[pessimizer]

This is a lot of words to say that a human can prompt an LLM to tell it what they want.

edit: it reminds me of all that I have to wade through after I've asked an LLM a straightforward question and the answer should have been "yes, you're right."