Click here to watch this full length movie / documentary
The History of Computer Chess: An AI Perspective
Computer History Museum 2 hr 5 min 57 sec – Jun 22, 2004
www.computerhistory.org
Playing chess by computer began in the early 1950s, nearly as soon as computers became available. As a human activity, chess is believed to require thinking, yet in 1997 a massively-parallel supercomputer, drawing on over four decades of continual advances in both hardware and software, defeated the best human player in the world.
Does playing chess require thinking? Or is human thinking perhaps a form of calculation, parts of which a computer can mimic? What is the tradeoff between knowledge and search? Was Claude Shannons 1950 prediction that studying computer chess might lead to applications in other areas fulfilled?
This panel, comprising seminal contributors to the solution of this challenge including two of AI’s leading pioneers will discuss these and other questions as well as the origin and development of computer chess and what it tells us about ourselves and the machines we build.”
This is a fascinating video. It’s incredible listening to these very smart individuals.
Bronstein said : “Computer calculate moves and human play chess.”
What else ?
susan,
your fellow hungarian, mathematician johnny von neumann wrote a little book called “the computer and the brain.” if i recall correctly, it was meant to be a lecture, but he was too sick (cancer) to give it. his notes were posthumously used to make the book.
william
I will never forget my very first encounter with computer chess. Back in the 70’s, already in USA (I was born and raised in Hungary, same country where Susan is from). One day with my wife we were just took a walk. I noticed a small new store with “hi-tech” (at the time) items. Since I was alway fascinated by technology, I just had to visit the store. We went in and I notice that the salesperson is showing something to another costumer. What’s that? A small chessboard? What does it do in a high-tech items store? I am looking…..what is that keypad doing on a chessboard. Hey look, something lights up to, some LED display (which I knew at the time). I go there, stick my nose into the conversation and ask “what is this”. A chess computer, says the salesperson. A chess WHAT, I am asking? A chess computer. A computer which plays chess.
For me, despite my reasonably good being up to date with high tech items, this was a contradiction in terms. Chess is for a high level intellectual activity for people. Computer is a machine which can do number crunching. How does the two get together? I express my doubts to the salesperson, who angrily looks at me, because the other customer took off. But he becomes nice and smooth, perhaps I am another potential customer. And explains that inside of this chess board there is a computer. If you “tell” the computer what your move is, the computer will display here (and points to the LED display) what it moves. I am absolutely fascinated. A computer which can think like a human. Otherwise, how could it play chess? With a definite “fear” I ask how much it costs, suspecting that something like this must be beyond my budget at the time. 300 dollars, comes the answer (or something in that range). I kind of knew that something like this must be more expensive than what I can afford. But my wife sees my absolute fascination whispers to me “buy it if you really like it”. I am tempted beyond belief. But as always, I must clear my head before jumping into “big financial decisions” and decide to test this miracle machine. So, I press the reset button, enter F2-F3 (intentionally) as the opening move.
A display flickers and shows E7-E5.
I punch in G2-G4. I want to see the “I lose” light to show up, since Qh4 is mate. To my total disappointment (and relief that I don’t have to spend an ill afforded 300 bucks) the machine displays G8-F6, missing the mate in one.
Oh well…..I mumble…..I will wait for the next one….I tell the salesperson (while I had no idea whether there will be a next one or not).
I leave the store with considerable disappointment, fascination and relief all at the same time.
——————-
And in case if somebody didn’t figure it out yet, I was encountering the very first chess computer ever came to the market, Chess Challenger 1. I resisted CC3, but I gave in when Boris came to the market, that was my very first chess computer. Good old days when Boris made me think that I am a reasonably good chess player. As technology went forward, I bought many other chess computers, each being better than the previous, making me feel less and less of a “good chess player”. Despite computers making me realize just how bad of a chess player I really am, up to date I am fascinated by the fact, that a “machine” can play chess. At all. Now that it can even beat the world champion, makes chess computers almost surrealistic to me.
Sorry for being so long-winded.
Gabor your post was great.
What I will explain probably scandalizes a lot of people.Sorry about that. I play chess as you, I have an elo of 2150, I am an engineer and although I am not devoted professionally to it I program in several languages.
Years ago I read a book of David Levy about creating a chess program… it seemed so easy, to calculate some functions, to make that the program plays against itself to improve them… with some few advice and allowing to play many times to my computer I got it. It is not a great program as Fritz, but it plays at a level of 1850.
An example for everybody: I decide that, to win, I should make as big as possible a function, called function victory, F. If the function victory is positive, I am winning and if it is negative my rival wins.
Now I decide that that function depends of other two (this voucher for any number, I explain just with two) f1 = material and f2 = development. And f1 is the values of my pieces (pawns 1, knight and bishop 3, Rook 5, Queen 9, king 1000) less the values of those of my rival. f1 has this way positive value if I have material advantage and negative if my rival has it . I call f2 the total sum of movements possible of my pieces less the movements possible of the pieces of the rival . If f2 is positive I have more development than him and if not I am underdeveloped.
Notice that I have defined the f1 and f2 like I have wanted.
Now then, don’t we know which the relationship is between f1 and f2, that is to say “as much as development compensate a pawn of less…” What can we do?. This: F= f1 +Number x f2 I Know that should be a number “Number” that can solve the relationship that we are looking for, And now? Solution: the computer plays against itself.
It is necessary to play a player computer with the function F=f1+2xf2 and to another with F=f1+3xf2. I suppose that second wins (they play 10 times), then already you know “Number” has to be bigger than three. Now I prove with F=f1+5xf2 against F=f1+3xf2 and it wins again. Now you Know the “Number” is between 3 and 5. Making tests ( a lot of them) you find the exact relationship and then you can put other functions f3 (the king’s security), f4 (control of the center), etc. I did my program with f1 until f10, but believe me when I tell you that they are functions so simple as f1 and f2 and that they were calculated in this way… for that reason when I play with somebody of 1850 or less(less than my cheap program), I cannot avoid to think that he/she knows very little of this game that I love.
As always:just think about that 🙂
hi..Im college student, thanks for sharing 🙂