Walter Quattrociocchi (Sci-Am, May 2026) certainy thinks that AI and human judgment are really different. For him a LLM cannot “distinguish a reliable claim from an unreliable one except by analogy to prior linguistic patterns. In short, it cannot do what judgment is fundamentally for.” AI, in his judgment, is defective in this: that it lacks a real-world understanding of the subject in question. It can, for sure, search for, and report on just about all that has been written on the subject; but AI’s knowledge is limited to what his sources tell him. Pretty hard to refute this claim. Yet AI is not alone in this. Think about a man: he has no real knowledge of what it is like to be a woman. Does this exclude him from the set of people who can pass judgment on women? Think of a priest: he has no direct knowledge of marriage. Should therefore his counselling on marital matters not be given any regard?Indeed, should any human being ever trust another human being, given that no two human beings ever think the same way?
No comments:
Post a Comment