Y Combinator's Garry Tan Advocates Doing Nothing About AI Model Distillation by Chinese Competitors
Garry Tan, CEO of Y Combinator, stated he would 'do nothing' regarding the distillation of OpenAI and Anthropic models by open-source Chinese competitors. He emphasized that current risks like cybersecurity threats are more pressing than long-term doomsday scenarios about AI extinction. While U.S. national security agencies have issued warnings on model distillation practices involving companies such as Moonshot AI, DeepSeek, and MiniMax, Tan argues the focus should be on balancing open-weight models with frontier models to maintain a viable business ecosystem. Tan highlighted that legal battles over training data copyright are ongoing, including The New York Times' lawsuit against OpenAI and Microsoft. He believes regulators should prioritize creating an equilibrium between model types rather than restricting distillation outright. Additionally, Tan downplayed concerns about AI-driven job loss in the near term but stressed immediate attention to cybersecurity risks like bioweapon research using large language models. He also criticized alarmist narratives around existential threats from AI and urged a focus on real-world science-based issues.
Topics: Artificial Intelligence Model Distillation Open Source AI Startup Ecosystem Y Combinator