It does in some cases go a bit mad, and OpenAI continues to be straightforward in regards to the 'hallucinations' that ChatGPT can have, and the problems inherent in these LLMs. Be wary of relying wholly on just about https://youtu.be/aGubCnA5Qyk