Unless you prompt them really well, large language models (LLMs) are not that clever. You can’t punch in a few keys and expect War and Peace to come out the other side. It doesn’t work like that.