Every story and comment from the Hacker News API, in a layout borrowed from 2001.
A comment on What Is ChatGPT Doing and Why Does It Work?. Parent comment
The problem with many early LLMs is that they could be persuaded into believing something if you stated that it’s true.
Maybe the default prompt in GPT-4o includes something to the effect of “You are a helpful, critical-thinking chatbot”…
There are no comments on this comment yet.
The problem with many early LLMs is that they could be persuaded into believing something if you stated that it’s true.
Maybe the default prompt in GPT-4o includes something to the effect of “You are a helpful, critical-thinking chatbot”…