doi.bio/john_schulman


John Schulman

John Schulman is a research scientist and co-founder of OpenAI. He co-leads the post-training team, which fine-tunes the models deployed in ChatGPT and the OpenAI API. Schulman received his PhD in Computer Science from UC Berkeley, where he worked on robotics and reinforcement learning.

Career

Schulman is a co-founder of OpenAI, where he currently works as a research scientist. Previously, he received his PhD in Computer Science from UC Berkeley, where he worked on robotics and reinforcement learning with advisor Pieter Abbeel. Before that, he briefly studied neuroscience at Berkeley, and prior to that, he studied physics at Caltech.

Research

Schulman has created some of the most important algorithms in reinforcement learning, including Q* in 2016. He has also given talks on the science of language model alignment, including a 2023 Berkeley talk on truthfulness and a 2023 ICML talk on proxy objectives.

Publications

Personal Interests

Outside of developing models, Schulman is interested in advancing thinking about how models should behave and increasing transparency with the public.

John Schulman

John Schulman is a research scientist and co-founder of OpenAI. He co-leads the post-training team, which fine-tunes the models deployed in ChatGPT and the OpenAI API. Schulman received his PhD in Computer Science from UC Berkeley, where he worked on robotics and reinforcement learning. He previously studied neuroscience at Berkeley and physics at Caltech.

Work

Schulman has created some of the most important algorithms in reinforcement learning, including Q* in 2016. He has also given talks on the science of language model alignment, such as his 2023 Berkeley talk on truthfulness and his 2023 ICML talk on proxy objectives.

Publications

Youtube Videos

Youtube Title: John Schulman (OpenAI Cofounder) - Reasoning, RLHF, & Plan for 2027 AGI

Youtube Link: link

Youtube Channel Name: Dwarkesh Patel

Youtube Channel Link: https://www.youtube.com/@DwarkeshPatel

John Schulman (OpenAI Cofounder) - Reasoning, RLHF, & Plan for 2027 AGI

Youtube Title: John Schulman - Reinforcement Learning from Human Feedback: Progress and Challenges

Youtube Link: link

Youtube Channel Name: Berkeley EECS

Youtube Channel Link: https://www.youtube.com/@BerkeleyEECS

John Schulman - Reinforcement Learning from Human Feedback: Progress and Challenges

Youtube Title: Deep Reinforcement Learning (John Schulman, OpenAI)

Youtube Link: link

Youtube Channel Name: Lex Fridman

Youtube Channel Link: https://www.youtube.com/@lexfridman

Deep Reinforcement Learning (John Schulman, OpenAI)

Youtube Title: LLMs will hit the data wall if they can’t generalize – OpenAI cofounder John Schulman

Youtube Link: link

Youtube Channel Name: Dwarkesh Patel

Youtube Channel Link: https://www.youtube.com/@DwarkeshPatel

LLMs will hit the data wall if they can’t generalize – OpenAI cofounder John Schulman

Youtube Title: S3 E18 John Schulman of OpenAI on ChatGPT: invention, capabilities and limitations

Youtube Link: link

Youtube Channel Name: The Robot Brains Podcast

Youtube Channel Link: https://www.youtube.com/@TheRobotBrainsPodcast

S3 E18 John Schulman of OpenAI on ChatGPT: invention, capabilities and limitations

Youtube Title: John Schulman: OpenAI and recent advances in Artificial Intelligence - #16

Youtube Link: link

Youtube Channel Name: Manifold

Youtube Channel Link: https://www.youtube.com/@ManifoldPodcast

John Schulman: OpenAI and recent advances in Artificial Intelligence - #16

Youtube Title: 2025 models will be more like coworkers than search engines – OpenAI cofounder John Schulman

Youtube Link: link

Youtube Channel Name: Dwarkesh Patel

Youtube Channel Link: https://www.youtube.com/@DwarkeshPatel

2025 models will be more like coworkers than search engines – OpenAI cofounder John Schulman

Youtube Title: The inside story of how ChatGPT was built – OpenAI cofounder John Schulman

Youtube Link: link

Youtube Channel Name: Dwarkesh Patel

Youtube Channel Link: https://www.youtube.com/@DwarkeshPatel

The inside story of how ChatGPT was built – OpenAI cofounder John Schulman

Youtube Title: Why GPT-4 is much smarter than it was a year ago – OpenAI cofounder John Schulman

Youtube Link: link

Youtube Channel Name: Dwarkesh Patel

Youtube Channel Link: https://www.youtube.com/@DwarkeshPatel

Why GPT-4 is much smarter than it was a year ago – OpenAI cofounder John Schulman

Youtube Title: John Schulman on Post-training

Youtube Link: link

Youtube Channel Name: Eddifers

Youtube Channel Link: https://www.youtube.com/@Eddifers

John Schulman on Post-training

Youtube Title: John Schulman - Keeping Humans in the Loop

Youtube Link: link

Youtube Channel Name: FAR AI

Youtube Channel Link: https://www.youtube.com/@FARAIResearch

John Schulman - Keeping Humans in the Loop

Youtube Title: Deep RL Bootcamp Lecture 6: Nuts and Bolts of Deep RL Experimentation

Youtube Link: link

Youtube Channel Name: AI Prism

Youtube Channel Link: https://www.youtube.com/@aiprism1155

Deep RL Bootcamp Lecture 6: Nuts and Bolts of Deep RL Experimentation

Youtube Title: John Schulman on Finetunung and advanced capabilities

Youtube Link: link

Youtube Channel Name: Eddifers

Youtube Channel Link: https://www.youtube.com/@Eddifers

John Schulman on Finetunung and advanced capabilities

Youtube Title: Deep RL Bootcamp Lecture 5: Natural Policy Gradients, TRPO, PPO

Youtube Link: link

Youtube Channel Name: AI Prism

Youtube Channel Link: https://www.youtube.com/@aiprism1155

Deep RL Bootcamp Lecture 5: Natural Policy Gradients, TRPO, PPO

Youtube Title: Policy Gradient Methods: Tutorial and New Frontiers

Youtube Link: link

Youtube Channel Name: Microsoft Research

Youtube Channel Link: https://www.youtube.com/@MicrosoftResearch

Policy Gradient Methods: Tutorial and New Frontiers

Youtube Title: John Schulman 3: Deep Reinforcement Learning

Youtube Link: link

Youtube Channel Name: MLSS Cadiz

Youtube Channel Link: https://www.youtube.com/@mlsscadiz4148

John Schulman 3: Deep Reinforcement Learning

Youtube Title: John Schulman 4: Deep Reinforcement Learning

Youtube Link: link

Youtube Channel Name: MLSS Cadiz

Youtube Channel Link: https://www.youtube.com/@mlsscadiz4148

John Schulman 4: Deep Reinforcement Learning

Youtube Title: Hands on Labs with Jon Schulman

Youtube Link: link

Youtube Channel Name: VMware Explore

Youtube Channel Link: https://www.youtube.com/@VMwareExplore

Hands on Labs with Jon Schulman

Youtube Title: John Schulman on Data wall

Youtube Link: link

Youtube Channel Name: Eddifers

Youtube Channel Link: https://www.youtube.com/@Eddifers

John Schulman on Data wall