Here’s my plan. We regard a computer as one long binary number—RAM, disk drives, registers, all of it. The order is arbitrary but doesn’t matter as long as it doesn’t change during the process. Set it to zero. Then begin incrementing it by one. At each iteration, write the number down somewhere.
This will generate all possible machine states, and therefore all programs possible to run on that computer.
Of course most of these numbers will not represent programs. It will therefore be necessary to have graduate assistants check each stored number to see whether it does anything. Perhaps they could run it through a disassembler to get readable code, or maybe try each to see what happened. Those which turned out to be programs they would store in an archive, labeled by the program’s function.
Now, it’s true that there would be a considerable number of these long numbers, specifically 2 ^ (number of bistable devices). On a machine with two terabytes of storage, the number would approximate 2 ^ (8 x 2 x 10 ^ 12). Thus analyzing them all would take a certain amount of effort, but that’s what graduate assistants are for. Give them lots of strong coffee and potato chips and they’ll go at it like badgers.
You would end up with a list of all programs possible, ordered by function. Programming would become a simple matter of referencing a look-up table. Calculations from the area of a circle to abstruse matters of computational fluid dynamics would be there at your fingertips. No more onerous fiddling with FORTRAN or C++. This will be a boon for people with a phobia for curly brackets. I am sure that the American Psychological Association will be interested.
Having thought of all this, I was basking in my own brilliance when it occurred to me that I had only looked into the shallow end of the pool. The list of machine states, I realized, would contain not only all programs, but also the state of the machine after running each of these programs. Yes! We would have not only all programs, but all answers! It would be necessary to have the graduate assistants associate each program with its results, but, well, brewing coffee is cheaper than hiring programmers. Even Indian ones. When they had finished, you would simply look up your program, versions of which would include the outcome for all possible data, and find the answer next to it.
An additional advantage would be that you would no longer need the computer, just the look-up table. This would be hard on IBM, but it’s a dog-eat-dog world, and none can advance unless some fall behind. Otherwise you couldn’t tell that they had advanced.
Another service that I propose to undertake for the betterment of society is to sue Intel Corporation for fraud, licentious commercialism, mopery with intent to gawk, and deceptive business practices. The company has been abusing the public for too long. It is time to act.
Intel is now selling a chip which, it claims with a straight face, in front of God and everybody, contains two billion transistors. Sure it does. And I’m Ivar the Boneless.
I mean, think about it. A CPU is somewhere between the size of a thumbnail and a postage stamp. Do you think two billion of anything can fit on something that size? Especially transistors. I remember seeing my first transistor in 1957. It looked like a three-legged aspirin tablet—a little pill with wires coming out of it—or maybe a miniature milking stool. Just possibly you could have fit three of them on a chip, if you used a hammer. Two billion? Maybe on a tennis court.
That people believe this sort of thing is a measure of the decay of the American mind. You couldn’t have fooled Davy Crockett with such stuff. Today the schools take malleable kids and teach them so much obvious rubbish, or rubbish that ought to be obvious, that they begin to believe anything at all. Such as that they need a new iPod every week or their lives will be blighted. They believe in molecules, though they have never seen one, in electrons, though nobody has ever seen one. They believe in astrology, evolution, Creation, and the FDIC. They believe in Mars, for God’s sake. (I put one of those NASA photos in Photoshop and blew it up, Way in the distance I found a billboard, blurry but legible, saying “Pedro’s Cat Tacos.” Arizona, I tell you.)
Intel’s trickery is obvious. How do they know that the chip has two billion transistors? Who has counted them? Assume that you could count a thousand an hour, or about twenty-five thousand a day—if you did it in shifts. Call it two hundred thousand a week. The chip hasn’t been in existence long enough for us to know how many transistors it has.
“Ah, but Fred,” you say, “the bugger works. How do you explain that?
I didn’t say it didn’t work, just that transistors have nothing to do with it. I figure it’s powered by spirits from a parallel universe—little tiny spirits. They may be green. They work for nothing and don’t require visas. Think graduate assistants, but more compact.
Actually, I think all of the space program is a scam. I mean, suppose I came to you and asked you to give me three billion inflating green ones to build a Jupiter probe.
“What’s Jupiter?” you might ask, wisely.
“Oh, it’s a planet.”
“Yeah? Who says?”
“Uh, well, Dr. Fulano de Tal. He’s very important and resounding, and works for NASA.”
“Uh-huh. You’re all in it together.” It’s true. Everybody who works on a Jupiter probe is connected to the space program. Pretty suspicious.
If you gave me the money, I’d get a web server somewhere in the Central African Republic and put some pictures of the Sahara on it with a red filter, and tell you we had landed on the Red Spot of Jupiter and what a splendid triumph it was for the United States, sing the anthem. Then I’d make a Swiss bank very happy.