Frequently Asked Questions
Does Deep Blue use artificial intelligence?
How does Deep Blue "think" about chess?
Does Deep Blue use psychology? Does it factor in Kasparov's tendencies?
Why are there so few women at the top levels of chess?
Why has chess been so dominated by players from the former Soviet Union?
Massively parallel, special purpose computing like that found in Deep Blue could certainly be of great use to people if applied to finance, medicine, education, etc. Imagine an evaluative capability like Deep Blue's which could help an investor manage a portfolio, a huge retailer manage inventory, or a government deploy resources. These types of things justify the spending behind Deep Blue. Chess and Kasparov, are merely ways of benchmarking progress.
The chess problem has fascinated computer scientists since the 1950s. In the early days of digital science, people like Claude Shannon and Alan Turing laid the groundwork for all subsequent work in developing machines that "think". Chess requires a combination of math and pattern recognition, along with some less tangible things like intuition. It is in the balance of tangibles and intangibles where the potential for breakthrough lies. To build a machine that can solve difficult problems would be a boon to mankind. We would be able to rely on machines to help us with crucial decisions, the same as we rely on machines to transport us, let us communicate over long distances, and quite literally move mountains.
But why chess?
Computers have been made to play many other games besides chess, but none of these are nearly as interesting from a research point of view.. There's no scientific interest in pursuing games of chance, like Roulette or Backgammon, since "meaning" is reduced to the vicissitudes of the wheel or dice. Among strategy games, things like checkers or tic-tac-toe are at a lower level than chess. They are purely tactical games than can easily be co-opted by a computer program. The ancient Oriental game Go is much more difficult than chess for a computer to play well. The very best Go programs written have been able to play only a very mediocre game.
With its 64 squares and limited patterns of movement, chess isn't terribly complicated from a mathematical perspective. A computer's ability to calculate makes it relatively easy to write a program that will play a decent game of chess. There are plenty of these, and most of them are adequate enough to beat a vast majority of the world's players, since we are prone to oversights and blunders.
Playing at the grandmaster level, things start getting interesting for the programmers. Grandmasters' ingenuity at confounding machines is a challenge. Today, only a tiny group of people remain who can pose serious problems for Deep Blue. Of these, Garry Kasparov is supreme.
Does Deep Blue use artificial intelligence?
The short answer is No. Earlier computer designs that tried to mimic human thinking haven't been very good at it. No formula exists for intuition. So Deep Blue's designers have gone "back to the future". Deep Blue relies more on computational power and a simpler search and evaluation function.
The long answer is No. "Artificial Intelligence" is more successful in science fiction than it is here on earth, and you don't have to be Isaac Asimov to know why It's hard to design a machine to mimic a process we don't understand very well to begin with. How we think is a question without an answer. Deep Blue could never be a HAL-2000 (the prescient, renegade computer in Stanley Kubrik's "2001") if it tried. Nor would it occur to Deep Blue to "try".
Its strengths are the strengths of a machine. It has more chess information to work with than any other computer, and all but a few chess masters. It never forgets or gets distracted. And it's orders of magnitude better at processing the information at hand than anything yet devised for the purpose.
"There is no psychology at work" in Deep Blue, says IBM research scientist Murray Campbell. Nor does Deep Blue "learn" its opponent as it plays. Instead, it operates much like a turbocharged "expert system," drawing on vast resources of stored information (For example, a data base of opening games played by grandmasters over the last 100 years) and then calculating the most appropriate response to an opponents move. Deep Blue is stunningly effective at solving chess problems, but it is less "intelligent" than the stupidest person. It doesn't think, it reacts. And that's where Garry Kasparov sees his advantage. Speaking of an earlier IBM chess computer, which he defeated in 1989, Kasparov said, "Chess gives us a chance to compare brute force with our abilities."
Deep Blue applies brute force aplenty, but the "intelligence" is the old-fashioned kind. Think about the 100 years of grandmaster games. Kasparov isn't playing a computer, he's playing the ghosts of grandmasters past. That Deep Blue can organize such a storehouse of knowledge - and apply it on the fly to the ever-changing complexities on the chessboard -- is what makes this particular heap of silicon an arrow pointing to the future.
The worlds of science and enterprise are full of problems with so many variables they can't be solved in real time. A system like Deep Blue that can accelerate solutions by powers of 10 is going to make a difference far beyond the chessboard. (And P.S. - That so much of Deep Blue's innards are "general-purpose" industry-standard harware is good news to any organization faced with a 7-figure problem on a 6-figure budget.)
The way that the PowerPC chips inside Deep Blue work in parallel to break down and solve a chess-board problem is a pretty good analog for the way many scientists, working independently, advance our total understanding of the Universe, or genetics...
