Deep Blue | Down the Rabbit Hole

The name

deep

blue

is revered in the field of computer engineering, sometimes invoked when discussing the rise of machine learning, and although the way it relates to these fields is often vaguely inaccurate or imprecise, it is also used as a challenge to anthropocentrism and to express fears that machines and computers advance beyond human capabilities, especially in creative work, and sometimes relates to anxieties associated with human obsolescence in similar unspecified ways. Misunderstandings about the nature of

deep

blue

, its construction and its history are common, so what was deep? Blue In 1949, American electrical engineer Claude E Shannon submitted an article to the scientific journal Philosophical Magazine entitled Programming a Computer to Play Chess.

The idea of ​​creating a chess-playing computer had been floated for many years, but in this article Shannon listed the challenges associated with chess automata. The fascination with chess automata had begun more than a century earlier, when a series of talented scammers created machines. that would be secretly controlled from another room or from within the building by a skilled chess player who beat the players. unsuspecting challengers one of these machines supposedly played against napoleon bonaparte and won a real chess machine was built in 1914 that could force victory in a specific endgame scenario by calculating its moves through mechanical devices shannon claimed it would be simple to create a computer program that chose a random legal move to play each quarter, but his and others' real interest was in creating a chess program that could play at a competitive level.

In the opening of his article, Shannon justified his interest in this topic by suggesting that quote, although it may not be important. Practical importance The issue is of theoretical interest and it is expected that a satisfactory solution to this problem will act as a wedge and attack other problems of a similar nature and of greater importance. To quote, this clinical approach, however, does not comprehensively describe scholars' fascination with chess. For hundreds of years, the ability to play strategy games had been considered evidence of the exceptional human quality of intelligence and, in In particular, the ability to play chess was directly correlated with mental acuity, as such it proved to be a tantalizing challenge for early computer and electrical scientists.

Many engineers believe that developing a high-quality chess program would help make significant progress toward realizing true artificial intelligence, a term that would be popularly adopted later that decade; even Shannon's contemporary Alan Turing was developing his own rudimentary chess program that played a simplified version of the Game Shannon went on to offer suitable methods and data that the program should have to analyze a position and suggest how such a program could do so. However, this analysis depended on solving another, much more extreme problem: the absurd exponential complexity of chess using a simple method. Shannon's equation established a lower limit for the number of possible chess games determining that after only five moves there are approximately 70 billion possible chess games that can be played and for the 40th move the lower limit is approximately 10 to the power of 120. power exceeding the number of atoms in the observable universe by 10 duodecillion times this incomprehensible variation arises from the relatively simple rule set of chess chess is played on an 8x8 grid with each player taking control of 16 pieces there are six types half Of these the pawns are the most complicated pieces that can move one or two squares in their first move than one square thereafter only forward capturing pieces diagonally in front of them and promoting to a stronger piece if they reach the other side of the board towers can move vertically and horizontally without limit unless another piece blocks their This way, if they fall on an enemy piece, then it is captured.

Bishops move similarly to rooks, except that diagonally the most powerful piece is the queen, which combines the movement capabilities of rooks and bishops. There are also horses that move in what is often called l. -forms two squares, either horizontally or vertically, then one square perpendicularly, the knight is the only piece that can jump over others but has a limited range of influence, the final and most important piece is the king, it moves similarly to the queen, except that there is only one square in at a time making it somewhat vulnerable, this is important since the object of the game is to trap the king, if the king is under attack by an enemy piece, it is said to be in check and If there is no way to remove the check, then the position is checkmate and the game ends, however, a player cannot put his own king in check and if he has no legal moves, then the position is a tie or a point. dead, meaning a losing player can try to tie the game.

Stalemate can also occur if checkmate is no longer possible, as if everyone except the kings has abandoned the board in order to select appropriate moves. A program like the one Shannon proposed would need to understand both the rules and the basic strategic and tactical concepts of chess and use them. evaluate the relative strength of each side, but in developing this method there was a final choice about how to allocate the time given to the evaluation. Shannon gave two possibilities, type a and type b, according to his definition, a type of program that would analyze all movements. up to a predetermined limit based on your computing power, evaluate each position and choose a continuation based on which sequence gives you the best odds, assuming the opponent will make good moves;

However, with each person's movement, the game becomes exponentially more complicated and so on for each level. deep, the computer must become exponentially more powerful to compensate for the enormous computing power required for this method, meaning it would need significant amounts of time to evaluate its positions. Unfortunately, a machine operating according to the type of strategy would be both slow and weak. player, it would be slow since even if each position was evaluated in a microsecond, very optimistic, there are about 10 to the 9th power evaluations that must be done after 3 moves for each side, so it would take more than 16 minutes for one move or 10 hours for is half of a 40-move game, it would be weak in terms of playing skill because it only sees three moves deep and because we have not included any condition on inactive positions for the evaluation, the machine is operating in an extremely inefficient manner, calculates all variations for exactly three moves and then stops even though he or the opponent is in check a good human player examines just a few selected variations and carries them out to a reasonable stopping point a champion world can build at best combinations, say 15 or 20 moves, quote this The type would later be defined as strong but stupid.

Another suggestion from Shannon, program type B was the one he preferred. In this method, the program would only evaluate the most promising lines and explore them extremely deeply, while this would require a much more sophisticated analysis method. than the computer type, it would save calculation time by eliminating bad moves that its counterpart would fruitlessly explore. This was clearly modeled on the way humans view the chess board, discarding moves that show no promise while reserving their attention for those that can lead to an advantage, by definition this method would have to search for fewer positions to save time. of processing and to illustrate this contrast, type b programs would be labeled as weak but intelligent, however, Shannon emphasized the contrast between human and machine thinking writing, it can be seen that a machine has several advantages over humans 1 . high-speed operations in individual calculations 2 absence of errors the only errors are due to program deficiencies, while human players are continually guilty of very simple and obvious errors 3. absence of laziness, it is also easy for a player. human make instinctive movements without proper analysis of the position. 4.

