Extinction-path argument: superhuman hacking plus billions of AI agents

davidmanheim · x · 2026-09-15

davidmanheim pushes back on "AI won't kill literally everyone" with a concrete escalation chain: humans use AI to build vision/robotics systems, AI becomes superhuman at hacking, hackable systems get wired into military C&C and autonomous weapons, billions of AI agents get deployed — and then the outcome is profit or extinction. He argues the present already feels like sci-fi, but that doesn't mean the plot lets humans win.

Original post →

More from AGI Musings

AGI Musings channel →