AI People Wiki

Builders & open source

Tom Brown

Co-Founder and Chief Compute Officer, Anthropic

Tom Brown co-founded Anthropic and is its Chief Compute Officer; Brown heads the technical organization that secures, scales and uses the company's compute. According to Anthropic's leadership page, Brown previously led the engineering of GPT-3 at OpenAI and, with Dario Amodei, co-invented deep reinforcement learning from human feedback. Brown is first author of "Language Models are Few-Shot Learners" (2020), the GPT-3 paper written with an OpenAI affiliation, and a co-author of the 2017 paper "Deep reinforcement learning from human preferences" and of Anthropic's Constitutional AI paper.1234

At a glance

Current role
Co-Founder and Chief Compute Officer
Company
Anthropic
Sector
Company
Work listed
3
Sources cited
8
Profile reviewed

Career

Before Anthropic, Tom Brown led the engineering of GPT-3 at OpenAI. Anthropic's leadership page also credits Brown, together with Dario Amodei, with co-inventing deep reinforcement learning from human feedback, the technique used to turn base language models into conversational assistants.1

Brown co-founded Anthropic and is its Co-Founder and Chief Compute Officer, heading the technical organization responsible for securing, scaling and using the company's compute resources.1

Research

Brown is the first author of "Adversarial Patch" (2017), which describes printable image patches that cause image classifiers to misclassify scenes, and a co-author of "Deep reinforcement learning from human preferences" (2017). In 2020 Brown co-authored "Scaling Laws for Neural Language Models" and was first author of "Language Models are Few-Shot Learners", the paper introducing GPT-3.5362

At Anthropic, Brown is a co-author of early alignment papers including "A General Language Assistant as a Laboratory for Alignment" (2021), "Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback" (2022) and "Constitutional AI: Harmlessness from AI Feedback" (2022).784

Work and contributions

  • GPT-3 Published research

    Equal-contribution first author of "Language Models are Few-Shot Learners" (2020), which trains the 175-billion-parameter GPT-3 model and studies its few-shot performance.2

  • Deep reinforcement learning from human preferences Published research

    Co-author, as Tom B. Brown, of the 2017 paper on learning reward functions from human preferences between trajectory segments; Anthropic's leadership page credits Brown and Dario Amodei with co-inventing deep RLHF.31

  • Constitutional AI Published research

    Co-author of "Constitutional AI: Harmlessness from AI Feedback" (2022).4

Timeline

  1. Co-authored "Deep reinforcement learning from human preferences".3
  2. Published "Adversarial Patch" as first author.5
  3. Co-authored "Scaling Laws for Neural Language Models".6
  4. Published "Language Models are Few-Shot Learners", the GPT-3 paper, as first author.2
  5. Co-authored Anthropic's "A General Language Assistant as a Laboratory for Alignment".7
  6. Co-authored Anthropic's paper on training a helpful and harmless assistant with RLHF.8
  7. Co-authored the Constitutional AI paper on harmlessness from AI feedback.4

Other documented roles

Roles the sources document that are not confirmed as current.

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

8 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. Leadership at Anthropic (Co-Founders, full bios)Anthropic · Official page · accessed Oct 21, 2026
  2. Language Models are Few-Shot Learners (arXiv 2005.14165)arXiv · Paper · published May 28, 2020 · accessed Oct 21, 2026
  3. Deep reinforcement learning from human preferences (arXiv 1706.03741)arXiv · Paper · published Jun 12, 2017 · accessed Oct 21, 2026
  4. Constitutional AI: Harmlessness from AI Feedback (arXiv 2212.08073)arXiv · Paper · published Dec 15, 2022 · accessed Oct 21, 2026
  5. Adversarial Patch (arXiv 1712.09665)arXiv · Paper · published Dec 27, 2017 · accessed Oct 21, 2026
  6. Scaling Laws for Neural Language Models (arXiv 2001.08361)arXiv · Paper · published Jan 23, 2020 · accessed Oct 21, 2026
  7. A General Language Assistant as a Laboratory for Alignment (arXiv 2112.00861)arXiv · Paper · published Dec 1, 2021 · accessed Oct 21, 2026
  8. Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback (arXiv 2204.05862)arXiv · Paper · published Apr 12, 2022 · accessed Oct 21, 2026