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

42 claims

42 claims, sorted by newest said. Choose a claim to read it in full.
VerdictKindSaidClaimTopic
Mostly true Factual 27 May 2026 Isomorphic Labs has a number of its drug programs in the pre-clinical stage. Factual · 27 May 2026 Their own work or organisation
Mostly true Factual 20 May 2026 Gemini 3.5 Flash performs better than Gemini 3.1 Pro on coding and agentic tasks. Factual · 20 May 2026 AI capabilities
Mostly true Factual 19 May 2026 Gemini Omni can take photos, video and audio and build entirely new scenes. Factual · 19 May 2026 AI capabilities
Mostly true Factual 10 Apr 2026 Deep Blue was world-champion level at chess but could not do anything else. Factual · 10 Apr 2026 AI capabilities
Mostly true Factual 10 Apr 2026 Demis Hassabis decided from a very young age to work on AI because he was obsessed with big questions as a schoolchild. Factual · 10 Apr 2026 Their own work or organisation
Mostly true Factual 10 Apr 2026 A protein's 3D structure is what determines, or partially determines, its function. Factual · 10 Apr 2026 Science and research methods
Mostly true Factual 25 Nov 2025 The AlphaFold Server is available to non-commercial researchers globally for predicting structures and interactions. Factual · 25 Nov 2025 Science and research methods
Mostly true Factual 25 Nov 2025 AlphaProteo can design novel, high-strength protein binders targeting diverse molecules, including ones associated with cancer and diabetes. Factual · 25 Nov 2025 Science and research methods
Mostly true Factual 3 Aug 2025 Proteins are the basic building blocks of life. Factual · 3 Aug 2025 Science and research methods
Mostly true Factual 3 Aug 2025 Current AI systems still can't really yet go beyond asking novel questions or conjectures, or come up with new hypotheses not thought of before. Factual · 3 Aug 2025 AI capabilities
Mostly true Factual 3 Aug 2025 On average it takes ten years and billions of dollars to design one drug. Factual · 3 Aug 2025 Science and research methods
Mostly true Factual 3 Aug 2025 Everything in biology and the body depends on proteins. Factual · 3 Aug 2025 Science and research methods
Mostly true Factual 23 Jul 2025 Video models such as Veo 3 can pretty accurately generate about eight seconds of video that, by eye and at a glance, looks consistent. Factual · 23 Jul 2025 AI capabilities
Mostly true Factual 23 Jul 2025 Veo, Google DeepMind's video generation model, can model liquids surprisingly well. Factual · 23 Jul 2025 AI capabilities
Mostly true Factual 23 Jul 2025 Google DeepMind built DolphinGemma, a version of its system trained on dolphin and whale sounds. Factual · 23 Jul 2025 Science and research methods
Mostly true Factual Dec 2024 Hassabis's first career was writing AI for commercial computer games. Factual · Dec 2024 Their own work or organisation
Mostly true Factual Dec 2024 Little progress was made in CASP protein structure prediction accuracy for a decade before AlphaFold. Factual · Dec 2024 Science and research methods
Mostly true Factual Dec 2024 AlphaFold was open sourced and made freely available for the world to use. Factual · Dec 2024 Science and research methods
Mostly true Factual Dec 2024 Proteins are the building blocks of life. Factual · Dec 2024 Science and research methods
Mostly true Factual Dec 2024 Over two million researchers have used AlphaFold and its predicted structures. Factual · Dec 2024 Science and research methods
Mostly true Factual Dec 2024 Theme Park had AI at the heart of its gameplay. Factual · Dec 2024 Their own work or organisation
Mostly true Factual Dec 2024 Hassabis first came across the protein folding problem as an undergraduate at Cambridge in the 1990s. Factual · Dec 2024 Their own work or organisation
Mostly true Factual Dec 2024 AlphaFold2 was re-architected for CASP14 to reach atomic accuracy, leading the CASP organisers to declare the problem solved. Factual · Dec 2024 Science and research methods
Mostly true Factual 28 Feb 2024 DeepMind has always bet on generality and learning as the core of its techniques, focusing on reinforcement learning, search and deep learning. Factual · 28 Feb 2024 Their own work or organisation
Mostly true Factual 28 Feb 2024 Demis Hassabis thinks human brains plan moves differently from AlphaGo because brains are not built for Monte Carlo tree search. Factual · 28 Feb 2024 Science and research methods
Mostly true Factual 28 Feb 2024 Demis Hassabis thinks that improving in one specific domain sometimes produces surprising improvement in other domains; as an example, he said that when large models improve at coding, that can improve their general reasoning. Factual · 28 Feb 2024 AI capabilities
Mostly true Factual 28 Feb 2024 DeepMind and Google Brain previously collaborated project by project, and now collaborate more deeply and broadly. Factual · 28 Feb 2024 Their own work or organisation
Mostly true Factual 28 Feb 2024 Gemini is the first product of the DeepMind–Google Brain collaboration, and its name implies twins. Factual · 28 Feb 2024 Their own work or organisation
Mostly true Promise 20 Apr 2023 Google DeepMind is creating a new Scientific Board, led by Koray Kavukcuoglu with representatives from across the orgs, to oversee research progress and direction. Promise · 20 Apr 2023 Their own work or organisation
Mostly true Factual 20 Apr 2023 Eli Collins will join Demis Hassabis's leads team at Google DeepMind as VP of Product. Factual · 20 Apr 2023 Field building and careers
Mostly true Factual 20 Apr 2023 When DeepMind launched in 2010, many people thought general AI was farfetched science fiction decades away. Factual · 20 Apr 2023 Their own work or organisation
Mostly true Factual 20 Apr 2023 Research advances from the Brain and DeepMind teams, from Deep Reinforcement Learning to Transformers, laid much of the foundations of the current AI industry. Factual · 20 Apr 2023 Their own work or organisation
Mostly true Factual 1 Feb 2023 More than one million researchers have used AlphaFold in the 18 months since DeepMind made it freely available. Factual · 1 Feb 2023 Science and research methods
Mostly true Factual 28 Oct 2021 Demis Hassabis gave the Kendrew lecture and visited the MRC Laboratory of Molecular Biology, and the talk has been posted online. Factual · 28 Oct 2021 Their own work or organisation
Mostly true Factual 13 Oct 2021 Demis Hassabis was giving the John Kendrew lecture on AlphaFold at the MRC LMB at 2.30pm UK time on 13 October 2021. Factual · 13 Oct 2021 Their own work or organisation
Mostly true Factual 22 Jul 2021 The AlphaFold work represents the culmination of five years of effort by the DeepMind AlphaFold team, including overcoming setbacks and developing new algorithmic innovations. Factual · 22 Jul 2021 Their own work or organisation
Mostly true Factual 22 Jul 2021 When DeepMind announced AlphaFold 2 in December, it was hailed as a solution to the 50-year-old protein folding problem. Factual · 22 Jul 2021 Science and research methods
Mostly true Factual 22 Jul 2021 DeepMind's AlphaFold team worked for five years and closely with EMBL-EBI for the past few months to be able to share the AlphaFold resource. Factual · 22 Jul 2021 Their own work or organisation
Mostly true Promise 22 Jul 2021 In the coming months DeepMind plans to expand the database to over 100 million structures covering most of the UniProt reference database. Promise · 22 Jul 2021 Science and research methods
Mostly true Factual 2004 Lead artist Brian Gillies wanted a shiny "atomic age" 1960s style for Evil Genius. Factual · 2004 Other
Mostly true Factual 2004 In Evil Genius, the more the player gloats, the more notoriety is gained, and notoriety is key to progress. Factual · 2004 Other
Mostly true Factual 2004 Demis Hassabis first conceived the kernel of the idea for Evil Genius before Elixir Studios had started. Factual · 2004 Other
Back to the list