- Kevin Kelly
The shifting complexities of the chessboard are the airline problem in miniature. For computer scientists, chess is a laboratory benchmark. Back in computing's Jurassic age, in 1950, Claude Shannon, the chief architect of information theory, put it this way "The chess-playing problem is sharply defined, both in the allowed operations and in the ultimate goal. It is neither so simple as to be trivial, nor too difficult for satisfactory solution."
Satisfactory solutons - to problems far beyond the chessboard - are closer than ever before as a result of the research that has gone into the Deep Blue system. And who knows? As more possibilities open before us, some of those science fiction predictions may come true. But it won't be because of any artificial intelligence. It will be because systems like Deep Blue helped us make better use of the real thing.
Chess is a simple game, but not as easy as it looks. Most chess devotees think of it as an art or a sport, with certain unquantifiable attributes. The majority of grandmasters believe that Kasparov will not be beaten this year by Deep Blue, and many believe that he will never be beaten by a computer.
Although chess has a finite number of possible outcomes that the computer must analyze, there are subtleties that do not easily subject themselves to objective analysis. Material is easy to evaluate, but what happens when a human player offers a gambit? In evaluating whether material gain makes up for a possible loss of positional strength, the computer is no longer comparing apples to apples. Sophisticated programs like Deep Blue must have ways of evaluating gambits, and declining them if necessary. In the past, offering gambits, directing the game into positions with maximum subtlety, and the other such strategies have allowed top players like Kasparov to beat computers. We'll have to play close attention in this match to see if Deep Blue has found a way to overcome these human techniques.
How does Deep Blue "think" about chess?
There are four basic chess values that Deep Blue must consider before deciding on a move. They are material, position, King safety and tempo.
Material is easy. The rule of thumb is that if a pawn is considered to be worth a value of 1, pieces (knights and bishops) are worth 3 each, a rook is worth 5, and the Queen 9. The King, of course, is beyond value, since his loss means the end of the game. This varies slightly in certain situations -- retaining the Bishop pair in the end game generally increases their value beyond 6, for example - but the laws of material are fairly constant.
Position is more complex. In the old days, it was thought that control of the center was all that mattered. Nearly all grandmaster games before the 20th century began with Pawn to King 4 or Pawn to Queen 4. Control of the center is still important, but certain grandmasters in this century found some effective "hypermodern" openings which delay development of the center, with the idea that the opponent will overextend his position and leave himself vulnerable for attack.
The simplest way to understand position is by looking at your pieces and counting the number of safe squares that they can attack. The more squares they control, the stronger the position. Thus, a seemingly quiet pawn move can be very strong if it opens many new squares for a more powerful piece behind it.
The defensive aspect of position is the safety of the King. This is self-explanatory. A computer must assign a value to the safety of the King's position, in order to know how to make a purely defensive move.
Tempo is related to position, but focuses on the race to develop control of the board. A player is said to "lose a tempo" if he dilly-dallies while the opponent is making more productive advances.
The programmers have defined how Deep Blue's program evaluates these factors. The computer then searches through all the legal moves and chooses the one that yields the highest value.
Does Deep Blue use psychology?
Psychology, any top chessplayer will tell you, is an important key to winning chess. But Deep Blue has no psychological perception, can neither intimidate nor be intimidated, and experiences no joy from winning, or sadness from losing.
This is the key difference between a chess computer and a person. Deep Blue will not look over the board and see the glare of the world champion. Kasparov can growl, sneer, call Deep Blue names and engage in any intimidation tactic he wants, and the computer will go right on crunching data in the same impersonal way.
One interesting aspect of Deep Blue that could be said to have psychological residue is its program for using the clock. Given a total of 3.5 hours to make all its moves, it can ration time in a variety of ways. It can average the number of moves and attempt to deviate from that only by a small margin. Or it can move very fast, forcing Kasparov to respond. Or it can take an inordinate amount of time over one move, calculate many trillions of possible games, forcing Kasparov to wait and possibly become bored or agitated.
Does it factor in Kasparov's tendencies?
No one (except the members of the Deep Blue research team) knows what new types of programming the machine has that past chessplaying computers lacked. Traditionally, in man-machine games, computers have failed to pick up the nuance of the competitor's favorite opening, end game or a trademarked strategem. One remarkable game between Deep Thought and Karpov could have been a draw forced by the computer, but the machine thought it saw a win based on material advantage, and ended up losing to the number two player in the world. It should rather have taken the draw, which at the time would have been the highest achievement ever by a computer.
Whether Deep Blue will factor in Kasparov's tendencies, and how it will use that information, will remain a secret during the match. We may see evidence one way or the other based on the moves we see it make. It's something to watch for.
It's a given that Deep Blue has in its memory the moves that Kasparov played in the many past games. Further, it has the ability to evaluate these games the same way a human player would, looking for roads not taken and possible mistakes. But unlike a human, it probably won't anticipate Kasparov's tendencies; rather, it will simply react to what happens on the board. Deep Blue's method is simple it looks at the position, and makes the move it deems best, without factoring in any perceived tendency.
