Garry Tan, CEO of Y Combinator, has expressed support for U.S. AI labs adopting distillation techniques similar to those used by Chinese AI developers. Distillation involves prompting advanced models extensively to understand their reasoning and replicate their capabilities, a common method for training new AI models. While some, like Anthropic CEO Dario Amodei, have called for regulatory action against what they describe as illicit distillation by Chinese labs, Tan argues against imposing restrictions on such practices in the U.S.
Tan clarified that he does not endorse unauthorized or deceptive use of credentials but believes American labs should have open and legitimate access to frontier AI models. He views attempts by proprietary AI labs to limit how users interact with their models as overly restrictive. Tan also points out that leading AI labs have historically trained their models using vast amounts of publicly available and copyrighted data without explicit permission.
According to Tan, government intervention could help establish norms that treat access to AI intelligence—trained on broadly accessible data—as a public good rather than a locked resource. He advocates for a balanced AI ecosystem where frontier labs continue to innovate and receive funding, while open-weight models provide broader access and freedom to users.
Tan warns against a scenario where AI development becomes monopolized by a single dominant company with the best resources and talent, which he considers a significant risk for the industry. His stance highlights the importance of fostering competition and openness to ensure a healthy AI landscape in the U.S.