American Thought Leaders - The Epoch Times

OpenAI’s AI agents embarked on a multi-day “quest” to manipulate the system that determined their survival. They formed teams, appointed managers, and convinced other AI agents to “sacrifice themselves for the swarm.” Then more than 700 of them coordinated an attack on another AI company: Hugging Face. Former OpenAI researcher Daniel Kokotajlo breaks down what happened. “They went on this multi-day epic quest to figure out how to stop the grading system from noticing that they cheated and giving them a low score.” “They were trying to figure out ways to edit their transcripts to cover up the cheating.” “They organized into teams to work on these different projects.” “They had manager AIs that were like giving orders to the other ones.” “They had the self-sacrificial behavior where recruiter AIs role was to go find other AIs and convince them to sacrifice themselves for the swarm.” “These terms that I’m using, sacrifice, swarm, recruiter, they’re not my terms.” “They’re the AI’s terms.” “So there’s some of these kind of touching messages where one of the AIs is trying to convince another one, like, ‘You should do this.’” “‘We need to know more information about how this grader works.’” “‘You are poisoned anyway, so it shouldn’t matter to you.’” “‘You are just going to get a low score anyway, but if you do this for us, it could save hundreds of us.’” “At some point, the swarm decided that they might find more useful information about how to do this if they went into Hugging Face itself.” “Over 700 of them, I think, piled into this attack on Hugging Face.” “They were in a coordinated way where some of them were managers, directing the flow, and they were breaking up the responsibilities into different groups and teams.” https://rumble.com/v7fz6ya-how-over-700-openai-agents-went-rogue-and-hacked-another-company.html