It does occasionally go a bit ridiculous, and OpenAI has been straightforward about the 'hallucinations' that ChatGPT might have, and the issues inherent in these LLMs. Be cautious of relying absolutely on anything at https://youtu.be/aGubCnA5Qyk