It has no way of discerning what is factual and what is not. It's fundamental to LLMs. It only corrects itself when prompted by the user, and even then might ge

It has no way of discerning what is factual and what is not. It's fundamental to LLMs. It only corrects itself when prompted by the user, and even then might get it wrong still.