PeopleofSuperintelligence

Researcher

Andrew M. Bean

Also known as Andrew Bean, Andrew Michael Bean

Andrew M. Bean studies cooperation between people and AI and how language models should be evaluated. He earned a BSc in Applied Mathematics at Yale in 2017, developed fixed-income algorithms at Bridgewater, and later worked in data science, completed an Oxford MSc, taught AI ethics at Stanford, and trained a custom LLM at Thomson Reuters. His research includes LingOly, a multilingual reasoning benchmark; PRISM, a dataset of human preferences and LLM feedback that won NeurIPS 2024's Best Paper Award in Datasets and Benchmarks; and a review that found problems with LLM benchmark validity. In 2026, he coauthored a randomized study of medical chatbots and is a DPhil student at Oxford and a Senior Researcher in the Thomson Reuters–Imperial Frontier AI Research Lab.

Now at
HealthSherpa
Badges
Links
Website, Google Scholar, LinkedIn and GitHub

Timeline

  1. 2017–2021

    Yale mathematics and Bridgewater

    He earns a BSc in Applied Mathematics at Yale in 2017. From 2017 to 2021, he works as an investment associate at Bridgewater Associates and develops proprietary algorithms for fixed-income markets.

  2. 2021

    Machine learning at HealthSherpa

    In 2021, he works as a data science consultant at HealthSherpa and trains machine-learning models to assess investment opportunities.

  3. 2022

    An Oxford MSc thesis

    He completes an MSc in Social Data Science at Oxford's Internet Institute in 2022. His thesis, Teaching AI to Cooperate with Human Partners through Rule-based Play, studies rule-based cooperation between people and AI.

  4. 2023–2025

    Stanford teaching and Thomson Reuters

    From 2023 to 2025, he teaches at Stanford's Bing Overseas Study Program and designs an undergraduate course on AI ethics. In 2025, he joins Thomson Reuters' Foundational Research team as a research scientist intern and trains a custom LLM for client applications.

  5. June 2024

    LingOly tests language reasoning

    In June 2024, he coauthors LingOly, a benchmark of 1,133 puzzles across more than 90 mostly low-resource languages. It tests LLMs on linguistic patterns and complex task instructions; the top model reaches 38.7% on harder problems.

  6. December 2024

    PRISM wins a NeurIPS award

    In December 2024, he coauthors PRISM, a dataset linking preferences and feedback from 1,500 people in 75 countries to 8,011 conversations with 21 LLMs. It wins NeurIPS 2024's Best Paper Award in the Datasets and Benchmarks track.

  7. November 2025

    Reviewing LLM benchmarks

    In November 2025, he coauthors a systematic review of 445 LLM benchmarks with 29 expert reviewers. The team finds patterns in measured phenomena, tasks and scoring metrics that undermine benchmark claims.

  8. As of October 2026

    DPhil student and lab researcher

    He has been a DPhil student in Social Data Science at Oxford's OII since October 2022, studying cooperative AI. In February 2026, he coauthors a randomized study of medical chatbots with 1,298 UK participants selected to represent the adult population; chatbot users identify conditions in less than 34.5% of cases and the right course of action in less than 44.2%. As of October 2026, he is a Senior Researcher in the Thomson Reuters–Imperial Frontier AI Research Lab, working on LLM evaluation, construct validity and algorithmic bias.

Career

  1. From 2025Thomson ReutersResearch scientist intern
  2. 2023–2025Stanford University
  3. From 2021HealthSherpaData science consultant
  4. 2017–2021Bridgewater AssociatesInvestment associate

Education

  1. 2022University of OxfordMSc in Social Data Science
  2. From 2022University of OxfordDPhil student in Social Data Science
  3. 2017Yale UniversityBSc in Applied Mathematics

From the biography.

  1. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Stanford University. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.
  2. Google had the most people shaping AI every year from 2014 to 2025Includes Stanford University. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  3. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Stanford University. We estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.
  4. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Stanford University. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Andrew M. Bean yet. Send us a free-licensed photo.

  1. cv | Andrew M. Bean am-bean.github.io. Checked 21 October 2026.
  2. OII | Andrew Bean oii.ox.ac.uk. Checked 21 October 2026.
  3. OSPOXFRD 85: Practical Ethics for Artificial Intelligence (Way: ER) | Stanford in Oxford oxford.stanford.edu. Checked 21 October 2026.
  4. LINGOLY: A Benchmark of Olympiad-Level Linguistic Reasoning Puzzles in Low-Resource and Extinct Languages arxiv.org. Checked 21 October 2026.
  5. Andrew M. Bean am-bean.github.io. Checked 21 October 2026.
  6. NeurIPS 2024 Awards neurips.cc. Checked 21 October 2026.
  7. Measuring what Matters: Construct Validity in Large Language Model Benchmarks arxiv.org. Checked 21 October 2026.
  8. OII | New study warns of risks in AI chatbots giving medical advice oii.ox.ac.uk. Checked 21 October 2026.
  9. Reliability of LLMs as medical assistants for the general public: a randomized preregistered study syntheticsociety.oii.ox.ac.uk. Checked 21 October 2026.
  10. People | Research groups | Imperial College London imperial.ac.uk. Checked 21 October 2026.