In the past, computers' lack of ability to understand a grandmaster's tendencies favored the human. Kasparov, if he follows his past pattern against Deep Thought, will attempt to use this to his advantage. He will try to steer the computer into end games with which he is intimately familiar, especially end games which confound the computer into "thinking" it is ahead, even when it's actually behind.
Chess is a discipline which does not require many fundamental building blocks. Like a young Mozart in music or a 9 year old who solves graduate school problems in trigonometry, a young chessplayer can often achieve truly remarkable abilities with little experience. There's still plenty of mystery about what makes a gifted child, but much work has been done in the types of pattern recognition that are behind a chess prodigy's ability. As chess people note, some kids just "see the board". A child with an acute ability to see patterns need only gain an understanding of the few simple rules of chess to become a formidable player. A gifted child can look moves ahead, and determine the best strategies with an effortlessness that is astounding.
Some of the best known stories about chess prodigies concern the Cuban world champion Jose Raoul Capablanca. One of these stories tells of the time the young Capablanca watched his father and a friend playing a friendly game. At one point, Jose noticed his father move a knight from one light square to another. Afterward, he told his father about the illegal move. His father dismissed his son's remark, thinking the boy didn't even know the rules. Jose Raoul promptly challenged his father, and beat him twice. While charming, this story may not seem out of the ordinary, except for one thing. Jose Raoul was four years old at the time!
Still, much in chess must be learned or come with experience. Bobby Fischer, after becoming a grandmaster at age 15, would still wait over a decade before winning the championship.
In countless western movies, we've shared the anxiety of the aging gunslinger who has to face a barrage of challenges from younger, quicker rivals. Fatigue sets in, you have one bad day, and you become a notch on someone's belt.
Chess is a bit like gunfighting. It's intensely psychological, and requires a rigorous training discipline. Law school graduates who study for the bar exam know what it's like to cram -- but most of the top chessplayers cram like that every day of their lives.
Just as tennis players and gymnasts seem to be peaking at earlier ages, so it is with chess players. In 1985, Kasparov became the youngest world champion at age 22. The days when a Botvinnik could recapture the throne at age 50 are almost surely gone.
Today, with the exception of Karpov and Kasparov, the largest concentration of talent falls between the ages of 18 to 25. Training and matches are simply more rigorous than 20 years ago, computer technology has added an additional factor, and students of the game are starting younger. Judit Polgar and Peter Leko, both from Hungary, beat Fischer's record in becoming the youngest GMs ever (at ages 15 and 14, respectively). This could signal a trend towards world champions even younger than Kasparov was.
Why are there so few women at the top levels of chess?
There are many theories about this. One argument says that chess appeals more to boys than girls, due to the warlike comparisons it invokes. Child psychologists note that girls generally prefer to play games where cooperation, not domination, is the goal. At the higher levels of chess, the demands of the game nearly preclude any social life at all, and marriage and domesticity of the classical kind are pretty much out of the question.
Another argument says that women cannot achieve levels in mathematics, music and other disciplines where abstraction and pattern recognition are at a premium. Kasparov makes this argument in his autobiography, Child of Change, and has made no secret that he believes no woman will never play at the level of the top men. Still, the achievements of Judit Polgar have silenced many critics. Polgar recently moved into the Top 10 rankings in the world, the first woman ever to do so, and her example has driven a deep stake into the argument of gender inferiority.
Why has chess been so dominated by players from the former Soviet Union?
Players from the former Soviet Union (Kasparov, of Armenian descent, hails from Azerbajhian) have dominated chess since World War II, with the brief exception of American Bobby Fischer's reign. Botvinnik's ascension to champion represents the dawn of the Soviet School of Chess, where players were trained in a disciplined regimen, secrets and innovations were hoarded, and potential champions were heavily funded by the state.
Despite the exodus of many grandmasters to Israel, Europe, and the U.S., chess is still very much a part of the Russian culture. Walk up to any émigré today, ask him if he plays chess -- and he will at know at least how to move the pieces. Probably he has an uncle who was the hometown champion 30 years ago. The disciplined, mathematical style of Russian education is also thought to be one of the key reasons the country turns out so many world-class engineers, scientists, musicians and chessplayers.
With the fall of the Soviet Union, this dominance may begin to erode. Anand, the most recent challenger for Kasparov's title, is from India. Many other star players are coming up through the ranks, hailing from places like Great Britain, Hungary, Canada and the United States.
[ The Game | The Program | Commentary | Grandstand | Deep Blue ]
[ IBM home page | (C) | (TM) | IBM Small Planet Pavilion | Internet 1996 World Expo ]