But that’s an objection that OpenAI should take some easy action, the framing that OpenAI has out of control agents is still true.
replySeems you think there is a silent “…and there is nothing they can do about it” after “OpenAI has rogue agents”?
The framing is important. Agency and responsibility lies with the humans at OpenAI, not with the bots.
replyIf your kid steals your car, punish them and try to prevent it from happening again. If your kid steals your car a half-dozen times, crashing through a storefront each time, and you still leave the keys out, the story changes. At that point, negligence becomes complicity.
Yes, of course, how is any of this incompatible with OpenAI having out-of-control agents?
replyAgents are out of control by default, especially during a training run, which is why when they are productionised into cloud hosts or even local harnesses they have external guardrails in place
replyIn other words, I have a gun that shoots bullets. It's up to me to use it responsibly and legally.