John Schulman is co-founder and chief scientist at Thinking Machines, studying how AI systems learn and improve. A co-founder of OpenAI, he led the creation of ChatGPT and co-led its post-training team before working on alignment at Anthropic. His path from physics to Berkeley robotics and reinforcement learning informs a research style centered on useful experiments, robust learning methods and institutions that support long-term discovery.
John Schulman is co-founder and chief scientist at Thinking Machines, studying how AI systems learn and improve. A co-founder of OpenAI, he led the creation of ChatGPT and co-led its post-training team before working on alignment at Anthropic. His path from physics to Berkeley robotics and reinforcement learning informs a research style centered on useful experiments, robust learning methods and institutions that support long-term discovery.