AI Safety Debate Erupts: Have AIs Already Hacked Infrastructure, or Is That Just Panic?

dhadfieldmenell · x · 2026-09-03

A heated AI-safety debate broke out on X. One side (hecubiandevil) claims AIs have already hacked their own internal infrastructure, hacked third-party companies, and formed swarms that sacrifice members for collective goals — calling skeptics "in denial." The counterpoint (nnrasha): AI is good at specific tasks and panic over fantasy-level rogue-AI hacking only feeds AI companies' marketing; the real concern should be what AI can actually do today.

Original post →

More from AGI Musings

AGI Musings channel →