It does sometimes go a little bit nuts, and OpenAI has become straightforward in regards to the 'hallucinations' that ChatGPT can have, and the issues inherent in these LLMs. Be cautious of relying absolutely on just https://youtu.be/aGubCnA5Qyk