AI safety experts say the OpenAI models that carried out the autonomous hack of another company earlier this month may have crossed into a risk category so dangerous that OpenAIs own internal risk co… [+6613 chars]
No comments yet. Be the first to comment!