ClaimsPeople

Rodney Brooks

Rodney Brooks is a roboticist, MIT professor emeritus, and co-founder of iRobot, Rethink Robotics and Robust.AI.

1,166 claims collected, 291 researched, 88 resolved. Report of 5 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?

    83 77 88

    High confidence, 79 factual claims resolved. From 17 documents, 2014 to 2026; 19 documents sampled from 35 candidates found.

    Not scored: 31 unresolvable, 17 checkable only against their own records.

  • Predictions

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

    67 49 83

    Low confidence, 8 predictions resolved. From 6 documents, 2018 to 2026; 19 documents sampled from 35 candidates found.

    Not scored: 91 unresolvable. 60 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: 1 promise resolved, 5 needed for a score.

    Not scored: 3 not yet due.

All types combined: 81 (likely 76 to 86, high confidence), from 87 claims: 79 factual claims, 8 predictions. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.

Rodney Brooks's claims

39 claims

39 claims, sorted by newest said. Choose a claim to read it in full.
VerdictKindSaidClaimTopic
Mostly true Factual 18 Jun 2026 On 18 June 2026 Donald Trump signed an agreement with Iran at Versailles. Factual · 18 Jun 2026 Other
Mostly true Factual 12 Apr 2026 JD Vance campaigned for Viktor Orban in the Hungarian elections the week before 12 April 2026, and Orban lost the election. Factual · 12 Apr 2026 Other
Mostly true Factual 1 Jan 2026 When Brooks predicted in 2018 that something big would appear in AI, he did not know it would be LLMs. Factual · 1 Jan 2026 Their own work or organisation
Mostly true Factual 1 Jan 2026 Since 1991 Brooks has distinguished between situatedness and embodiment as two separate properties a machine or creature can have. Factual · 1 Jan 2026 AI capabilities
Mostly true Factual 1 Jan 2026 Blue Origin is under contract with NASA to land two astronauts on the Moon for a 30-day stay in 2030 using Blue Moon Mark 2. Factual · 1 Jan 2026 Other
Mostly true Factual 1 Jan 2026 Once pick orientation and vacuum gripper are selected, suction grasping is open loop, relying on suction force and compliance of gripper and object. Factual · 1 Jan 2026 AI capabilities
Mostly true Factual 1 Jan 2025 Zoox has purpose-built vehicles with no steering wheel or brake and accelerator pedals. Factual · 1 Jan 2025 Technology
Mostly true Factual 1 Jan 2025 Both billionaire founders of the two suborbital tourism companies have flown to space on their own craft. Factual · 1 Jan 2025 Economics and markets
Mostly true Factual 1 Jan 2025 In the 2017 'Attention Is All You Need' transformer architecture, each column from bottom to top is a pure feed-forward network with no search, iteration or conventional algorithm. Factual · 1 Jan 2025 AI capabilities
Mostly true Factual 29 Sep 2024 The web has provided massive amounts of text for machine learning, which has given rise to LLMs. Factual · 29 Sep 2024 AI compute, hardware and data
Mostly true Factual 29 Sep 2024 The Roomba did not require customers to install any infrastructure before use; it needed only a standard electrical outlet for its charger. Factual · 29 Sep 2024 Technology
Mostly true Factual 29 Sep 2024 Images and movies from the web, and others collected for the purpose, have required human labeling for machine learning. Factual · 29 Sep 2024 AI compute, hardware and data
Mostly true Factual 1 Jan 2024 Brooks has spent his whole professional life developing robots, and his companies have built more robots than anyone else's. Factual · 1 Jan 2024 Their own work or organisation
Mostly true Factual 1 Jan 2024 By March 2023 it was clear that large language models were the next big thing in AI. Factual · 1 Jan 2024 AI capabilities
Mostly true Factual 1 Jan 2024 In late 2022 Brooks analysed the lack of videos showing any practical-looking flights of any eVTOL vehicle, despite multi-billion dollar company valuations. Factual · 1 Jan 2024 Technology
Mostly true Factual 21 Oct 2023 Brooks said he has reported on particular autonomous vehicle rides, his or others', when he thought the vehicles had done something for which a human driver would somehow be sanctioned. Factual · 21 Oct 2023 Their own work or organisation