Factual Science and research methods

“I think it’s different because our brains are not built for doing Monte Carlo tree search.”

Demis Hassabis, CEO of Google DeepMind 28 February 2024 Demis Hassabis — Scaling, superhuman AIs, AlphaZero atop LLMs, AlphaFold (Dwarkesh Podcast)

Demis Hassabis thinks human brains plan moves differently from AlphaGo because brains are not built for Monte Carlo tree search.

Mostly true True in substance, wrong in a detail.

What would count as true or false

The claim is hedged ('I think'). It holds that human planning in games such as Go differs from AlphaGo's because the brain is not built to do Monte Carlo tree search. The premise that AlphaGo uses MCTS is checkable from DeepMind's papers. TRUE: AlphaGo uses MCTS, and cognitive-science or neuroscience research broadly supports that human planning does not implement MCTS-style large-scale simulated rollouts or search, relying instead on far more limited, pattern-based or heuristic search. MOSTLY TRUE: the difference is supported, but the research shows that humans do perform some limited tree-search-like lookahead, so the contrast is one of degree. MIXED: the evidence is genuinely split on whether human planning resembles MCTS. MOSTLY FALSE: research has found substantial MCTS-like mechanisms in human planning, or the question is unresolved and the claim overstates it as settled. FALSE: AlphaGo does not use MCTS, or the evidence shows that human brains do implement MCTS-like planning in the relevant sense.

Evidence

  1. Mastering the game of Go with deep neural networks and tree search - Nature Nature · 27 January 2016 Own record
    “We also introduce a new search algorithm that combines Monte Carlo simulation with value and policy networks.”
  2. [PDF] THE GAME OF CHESS - Carnegie Mellon University Elsevier Science Publishers (Handbook of Game Theory, Vol. 1; authors H. A. Simon and J. Schaeffer, Carnegie-Mellon and University of Alberta) · 1992
    “we have known that a grandmaster, even in a difficult position, carries out a very modest amount of search of the game tree, probably seldom more than 100 branches”
  3. Expertise increases planning depth in human gameplay - Nature Nature · 31 May 2023
    “We use model fitting methods to show that human behaviour can be captured using a computational cognitive model based on heuristic search.”
  4. Neural Signatures of Tree Search - Cognitive Computational ... Cognitive Computational Neuroscience (CCN) 2025 conference · 26 September 2025
    “A leading theoretical framework proposes that human planning follows an algorithm akin to tree search.”

Reasoning

The AlphaGo paper confirms the premise: its search algorithm combines Monte Carlo simulation with value and policy networks (E1). Independent research finds that human experts search the game tree only a little (E2), and that a cognitive model built on heuristic search captures human play (E3). Both sources also show that humans do some limited tree-search-like lookahead, and E4 notes that a leading framework treats human planning as akin to tree search. The difference from MCTS is therefore supported, but it is one of degree, which matches the "mostly true" criterion.

Cite this claim

Demis Hassabis, “Demis Hassabis thinks human brains plan moves differently from AlphaGo because brains are not built for Monte Carlo tree search.” (said 28 February 2024). Verdict: mostly true. Reliability API, claim c_377074f60bdc, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_377074f60bdc

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.