PeopleofSuperintelligence

Researcher

John Schulman

John Schulman is a researcher known for work on reinforcement learning and training language models with human feedback. He coauthored Trust Region Policy Optimization in 2015 and Proximal Policy Optimization in 2017, two methods for updating a learning system's behavior. OpenAI's founding announcement names him among its original researchers, and He received his doctorate from Berkeley in 2016. He later contributed to InstructGPT and ChatGPT with large research and engineering teams. He was a Thinking Machines Lab cofounder and chief scientist in May 2025. In a September 2025 paper, he studied efficient model adaptation with LoRA.

Works
53

Publication numbers from OpenAlex, as of 14 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Photo of John Schulman

Timeline

  1. 2015

    Making learning updates more stable

    Schulman and colleagues publish Trust Region Policy Optimization.

  2. 2016

    Completing a Berkeley doctorate

    He received a doctorate in electrical engineering and computer sciences from Berkeley in 2016.

  3. 2017

    A more practical training method

    With Filip Wolski, Prafulla Dhariwal, Alec Radford and Oleg Klimov, he publishes Proximal Policy Optimization.

  4. 2022

    Using feedback to follow instructions

    Schulman co-authors the InstructGPT paper with Long Ouyang and a large team.

  5. May 2025

    A new research setting

    He was a Thinking Machines cofounder and chief scientist in May 2025.

Sources and credits

Photo: Eric Louis Haines, via Berkeley Engineering. License: Not stated by source. Source: engineering.berkeley.edu. Cropped to a square and resized. Ask us to remove or replace this photo.

  1. Schulman and colleagues: TRPO, February 2015 arxiv.org. Checked 10 October 2026.
  2. OpenAI founding announcement, December 2015 openai.com. Checked 10 October 2026.
  3. Berkeley Engineering: young innovators, June 2018 engineering.berkeley.edu. Checked 10 October 2026.
  4. Schulman and colleagues: PPO, July 2017 arxiv.org. Checked 10 October 2026.
  5. Ouyang and colleagues: InstructGPT, March 2022 arxiv.org. Checked 10 October 2026.
  6. OpenAI: ChatGPT launch, November 2022 openai.com. Checked 10 October 2026.
  7. Berkeley: 2025 commencement program, page 19 engineering.berkeley.edu. Checked 10 October 2026.
  8. Schulman and Thinking Machines: LoRA Without Regret, September 2025 thinkingmachines.ai. Checked 10 October 2026.