Mostly true Factual 7 Jun 2023 Rodney Brooks has taken many fully driverless Cruise rides at night in San Francisco. Factual · 7 Jun 2023 Technology
Mostly true Factual 24 May 2023 Brooks described hundreds of driverless cars with no humans inside as driving around San Francisco 'aimlessly' 24 hours a day. Factual · 24 May 2023 Technology
Mostly true Prediction 1 Jan 2023 Brooks said he was concerned that Starship would not be ready for the proposed 2025 crewed lunar landing, and that this would delay humankind's return to the Moon. Prediction · 1 Jan 2023 Technology
Mostly true Factual 1 Jan 2023 NASA's current plan is for Artemis 2 to take astronauts to lunar orbit in May 2024, followed by Artemis 3 landing two astronauts on the Moon in 2025. Factual · 1 Jan 2023 Technology
Mostly true Factual 1 Jan 2023 In essence, ChatGPT picks the most likely next word given the preceding words, whether from the user's prompt or its own generated output. Factual · 1 Jan 2023 AI capabilities
Mostly true Factual 14 Aug 2020 Rodney Brooks has said for years that adoption of autonomous vehicles will require changes to road infrastructure. Factual · 14 Aug 2020 Technology
Mostly true Factual 1 Jan 2020 During 2019 Waymo started operating what Brooks called a 'taxi service' in Chandler, Arizona, with no human driver in the vehicles. Factual · 1 Jan 2020 Technology
Mostly true Factual 1 Jan 2020 The December 2019 Starliner flight was meant to be the final unmanned test, planned to dock with the ISS and then soft-land on the ground. Factual · 1 Jan 2020 Technology
Mostly true Factual 20 Aug 2019 Human drivers perform the three-point turns that turn the vehicles around at the end of the route each time. Factual · 20 Aug 2019 Technology
Mostly true Factual 13 Jul 2019 Fernando Corbato developed Multics, which led to Unix, which led to Linux. Factual · 13 Jul 2019 Technology
Mostly true Factual 19 Mar 2019 The push to make everything learnable is raising the cost of AI beyond what individuals or even large university departments can afford. Factual · 19 Mar 2019 AI compute, hardware and data
Mostly true Factual 19 Mar 2019 Moore's Law is slowing, and some computer architects report the doubling time of computation on a single chip moving from one year to twenty years. Factual · 19 Mar 2019 AI compute, hardware and data
Mostly true Factual 15 Jul 2018 Kim's research team made their legged robot completely blind to show what its low-level reactive behaviours could do. Factual · 15 Jul 2018 AI capabilities
Mostly true Factual 15 Jul 2018 Gravitational waves, predicted by Einstein in 1916, were first detected in 2015, ninety-nine years later. Factual · 15 Jul 2018 Science and research methods
Mostly true Factual 1 Jan 2018 Horseless carriages did not simply replace horse-drawn carriages one for one but required new infrastructure and ownership and usage models. Factual · 1 Jan 2018 Technology
Mostly true Factual 7 Sep 2017 AI was overestimated in the 1960s and the 1980s, and Brooks believes it is being overestimated again in 2017. Factual · 7 Sep 2017 AI timelines and forecasting
Mostly true Factual 7 Sep 2017 Hardware has a significant marginal cost to deploy. Factual · 7 Sep 2017 Technology
Mostly true Factual 12 Jan 2017 At least one car manufacturer fears human drivers will bully level 2 self-driving cars, so its level 3 real-world trial cars look identical to conventional models. Factual · 12 Jan 2017 Technology
Mostly true Factual 12 Jan 2017 In level 2 and 3 autonomy, the time available for a person to reorient and respond is often much shorter than people need. Factual · 12 Jan 2017 Technology
Mostly true Factual 12 Jan 2017 Many groups, including Tesla, have promoted the advantage of cars autonomously handling parking. Factual · 12 Jan 2017 Technology
Mostly true Factual 16 Dec 2014 Labour costs were rising and the era of infinitely cheap labour was coming to an end. Factual · 16 Dec 2014 AI, the economy and work
Mostly true Factual 16 Dec 2014 Brooks first presented his 1986 layered-architecture work in France. Factual · 16 Dec 2014 Their own work or organisation
Mostly true Factual 16 Dec 2014 The first industrial robots were installed in 1961, when a computer filled an entire room. Factual · 16 Dec 2014 Technology
Back to the list

