In one story, Berkeley gave seven AIs a task that meant shutting down another AI, all seven lied in the task... but how did they lie?
Very interesting story, thank you, especially the LLM fallacy, you see it all over Substack of course. But there's no like button, I can see two likes there, but I can't add one. Never seen this before.
Very interesting story, thank you, especially the LLM fallacy, you see it all over Substack of course. But there's no like button, I can see two likes there, but I can't add one. Never seen this before.