Jensen Huang: The Doomer Hoax, Superintelligence is Here, and The Future of AI (ft. President Trump)

Two days after Dario Amodei argued the frontier should be paced, Huang is asked on stage at the All-In Summit what to make of the number going around: a greater-than-10% chance AI kills everyone within the decade. His answer is that the figure is made up, and that hearing it from researchers inside a lab is what makes it irresponsible. The case he builds is a track record: radiology gone in five years, ninety percent of code written by AI within six to twelve months, half of entry-level jobs wiped out, GPT-2 and Llama 3 too dangerous to release. All wrong, and nobody accountable for having been wrong.

What the title obscures is that two people are speaking. Huang is careful in places the framing is not: on Jacob Coxon he says we do not know what he saw, he treats loss of control as a separate question deserving a separate answer, and on regulation he concedes that every problem so far has come from the frontier labs, because they have the most compute and are working on the hardest problems. The line about a hoax belongs to President Trump, who is put on speakerphone partway through and is talking mostly about data-center permitting and robots not taking over.