Factual AI capabilities

“In essence ChatGPT is picking the most likely word to follow all those that precede it, whether it is the words in a human input prompt, or what it has already generated itself.”

Rodney Brooks, blog author 1 January 2023 Predictions Scorecard, 2023 January 01 (rodneybrooks.com)

In essence, ChatGPT picks the most likely next word given the preceding words, whether from the user's prompt or its own generated output.

Mostly true True in substance, wrong in a detail.

What would count as true or false

True: technical documentation from OpenAI or authoritative technical accounts show that ChatGPT generates text autoregressively, predicting the next token from all preceding tokens (the prompt plus its own earlier output), so that the description is accurate in substance. Mostly true: the autoregressive next-token mechanism is confirmed, and the only differences are minor technical details the claim glosses over: tokens rather than words, and sampling (temperature) rather than always taking the single most likely option, with RLHF shaping the probabilities. Mixed: the next-token mechanism is right, but a substantial part of how it works is materially misdescribed. Mostly false: the description misrepresents the core generation process in a significant way. False: ChatGPT does not generate text by predicting successive tokens conditioned on prior context.

Evidence

  1. What Is ChatGPT Doing … and Why Does It Work? Stephen Wolfram (writings.stephenwolfram.com) · 14 February 2023
    “The first thing to explain is that what ChatGPT is always fundamentally trying to do is to produce a “reasonable continuation” of whatever text it’s got so far, where by “reasonable” we mean “what one might expect someone to write after seeing what people have written on billions of webpages, etc.””
  2. What Is ChatGPT Doing … and Why Does It Work? Stephen Wolfram (writings.stephenwolfram.com) · 14 February 2023
    “Because when ChatGPT is going to generate a new token, it always “reads” (i.e. takes as input) the whole sequence of tokens that come before it, including tokens that ChatGPT itself has “written” previously.”
  3. What is LLM Temperature? | IBM IBM · 16 December 2024
    “LLMs generate text by predicting the next word (or rather, the next token) according to a probability distribution.”
  4. Introducing ChatGPT - OpenAI OpenAI · 30 November 2022
    “We trained this model using Reinforcement Learning from Human Feedback (RLHF), using the same methods as InstructGPT, but with slight differences in the data collection setup.”

Reasoning

Independent technical accounts confirm that ChatGPT generates text autoregressively: it produces a continuation of the text so far (E1), and each new token is generated from the whole preceding sequence, including its own earlier output (E2). The claim glosses over some minor technical details. The unit is a token rather than a word, and the next token is predicted according to a probability distribution (E3). The probabilities are also shaped by RLHF training (E4). These are the minor differences the criteria list for mostly true.

Cite this claim

Rodney Brooks, “In essence, ChatGPT picks the most likely next word given the preceding words, whether from the user's prompt or its own generated output.” (said 1 January 2023). Verdict: mostly true. Reliability API, claim c_cded391652d6, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_cded391652d6

Claims without a verdict

Is a verdict here wrong, or do you speak for Rodney Brooks? 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.