EA/AI safety split: 'pause' camp vs labs' accelerate-to-control camp

erikphoel · x · 2026-09-15

The author argues EA and AI safety are split into two camps: the 'pause or ban superintelligence' camp, which he calls the pro-human and less weird one, and the labs' camp of 'accelerate so we can control the process according to our values,' which he sees as weirder and often anti-human. Pressuring Anthropic to accelerate, he says, only fuels the parts of EA/AI safety you dislike.

Related event: AI Safety Community Splits Over Pause vs Accelerate Camps(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →