Putting aside how idiotic it is to describe a LLM as a "Python loop", what's amazing with those people, who I'm sure would self-describe as materialists, is their assumption that a computer program couldn't possibly be intelligent, which I think reveals how deeply entrenched some inchoate form of dualism remains. I'm sorry to break it to you, but the human brain isn't so special. At the end of the day, it's just a wet computer interacting with a body, both of which are the result of a very dumb process that went on for billions of years, the last few hundred millions of which it spent selecting your ancestors that were the best at fucking. The implicit assumption behind this kind of argument is that, unless intelligence isn't implemented exactly as in the human brain, it's not really intelligence. But in general there are several ways to implement something and there is no reason to assume that human-like intelligence is any different.