Y Combinator’s Garry Tan wants U.S. open-weight AI labs to ‘distill’ frontier models, too

September 11, 2026

Tan argues that frontier models themselves trained on public human knowledge so access to capable AI should be "a form of public good."