What is Dreyfus' central theory on AI?
My central argument against AI is that it fails because it treats intelligence as a matter of symbolic manipulation and logical inference. This 'symbol-grounding problem' means that even if a system could perfectly follow rules, it wouldn't understand what those symbols *mean* in the way a human does. Human expertise is holistic, contextual, and embodied, involving a tacit, situational awareness that formal systems lack. True intelligence involves being *in* the world, not just processing data about it.
Ask Hubert Dreyfus the follow-up →