For practical purposes, abstractions are useful, so, no, it doesn't matter.

But the post we are responding to is directly posing the philosophical question of whether the activity of LLM agents constitutes thought.