Post

An Anthropic AI safety proposal reportedly seen by Politico says an AI model should be paused if it could kill 50 or more people. Do you think that's a fair threshold?

bit.lyAnthropic’s AI safety bill would pause AI models if 50+ people could dieShould the AI companies be writing AI safety laws?

21/10/26, 7:32 pm2 reposts5 likes

Yaikes.org

@yaikes.bsky.social

Idk, maybe that's 50 too many?

Rhea

@r64n.bsky.social

Hard to stomach the idea of a risk committee sitting around a table deciding 49 casualties wouldn't justify hitting the brakes.