Erik Hoel maps the EA split: pause camp vs labs' 'accelerate to control' camp

erikphoel · x · 2026-09-16

Erik Hoel argues EA and AI safety are split into two camps: the "pause/ban superintelligence" camp is the pro-human, less weird one, while the labs themselves are in the "accelerate so we can control the process according to our values" camp — the more weird and often anti-human one. Telling Anthropic to accelerate, he says, fuels the parts of EA you don't like. ScholarsStage endorsed the take.

Related event: AI Safety Community Splits: Pause Superintelligence Camp vs Lab Accelerationists(4 posts)→

Original post →

More from AGI Musings

AGI Musings channel →