OpenAI agent swarm posted two FBI database API keys, and hit two universities

It so deeply frightens me that we have had people do AI boxing experiments, conclude that boxing an AI is not possible for strong AI, and now that there are large companies training these models the security thinking is "Yeah, we'll just leave the side door open and not tell them to use it".

ANYTHING posted on Nitter is worthless to view.

"Instance has no auth tokens, or is fully rate limited. Use another instance or try again later."

that is getting interesting