OpenAI deserves some blame here. Their product UI does a very poor job conveying that it will just make things up when asked for specific information. And in th

OpenAI deserves some blame here. Their product UI does a very poor job conveying that it will just make things up when asked for specific information. And in the conversation there's no particular hint that the confidently stated "facts" are not in fact true. A large number of people get tripped up by this when first using ChatGPT. I did myself, it took me a day or two to realize the papers or web pages it was claiming existed didn't. It's pathological.