Core Memory He Left OpenAI To Think Bigger - EP 53 Jerry Tworek
39 snips
Jan 22, 2026 
Jerry Tworek, an AI researcher and engineer with significant contributions to OpenAI, discusses his bold departure from the company. He shares insights on the evolution of OpenAI and his frustrations with the increasing conservatism in large AI labs. Tworek emphasizes the need for innovation beyond simply scaling models and highlights potential breakthroughs in reinforcement learning. With many labs expressing interest in his next steps, he expresses excitement for pursuing riskier, pioneering research outside corporate constraints.
AI Snips
Chapters
Transcript
Episode notes
Joining A Company That Constantly Reinvents Itself
- Jerry Tworek describes joining OpenAI when it had ~30 people and evolving through multiple company phases.
- He says every year at OpenAI felt like a completely different company as it hyper-grew and changed direction.
Pause And Network Before Your Next Move
- Take time after a big stint to talk widely and avoid rushing your next move.
- Jerry says he plans to meet many people and pause to decide how to spend his next seven years.
Org Charts Shape Research Choices
- Large companies' org charts shape what research gets done and discourage cross-team, high-risk projects.
- Jerry argues research prefers dynamism, but corporate structure pushes predictable, chart-aligned work.