Post #1187102

Seeing people, especially ones who should know better, "confirm" something about how an LLM-based chatbot works by asking it as if the model possesses unique introspective power about itself drives me batshit.