For those of you who are so intellectually disabled that you can't comprehend even the topic of the discussion, let me clarify. The discussion is about trustworthiness of LLMs, benchmarked against compilers. "Do you need to read code produced by an LLM or can you trust it like you trust the compiler?" Some say LLMs are close enough to compilers. They're wrong. Some say LLMs aren't anywhere near compilers, because the former are non-deterministic and the latter are not. They're wrong too, because non-determinism isn't the defining difference. I hope this was clear enough to pierce the wall of your dumbery.