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 -
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.
-
Promises
Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?
65 45 83
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
“It takes inspiration from our recent work in Reinforcement Learning to shed new light on how dopamine works in the brain.”
Demis Hassabis, DeepMind CEO, on X 15 January 2020 Post by @demishassabis on X, 2020-01-15 (X)
The DeepMind Nature paper takes inspiration from its recent reinforcement learning work to shed new light on how dopamine works in the brain.
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
True: the DeepMind Nature paper published around 2020-01-15 (Dabney et al., "A distributional code for value in dopamine-based reinforcement learning") draws on DeepMind's recent reinforcement learning work, such as distributional RL, to give new findings or hypotheses about how dopamine works in the brain. Mostly true: the paper matches this, but a minor detail differs, for example the RL work drawn on was not DeepMind's own or not recent. Mixed: the paper concerns dopamine but its link to RL work is only partial. Mostly false: the paper exists but is not based on RL work, or is not about dopamine. False: no such DeepMind Nature paper was published around that date.
Evidence
- Dopamine and temporal difference learning: A fruitful relationship between neuroscience and AI DeepMind · 15 January 2020 Own record
“Because distributional TD is so powerful in artificial neural networks, a natural question arises: Is distributional TD used in the brain? This was the driving question behind our paper recently published in Nature.”
- Dopamine and temporal difference learning: A fruitful relationship between neuroscience and AI DeepMind · 15 January 2020 Own record
“In summary, we found that dopamine neurons in the brain were each tuned to different levels of pessimism or optimism.”
- Dopamine and Learning: The Molecule That Teaches Your Brain Mindomax · 30 April 2026
“In January 2020, a team from DeepMind and Harvard published a paper in Nature that changed the conversation [10].”
- Dopamine and Learning: The Molecule That Teaches Your Brain Mindomax · 30 April 2026
“This was not a random observation. It matched a specific algorithm from artificial intelligence called distributional reinforcement learning, which the DeepMind team had itself developed.”
Reasoning
DeepMind's 2020-01-15 blog post says its recently published Nature paper asked whether distributional TD, a reinforcement learning algorithm, is used in the brain (E1). The post also reports that the paper found individual dopamine neurons tuned to different levels of optimism or pessimism (E2). An independent article confirms that a DeepMind and Harvard team published a Nature paper on dopamine in January 2020 (E3). The same article says the finding matched distributional reinforcement learning, which the DeepMind team had developed itself (E4).
Cite this claim
Demis Hassabis, “The DeepMind Nature paper takes inspiration from its recent reinforcement learning work to shed new light on how dopamine works in the brain.” (said 15 January 2020). Verdict: resolved true. Reliability API, claim c_09eef94b7bb4, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_09eef94b7bb4
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.