$25
Jersey Tee · from
Cut
Colourway
Size
Size guidePrinted on demand, usually ships in 2-5 business days. US only. Shipping is calculated at checkout from what's in your cart. Tees and Pro Jerseys ship separately from different US print shops.
Why #17
PPO, his most-used algorithm, hit arXiv in 2017.
Coming soonSeries 1 · Card 153/211
The trading card
Every player in the league gets a card. The front shows the kit and the back has the stat line: transfer history, scouting report and the story behind the number. Tap it to flip it over.
Printed cards aren't on sale yet. They're coming as random packs and team packs.
See the full setScouting report
Known for
- 01Invented TRPO (2015) and PPO (2017)
- 02OpenAI cofounder who led the RL team behind ChatGPT
- 03Chief scientist at Thinking Machines Lab
Transfer history
UC Berkeley (PhD)
2015–2024
OpenAI
2024–2025
Anthropic
2025–now
Thinking Machines Lab
Current club
Fact check
Receipts
Every fact on this page, with a link. Data as of 2026-10. Spot something stale? Tell us.
- TRPO, PPO, ChatGPT RL lead, career path, educationen.wikipedia.org ↗ (opens in a new tab)
- Cofounder and chief scientist at Thinking Machines (own homepage)people.eecs.berkeley.edu ↗ (opens in a new tab)
- Still chief scientist at Thinking Machines, April 2026thenextweb.com ↗ (opens in a new tab)
- Gave Inkling training timeline as TML staff, July 2026news.smol.ai ↗ (opens in a new tab)