I don't pretend to understand everything about this, but based on my limited experience with LLM based chatbots, might they not have some issues accurately following some of these prompts? Like, "If someone is rude..." Is the chatbot going to understand that?

Maybe this is actually an incredible art piece. Like, Jimmy's over here making poor people kill themselves over life changing levels of money that he's able to give away hundreds of times over. Maybe he wants us to start retching violently and rioting. I mean, it's possible, right?