Comment by thfuran

It depends what exactly you mean by "LLM". But an ANN is effectively a function approximator. If you made one big enough to very closely approximate the entire quantum state of a person interacting with an environment, would you still declare that nothing it could do is "thinking"?