The most powerful chipmaker in the AI race just told the doomers to show their work, with a presidential assist.
The Summary
- Nvidia CEO Jensen Huang said AI extinction predictions lack scientific grounding during Monday's All-In Summit in Los Angeles
- Trump called Huang mid-interview via speakerphone, calling AI doomerism "a hoax" and declaring "the robots will not be taking over"
- Huang praised former Anthropic researcher Jacob Coxon for whistleblowing courage while drawing a line between exposing company problems and making unscientific future predictions
- The US can "lead and do it safely" on AI without choosing between speed and safety, according to Huang
The Signal
The timing matters. Jacob Coxon resigned from Anthropic last week, writing on X that the people building AI "believe that it could kill us all by the end of the decade." Within days, the CEO of the company supplying the hardware for that AI buildout is on stage calling the existential threat claims unscientific. Not wrong. Not overblown. Unscientific.
Huang's framing is careful. He called AI safety "paramount" and said Coxon showed "great courage" as a whistleblower. He's not dismissing safety concerns. He's separating two things: concrete problems you can point to inside a lab versus predictions about AI killing everyone by 2030.
"I think the whistleblowing is fine. I think the science behind the doom predictions is not."
Then Trump calls in. Mid-interview. On speakerphone. The President tells the All-In Summit crowd that AI doomerism is "a hoax," that robots won't be taking over, and that while "we have to be a little bit careful," the US shouldn't pump the brakes on AI development. It's theater, but it's also policy signaling from the executive branch to Silicon Valley: build fast, we've got your back.
This isn't a random tech CEO opining. Nvidia's H100 and B200 chips are the literal substrate of the agent economy. Every foundation model, every reasoning system, every company racing toward AGI is buying Huang's hardware. When he says doomer predictions aren't grounded in science, he's not just offering an opinion. He's telling the builders financing his quarterly earnings that they can keep building without existential guilt.
Key dynamics at play:
- Huang supports whistleblowing on concrete safety issues but rejects unfalsifiable extinction scenarios
- Trump's call signals federal alignment with acceleration over AI safety regulation
- The man selling the shovels in the AI gold rush just told miners the mountain won't collapse
The Implication
Watch how this reframes the safety debate. Huang just gave every AI lab a playbook: take safety seriously, welcome whistleblowers on real problems, but don't let doomer scenarios slow shipping. Coxon gets praise for courage. His extinction timeline gets dismissed as unscientific. That's the new center.
For builders in the agent economy, this is permission structure. The hardware supplier and the White House just aligned on "move fast, be careful, but keep moving." If you're building AI agents, infrastructure, or applications, the message is clear: safety is on you, but existential handwringing won't get federal sympathy. Ship or get lapped.