Y Combinator’s Garry Tan wants U.S. open-weight AI labs to ‘distill’ frontier models, too
Y Combinator's Tan pushes for open distillation of frontier models—reframing AI access as public infrastructure.
Why it matters
A high-profile VC is publicly arguing for policy that would require frontier labs to enable open-weight distillation of their models. This is a frontier-lab strategy and policy play that affects how capability gets distributed, with implications for open-source competition and regulatory strategy.
The key facts
4 to knowGarry Tan (Y Combinator) advocates open-weight labs should have distillation access to frontier models
Argument: frontier models trained on public human knowledge; access should be 'a form of public good'
Policy angle: reframes distribution and access strategy for the lab-race era
Timing: September 2026—ongoing frontier model competition and open-source acceleration
Go to the source
TechCrunch AItechcrunch.com
Publisher excerpt: Tan argues that frontier models themselves trained on public human knowledge so access to capable AI should be "a form of public good."