PeopleofSuperintelligence

Researcher

Hunter Lightman

Hunter Lightman is an OpenAI researcher who works on mathematical reasoning in language models and belongs to the research team behind the o1 reasoning model. A CIS major at the University of Pennsylvania in the class of May 2019, he joined OpenAI in 2022 and worked on MathGen, a team teaching models to solve high school math competitions. In 2023, he was first author of "Let's Verify Step by Step," which found that process supervision, feedback on each reasoning step, significantly outperforms outcome supervision on the MATH dataset, and which released the PRM800K dataset. He worked on Project Strawberry, which led to o1, released in the fall of 2024; he went on to coauthor the o1 and GPT-5 system cards and a 2025 study of reasoning models in competitive programming. Today he is a Member of Technical Staff at OpenAI.

Timeline

  1. 2019

    A CIS major at Penn

    At the University of Pennsylvania, he joins Theta Tau's Phi pledge class and majors in CIS, in the class of May 2019.

  2. 2022

    Joining OpenAI's MathGen team

    He joins OpenAI as a researcher in 2022, shortly before his colleagues launch ChatGPT. On the MathGen team, he works on teaching OpenAI's models to solve high school math competitions, at a time when the models are weak at mathematical reasoning.

  3. 2023–2024

    Process supervision and PRM800K

    In May 2023, he is first author of "Let's Verify Step by Step," written with coauthors including Jan Leike, John Schulman, Ilya Sutskever and Karl Cobbe. The paper finds that process supervision, which gives feedback on each intermediate reasoning step, significantly outperforms outcome supervision on the MATH dataset, and it releases PRM800K, a set of 800,000 step-level human feedback labels. It appears at the International Conference on Learning Representations (ICLR) in 2024.

  4. 2023–2024

    Project Strawberry and o1

    He works on Project Strawberry, OpenAI's 2023 breakthrough that combines language models, reinforcement learning and test-time computation and leads directly to o1. OpenAI releases o1, its first reasoning model, in the fall of 2024, and he discusses it as part of the o1 research team with Noam Brown and Ilge Akkaya on Sequoia Capital's Training Data podcast. He calls the breakthrough one of the most exciting moments of his research career.

  5. December 2024

    The o1 system card

    In December 2024, he is among the coauthors of the OpenAI o1 System Card, which describes how the o1 model series is trained with large-scale reinforcement learning to reason using chain of thought.

  6. February 2025

    Competitive programming with o3

    In February 2025, he coauthors "Competitive Programming with Large Reasoning Models," which compares OpenAI o1 and an early o3 checkpoint with o1-ioi, a system built with hand-engineered inference strategies for the 2024 International Olympiad in Informatics. The paper reports that o3 achieves a gold medal at IOI 2024 and a Codeforces rating on par with elite human competitors.

  7. August 2025

    The GPT-5 system card

    He coauthors the OpenAI GPT-5 System Card, published alongside the GPT-5 launch in August 2025. In an August 2025 TechCrunch interview, he names training models on less verifiable tasks as research he is excited about.

  8. As of October 2026

    Member of technical staff at OpenAI

    He is a Member of Technical Staff at OpenAI, the company he joined as a researcher in 2022.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  2. The United States employs an estimated 53% of people in our data with a known employment countryOur estimates of where AI careers are concentrated now and how countries’ shares of dated positions changed over time.
  3. 46% of living people in our data with known birth and employment countries work outside their birth countryOur estimate of cross-border careers in AI, from country of birth and the country of each current organization.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionOur estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Hunter Lightman yet. Send us a free-licensed photo.

  1. Hunter Lightman — Theta Tau UPenn upennthetatau.com. Checked 19 October 2026.
  2. Inside OpenAI’s quest to make AI do anything for you | TechCrunch techcrunch.com. Checked 19 October 2026.
  3. [2305.20050] Let's Verify Step by Step arxiv.org. Checked 19 October 2026.
  4. Let's Verify Step by Step | ML Anthology mlanthology.org. Checked 19 October 2026.
  5. Noam Brown and Team on Teaching LLMs to Reason | Sequoia Capital sequoiacap.com. Checked 19 October 2026.
  6. [2412.16720] OpenAI o1 System Card arxiv.org. Checked 19 October 2026.
  7. [2502.06807] Competitive Programming with Large Reasoning Models arxiv.org. Checked 19 October 2026.
  8. [2601.03267] OpenAI GPT-5 System Card arxiv.org. Checked 19 October 2026.
  9. Hunter Lightman - Google Scholar scholar.google.com. Checked 19 October 2026.