Yes, completely agree. LLMs instantiate Searle's Chinese Room argument; symbols alone have no meaning. And yes, I buy the Kripkenstein argument. The meaning is the use, and intersubjectivity is key. My fear is that the humans who should know the use are having their grasp on that use degraded by production and consumption of AI slop text. Understanding requires application, and AI slop is letting the lazy off that hook.