AI safety figures clash over 'human extinction' meme: fearmongering or ignored threat?

round · x · 2026-09-10

Quoting Nathan Lambert's post that a human-extinction risk meme being viewed 100M times is "net bad" fearmongering absent evidence supporting even a 10% extinction chance, gfodor pushed back: the extinction meme is a thought-terminating infohazard that distracts from concrete threats — like a terrorist cell killing 10-20M people in 2028 with a chemical weapon invented by a model, or a kid shutting down a city power grid with a phone.

The exchange captures a real split in the AI safety community over how risk should be talked about: abstract extinction narratives vs. specific, tractable disaster scenarios.

Related event: AI community debates extinction-risk messaging(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →