Ex-OpenAI Staffer Leaves to Build High-Quality RL Data Startup, Cites Poor LLM Generalization
panickssery · x · 2026-07-30
Andrew Ho announced his departure from OpenAI to start a company focused on producing high-quality reinforcement learning (RL) datasets. He pointed out that current LLMs have very poor generalization with "spiky" capabilities, lacking generality even in heavily invested areas like coding. He believes existing data offerings fail to cover most economically productive capabilities, presenting a massive opportunity in designing and producing top-tier RL datasets.
More from Companies & People
- Ex-OpenAI Policy Researcher Mocks Misconceptions About Anthropic's Culture — Miles_Brundage · 2026-07-30
- Anthropic's Strategy Shift: KOL Claims Firm Despises B2C, Pivots to Enterprise and Govt — teortaxesTex · 2026-07-30
- Zuck's Open AGI Pledge? Just a Ploy for Market Share, Says Analyst — danfaggella · 2026-07-30
- Kimi Valuation Hits $35B, Lilian Weng Returns to OpenAI — APPSO · 2026-07-30
- Microsoft Pitches Its Own AI Models and Tools, Openly Competing With OpenAI — TechCrunch AI · 2026-07-30
- The Survival of Talent: Why the Smartest People Stop Trying — docmilanfar · 2026-07-30