concentrating on doing things that people don’t do well, and they should be concentrating on things that people do do well. The examples that you gave of what people do well were things like walking and talking and very simple things like this, whereas the examples of things that people don’t do well are things like pole vaulting and perhaps spending a great deal of money to go to Mars by the year 2000. Well, going to Mars is certainly not something people do well, but I don’t think that answers the question whether we ought to be doing it or not. The question is, is there any point in people going to Mars? Well, you go back and have a look at the origins of that program, and I think you get a pretty straight answer to that question. That program originally came out of the failure of the Bay of Pigs invasion. Kennedy needed a political victory to overcome the effect of that, and he decided to set up a big crusade, the Apollo program. We got this fantastic thing. We’re going to put a man on the moon. Was there any point to that? Well, it was a big point for the electronics industry, and they made a killing out of it. There was a big point for a lot of scientists who got huge research grants. But there is a question, did it make sense to put a man on the moon as distinct from a little receptor of some kind? You put a bunch of receptors on the moon. You don’t have to worry about an oxygen support system and precautions for survival and so on and so forth. It’s obviously going to be vastly cheaper and vastly easier from an engineering point of view when you get any less information. And whether is a man the best instrument to be on the moon? Well, scientists will have to answer that, but I’m quite certain the answer is no. But a human being is probably a very poor instrument to be on the moon. The choice of human beings on the moon was to have something dramatic. To have something that get people excited and to distract them from their own problems and subordinate them to power and give a shoot to the electronics industry and so on. That was promoted for it, and that doesn’t make any sense. But those seems to be the kind of questions that I want you to ask, not is it the kind of things people do well or badly. Sometimes the kinds of things people do badly are worth doing. Quantum physics is the kind of things that people do badly. That’s why there’s so few people who can do it well. It’s not like pole vaulting. Kind of a quirk. Some people can do it. Most people can’t, but that’s worth doing.
You put me in mind of a possible modification of Turing’s test, which if you follow cartoons, I’ll call it the Mark Alan’s [inaudible] test. I want to ask you what may not be quite a question. But, well, I guess the question or possibly nonquestion I want to ask is– there obviously is a good bit of research money going into these AI fold, and I mean, a lot of people pay good money to make chess-playing programs. It’s a funny thing. I was working with a friend of mine’s commercial one, and the thing was awfully cowardly. And I found that if I told it to make a move that I thought might work, and then let it keep me from getting killed, then it worked pretty well. But if you just let it work, it took forever to do anything.
Well, I must say on the– I’m just giving an opinion. On the question of chess-playing programs, is it an interesting question at the current moment of science to figure out how people play chess? The answer to that, that’s probably no. It’s not an interesting question at the current phase of our understanding of the mind, but let’s assume for the sake of our argument that it is an interesting question. So just for the sake of our argument. Then we take a look at the progress on program and we ask, “Does that program tell us anything about the way humans do it?” And the answer to that is flat no. They’re designed so as not to tell us anything about the way humans do it. The programs are designed to exploit the capacities of computers, which are different than the capacities of humans. So they have deep search and fast big memory and all that kind of stuff. So we, therefore, throw out the systems instantly, at least on scientific grounds, because they don’t do anything. So they’re studying what’s probably a dumb question, and they’re giving an answer, which is totally useless. So what’s the point? Well, as far as I can see, the only point is either advertising and getting research grants, or maybe taking the fun out of playing chess. If there’s some other point, I don’t know what it is