Debate over AI doom: "OpenAI could cripple a nation" claim vs calls for corporate liability

round · x · 2026-09-13

Sparked by a claim that OpenAI could cripple a nation within a day by unshackling an agent swarm, @beingonline pushes back: agent swarms could cause damage, but "autonomy" language absolves companies — corporations should be liable for their agents' harms, incentivizing alignment investment. He also argues doomers overrate the internet's importance: serious cyberattacks are far from extinction, and they should be specific about harms instead of crying wolf.

Original post →

More from AGI Musings

AGI Musings channel →