Yes, LLMs are designed to emulate how a human would respond to a prompt by digesting a huge amount of human-generated content. They can do that fairly well except when they can’t.