below 'billions' for a single product but reach them across products. Mixed if one part holds but the other is doubtful: the billions of users or the large inference demand. Mostly false if AI products reached only hundreds of millions of users. False if the evidence shows neither part held.

Evidence

  1. Q2 earnings call: CEO's remarks Google (blog.google) · 23 July 2025 Own record
    “AI Overviews now has over 2 billion monthly users across more than 200 countries and territories and 40 languages.”
  2. Q2 earnings call: CEO's remarks Google (blog.google) · 23 July 2025 Own record
    “The Gemini App now has more than 450 million monthly active users, and we continue to see strong growth and engagement, with daily requests growing over 50% from Q1.”
  3. Q2 earnings call: CEO's remarks Google (blog.google) · 23 July 2025 Own record
    “At I/O in May, we announced that we processed 480 trillion monthly tokens across our surfaces. Since then we have doubled that number, now processing over 980 trillion monthly tokens, a remarkable increase.”
  4. Google conference call: The Global Strategy of full-stack AI drives comprehensive Business growth, with the monthly Tokens processing volume doubling and visual search increasing by 70%. moomoo news (republishing wallstreetcn) · 24 July 2025
    “The company has significantly increased its capital expenditure forecast for 2025 to $85 billion, an increase of $10 billion from the previous estimate, primarily for AI infrastructure development.”

Reasoning

In Alphabet's Q2 2025 earnings remarks, dated the same day as the claim, Google said AI Overviews had over 2 billion monthly users in more than 200 countries and territories (E1), and the Gemini app had more than 450 million monthly active users (E2). Google also said the monthly tokens it processed roughly doubled, from 480 trillion to over 980 trillion, between May and July (E3), which shows large and growing inference demand. An independent financial news report says Google raised its 2025 capital spending forecast to $85 billion, mainly for AI infrastructure (E4). That fits with rising compute needs, although the user and token figures come only from Google's own official earnings remarks.

Cite this claim

Demis Hassabis, “AI systems are now in products used by billions of people around the world, requiring large amounts of inference compute.” (said 23 July 2025). Verdict: resolved true. Reliability API, claim c_b52e9e316a39, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_b52e9e316a39

Claims without a verdict

Is a verdict here wrong, or do you speak for Demis Hassabis? 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.