And? Do the arguments differ for LLM vs the other models?

I guess the arguments sometimes mention languages. But I feel like the core of the arguments are pretty much the same regardless?