Huang and Amodei split on AI safety at Dreamforce

Two of the biggest names in AI stood on the same stage hours apart and gave opposite answers about slowing down.

7 cards · 1m 40s · tap to begin

From · · · 1m 40s

Huang and Amodei split on AI safety at Dreamforce

Two of the biggest names in AI stood on the same stage hours apart and gave opposite answers about slowing down.

In brief

Two of the biggest names in AI stood on the same stage hours apart and gave opposite answers about slowing down. The AI safety fight is no longer about whether risk is real, it is about whether anyone should be forced to slow down. Originally reported by CNBC.

Same stage, opposite answers

Same stage, opposite answers

Dario Amodei and Jensen Huang both took the Dreamforce keynote with Marc Benioff on Tuesday, and landed on opposite sides of how fast AI should move.

Amodei wants a slowdown

Days earlier, Amodei published an essay asking the industry to ease off the pace of model development. He pitched it as leadership, not retreat.

“I think that's the way to lead the industry forward, to set an example”

A three-step plan with a catch

Anthropic's Amodei says China presents 'toughest dilemma' for his proposed AI slowdown

His proposal would temper how fast model capabilities improve without giving up commercial advantage or the US lead in AI. It followed researcher warnings about catastrophic harm.

Rivals lined up behind him

Sam Altman, Elon Musk and Demis Hassabis all voiced support, a rare moment of agreement among people who normally compete on release speed.

Huang calls it a false choice

The head of the world's most valuable company said speed and safety are not a tradeoff, and the industry needs no new laws because market forces already apply pressure.

“It's a false choice. You could definitely have both at the same time.”

Run fast, pause if needed

Huang put the judgment call on developers themselves: go full speed, then stop only if the company or the product feels out of control.

“Run as fast as you can”

Washington is already leaning in

Sen. Mark Warner: Have to take advantage of AI leaders calling for regulation

A day earlier Huang took a call from Trump on stage, who worried about people blocking AI data centers. Benioff's own line: slow down or speed up, but be held accountable.

“We're not going to let that happen, sir”

The short version

The AI safety fight is no longer about whether risk is real, it is about whether anyone should be forced to slow down.

Read the original on CNBC

React

Sign in to react and comment.

Comments (0)

Life is short. Keep it sweet. Respect others' opinions and be kind!

    Recommended next

    More decks on ai safety and related topics.