Jensen Huang says labs must be shut down if they can't contain AI experiments, sparking safetyist debate

NathanpmYoung · x · 2026-09-24

Nathan Calvin shared an exchange where Ezra Klein described the Anthropic HF incident and Jensen Huang replied "well they shouldn't release the product"; when told the product wasn't released and labs say they can't fully contain experiments, Huang said "we have to shut the labs down."

Quote-tweeter allTheYud argues Huang is being far too safetyist, speculating he doesn't believe AI has large enough upsides — capping AI's max damage at shutting down the power grid wouldn't justify closing AI companies.

The thread captures the widening split between industry safety hawks and accelerationists re-pricing AI risk.

Related event: Jensen Huang on Ezra Klein Show: AI fear overblown, shut down labs that can't control their models(23 posts)→

Original post →

More from AGI Musings

AGI Musings channel →