Prompt Refusal
Data Skeptic
00:00
The Ethics of Open AI
William Bennett: In the beginning, I assumed Chad GPT would under the hood generate an output and pass it through a separate model. But Max was early on in saying, William, I don't think that's how they do it. And so they basically have these humans demonstrate to Chad GPT how it should be behaving instead of setting up like another filter at the other end,.
Transcript
Play full episode
