My understanding is it learns from us. It's to be expected that it often speaks the way we speak, especially in places like this, since it's a free source of training data.

I'm not sure playing "spot the llm" is going to be a good use of time in many cases.