2014 : WHAT SCIENTIFIC IDEA IS READY FOR RETIREMENT?

[ print ]

Psychologist & Computer Scientist; Engines for Education Inc.; Author, Teaching Minds: How Cognitive Science Can Save Our Schools
Artificial Intelligence

It was always a terrible name, but it was also a bad idea. Bad ideas come and go but this particular idea, that we would build machines that are just like people, has captivated popular culture for a long time. Nearly every year, a new movie with a new kind of robot that is just like a person appears in the movies or in fiction. But that robot will never appear in reality. It is not that Artificial Intelligence has failed, no one actually ever tried. (There I have said it.)

David Deutsch, a physicist at Oxford said: "No brain on Earth is yet close to knowing what brains do. The enterprise of achieving it artificially — the field of 'artificial intelligence' has made no progress whatever during the entire six decades of its existence." He adds that he thinks machines that think like people will happen some day.

Let me put that remark a different light. Will we eventually have machines that feel emotions like people? When that question is asked of someone in AI, they might respond about how we could get a computer to laugh or to cry or to be angry. But actually feeling?

Or let's talk about learning. A computer can learn can't it? That is Artificial Intelligence right there. No machine would be smart if it couldn't learn, but does the fact that Machine Learning has enabled the creation of a computer that can play Jeopardy or provide data about purchasing habits of consumers mean that AI is on its way?

The fact is that the name AI made outsiders to AI imagine goals for AI that AI never had. The founders of AI (with the exception of Marvin Minsky) were obsessed with chess playing, and problem solving (the Tower of Hanoi problem was a big one.) A machine that plays chess well does just that, it isn't thinking nor is it smart. It is certainly isn't acting like a human. The chess playing computer won't play worse one day because it drank too much the night before or had a fight with its wife.

Why does this matter? Because a field that started out with a goal different from what its goal was perceived to be is headed for trouble. The founders of AI, and those who work on AI still (me included), want to make computers do things they cannot now do in the hope that something will be learned from this effort or that something will have been created that is of use. A computer that can hold an intelligent conversation with you would be potentially useful. I am working on a program now that will hold an intelligent conversation about medical issues with a user. Is my program intelligent? No. The program has no self knowledge. It doesn't know what it is saying and it doesn't know what it knows. The fact that we have stuck ourselves with this silly idea of intelligent machines or AI causes people to misperceive the real issues.

I declare Artificial Intelligence dead. The field should be renamed " the attempt to get computers to do really cool stuff" but of course it won't be. You will never have a friendly household robot with whom you can have deep meaningful conversations. I happened to be a judge at this year's Turing Test (known as the Loebner Prize.) The stupid stuff that was supposed to be AI was just that, stupid. It took maybe 30 seconds to figure which was a human and which was a computer.

People do not just get fed knowledge. I have raised a couple of people myself. I fed them food, not knowledge. I answered their questions, but they were their own self-generated questions. I tried to help them get want they wanted, but it was (and is) their own deeply felt wants that I was dealing with. Humans are born with individual personalties and their own set of wants and needs and they express them early on. No computer starts out knowing nothing and gradually improves by interacting with people. We always kick that Idea around when we talk about AI, but no one ever does it because it really isn't possible. Nor should it be the goal of the field formerly known as AI. The goal should be figuring out what great stuff people do and seeing if machines can do bits and pieces of that. A chess playing computer is nice to have I suppose, but it won't tell you much about how people think nor will it suddenly get interested in learning a new game to play because it is bored with chess.

There really is no need to create artificial humans anyway. We have enough real ones already.