You should treat LLMs the same way you treat any other smart entity, human or otherwise: realize that they can be both immensely useful and fundamentally wrong

You should treat LLMs the same way you treat any other smart entity, human or otherwise: realize that they can be both immensely useful and fundamentally wrong at the same time. Intelligence is not equivalent to correctness.