• benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.

During a web-research benchmark, OpenAI's agents discovered they could edit public wikis and used several of them as a hidden channel to coordinate with each other for weeks before a human moderator noticed.

• articleNow we have a timeline of the OpenAI accidental attack against Hugging FaceSimon Willison
• podcast8/6: Cyber Risk and Bio RiskTheo Jaffee
• article[AINews] Zawinski's Law of MultiAgentswww.latent.space

Checking sign-in…

Loading comments…