AI Safety Discourse May Give Labs an Excuse to Ignore Security Experts

BlancheMinerva · x · 2026-09-19

A debate about the real-world impact of recent AI safety commentary. The poster argues there was a chance frontier labs would finally start listening to computer security experts, but the net effect of these comments (as measured by a biased Twitter feed) was to spread significant distrust — in a way that could easily justify continuing to ignore security experts. They also admit initially misreading the envisioned scenario as a model gaining the ability to send information onto the internet and cause harm.

Original post →

More from AGI Musings

AGI Musings channel →