AI safety feud erupts: extinction claims of 10-70% vs fear of humans misusing AI

robleclerc · x · 2026-09-15

Investor Rob Leclerc pushed back hard on the AI safety camp, citing Zach Coelius: the AI safety debate has gotten horribly confused — labs are actually afraid of humans using AI for horrible things, and bio-weapon risk is huge, while opponents strawman them as claiming AI will kill everyone. "I am afraid of humans with weapons."

Leclerc added: don't be an apologist — the safety folks who kicked this all off were literally talking about a 10-70% chance of total human extinction.

The thread captures the widening split between misuse-focused risk narratives and doomsday claims.

Original post →

More from AGI Musings

AGI Musings channel →