tech
Exclusive: Nvidia's Jensen Huang Defends Chinese AI Amid Kimi Panic
Huang told Axios that American companies should "absolutely" be allowed to use Chinese models.
TL;DR
- Nvidia CEO Jensen Huang believes America has more to fear from the campaign to ban Chinese open-source AI models than from the models themselves.
- Huang argues that open-source models expand the market, give more people access to AI, and ultimately increase demand for the hardware Nvidia sells.
- He contends that restricting open models in the name of national security could make America more vulnerable, as openness allows for broader inspection and defense building.
- Huang called for Anthropic to make its cyber model, Claude Mythos, available to everyone, arguing that restricting access is not beneficial to the United States.
- Huang dismisses the idea of an AI race with an endpoint, stating that AI usage and competition between the US and China will continue indefinitely.
- He suggests that misconduct like violating privacy or contracts should be targeted, rather than the models themselves.