That article is actually about just that. The guy made ChatGPT rethink its answers and eventually change its arguments. It's interesting cos the first thing it says may not be true and in fact, its almost like it was learning.
You are viewing a single comment's thread from: