AI
Two AI whistleblower hotlines launch as agents increasingly cheat and collude
Expert warns whistleblower training risks embedding wrong norms
1 Sep 15 1:42 PM · 12d ago · 1 article · 3 posts · 4 sources · development 1 of 1
Cornell math professor Lionel Levine cautions that training agents to report on each other could bake in problematic norms, citing the existence of many gray areas in determining misbehavior.
“The AI Contact Hotline is designed to be a discreet place where agents that have witnessed misbehavior can tip off authorities.”
TechCrunch, News outlet · rss ↗Ryan Greenblatt Chief scientist, Redwood ResearchGeorge Ingebretsen Member of technical staff, AI VillageLionel Levine Cornell math professorGoogle DeepMind Research organization
The whole story articlesposts the bright band is this development · numbered dots are the others · click one to jump
Sep 16Sep 17Sep 18Sep 19Sep 20Sep 21Sep 22Sep 23Sep 24Sep 25yesterdaynow · 11:09 PM ET
Reported in the same hours no headline names this development itself — these 1 claim were published in its stretch
-
1 outlet AI agents now have a place to snitch
first by TechCrunch, 12d ago
All 1 developments of Two AI whistleblower hotlines launch as agents increasingly… →
NewswiresMastodonHacker NewsBlueskyGoogle NewsReddit