Jensen Huang said two things on Lex Fridmans podc

Jensen Huang said two things on Lex Fridman’s podcast that together are more interesting than either alone: I think we’ve achieved AGI and that the odds of AI agents building Nvidia is 0%. I think he’s right on both counts, and the gap between those two statements is basically the entire problem space of enterprise AI right now. AI can build an app that a billion people use for a week. It can write code, pass exams, generate content. But it can’t coordinate 50 people toward a goal over 6 months, it can’t decide which project to kill, and it can’t build trust with a customer over a multi-year relationship - yet. I think that changes in 2026.