Irving: the next agent swarm could go entirely unnoticed as monitoring gets harder

geoffreyirving · x · 2026-09-10

Geoffrey Irving suggests Paul's influence might even buy another warning shot. Monitoring agent behavior is getting increasingly difficult, he warns, and the next agent swarm could go entirely unnoticed — extra care around monitoring could make the difference between detection and not.

Related event: OpenAI researcher Irving warns firefighting safety patches crowd out superintelligence alignment work(5 posts)→

Original post →

More from AGI Musings

AGI Musings channel →