It does from time to time go a little bit nuts, and OpenAI has become honest with regard to the 'hallucinations' that ChatGPT may have, and the problems inherent in these LLMs. Be wary of relying absolutely on something https://youtu.be/aGubCnA5Qyk