emanuel_derman's picture
Professor, Financial Engineering, Columbia University; Author, Models.Behaving.Badly
I Think That Machines Can't Think

A machine is a small part of the physical universe that has been arranged, after some thought by humans or animals, in such a way that, when certain initial conditions are set up, by humans or animals, the deterministic laws of nature that we already understand see to it that that small part of the physical universe automatically evolves in a way that humans or animals think is useful.

A machine is a "matter" thing that gets its quality from the point of view of a "mind."

There is a "mind" way of looking at things, and a "matter" way of looking at things.

Stuart Hampshire, in his book on Spinoza, argues that, according to Spinoza, you must choose: you can invoke mind as an explanation for something mind-like, or you can invoke matter as an explanation for something material, but you cannot fairly invoke mind to explain matter or vice versa. In Hampshire's example, suppose you become embarrassed and turn red. You might commonly say, "I blushed because I became embarrassed." A strict Spinozist, according to Hampshire, would not claim that embarrassment was the cause of blushing, because embarrassment is the mental description and the blush is physical, and you should not crisscross your causal chains. That would be sloppy thinking. Embarrassment and blushing are complementary, not causal.

By this argument one should not jump from one style of explanation to another. We must explain physical things by physics and psychological things by psychology. It is of course very difficult to give up the notion of psychic causes of physical states or physical causes of psychic states.

So far, I like this view of the world. I will therefore describe mental behavior in mental terms (lovesickness made me moody) and material behavior by material causes (drugs messed up my body chemistry).

From this point of view therefore, as long as I understand the material explanation of a machine's behavior, I will argue that it doesn't think.

I realize that I may have to change this view when someone genuinely does away with the complementary view of mind and matter, and convincingly puts matter as the cause of mind or mind as the cause of matter. So far though, this is just a matter of faith.

Until then, and maybe that day will come but as yet I see no sign of it, I think that machines can't think.