Never-human freedom players are prone to making mistakes due to overconfidence in a position or defeatism and self-recrimination in losing positions. These must be balanced with the flexibility, imagination and inductive and learning capabilities of the human mind. He concluded his article with an ominous statement that invokes biblical language: it is not suggested that we should design strategy in our own image, but that it should adapt to the capabilities and weaknesses of the computer that is the computer. strong in speed and precision and weak in analytical skills and recognition, it should therefore make more use of brutal calculation than humans, but with possible variations increasing by a factor of 10 to a third.

Each move, a small selection, goes a long way toward improving blind trial and error end of quote this paper proved pivotal and would be cited frequently by other scholars researching computer chess, including its lower limit for the number of chess games would be called Shannon's number, however some of his assumptions would prove drastically wrong in a way that many at the time could hardly have predicted and the computers that emerged from this work would eventually consume the zeitgeist in an existential crisis. In 1982, 33 years after Shannon's article was published, a man named Feng Shiong Shu enrolled in the graduate program at Carnegie Mellon University in Pennsylvania, traveling from Taiwan.

There he earned his graduate degree and soon after was ready to receive his doctorate, he just needed to complete a thesis. His search for a topic was interrupted; However, a Carnegie Mellon professor named Hans Berliner Berliner had become a chess master before his time at Carnegie. Mellon earned the title of grandmaster in correspondence chess, a special version of the game in which each person's move is sent to the other by mail and games sometimes take more than a year to complete. Finally, in 1974 he received his doctorate with a thesis that examined the forms of chess. computers to analyze tactics and since then he had become one of the most respected and well-known researchers in chess computing and now, in 1985, he was approaching fengxiongshu to ask a favor.

For years, the Berliner had been working on a new so-called high-tech chess computer with a group of graduate students, but he wanted to see what improvements could be made to the custom chipset used to evaluate board position. Berliner made it clear that the only reward would be a points quote, but despite this, the problem was interesting enough. a shu and so accepted at the time that there was a discussion in the world of computer chess, claude e shannon had suggested that type b computing was the most promising, eliminating the weak options in favor of looking at only the most immediately attractive lines , but like computers. became exponentially stronger, it was clear that experiments with type a programs were achieving better results and so for years these strong but stupid type a programs had dominated research;

However, it seemed that this method was reaching its limit and Berliner was now proposing that chess computers needed to revert to the smarter but weaker type b method to progress. Although Shu had not specifically researched chess computers, he suspected that There were still ways to improve the efficiency of chess computers with modern technology, so he began creating hypothetical designs using high-tech evaluation hardware. He analyzed 64 individual pieces on the board, one for each square, and worked together to determine the relative strength of a position and what opportunities might be available to each player. However, Shu thought that this 64-tile design seemed inefficient and so he began working on his own theories.

He believed that with a proper custom design he could consolidate the entire evaluation of 64 chips into just one and a basic exploration of the concept showed that this was indeed possible and that a computer could be filled with many of these chips. To confidently increase his power, he emailed Berliner and told him that he was no longer interested in helping him with the 64-chip design, but instead wanted to explore his single-chip method, but Berliner was not convinced despite Shu's insistence and said that he was still only interested. in the function oforiginal evaluation of 64 chips, at this point Xu left the project with no interest in entertaining Berliner's fixation, instead flying back to Taiwan to investigate his idea for a printer driver for Chinese characters, suspecting it might win a a lot of money if he designed one at this early stage of printing technology, however, his single-chip evaluation design continued to attract him and the more he thought about it, the more he believed he could create a chess machine that would dominate the others that he would later write. the basic blueprint to build the mother of all chess machines, a machine that could defeat the world champion, in other words, I had the opportunity to pursue one of the oldest holy grails of computing and possibly make history on the downside. , the printer driver seemed quite promising too and was not an idea that could wait.

There was some chance that the printer driver idea could make me financially independent. There was virtually no chance that solving the computer chess problem would provide any financial reward commensurate with the time and efforts required. If I dedicated myself to the problem of computer chess, I would seek the glory of knowing that I solved the problem. I probably wouldn't be financially destitute, but there would be great personal sacrifices. The opportunity cost would be extremely high. It could be wasting a lot of time. The most productive years of my life on a project that had very little reward.

The decision was ultimately difficult after some serious soul searching. I decided to go for the glory. You don't get the chance to make history every day. End of appointment as Shu returned to Carnegie Mellon and began work on his chest ship recognizing that he would have to work quickly to retain the good favor of the university, especially considering that in a sense he was creating a project that competed against one of the most respected professors at the university. He estimated he had six months before professors would begin to reconsider his enrollment there and so he got to work after a few months struggling to adapt his designs to a standard chip size and battling with manufacturers who repeatedly made mistakes when producing their chips. .

He finally had an outfit he could try out after convincing his colleague Thomas Anand Theriman to wear his custom one. chess program on the tiles, the two discovered that the system was calculating 30,000 positions per second shu then juxtaposed this achievement by saying, quote, the fastest chess programs in 1986 were bell high tech and cray blitz bell and high tech were machines special-purpose chess program about the size of an office refrigerator, Gray Blitz was a chess program that ran on a multimillion-dollar Cray supercomputer. The best programs at the time were looking in the range of 100,000 chest positions per second.

Thomas was getting 30,000 chest positions per second with a basic office workstation plus a chest chip interface card the size of a paperback novel, albeit with a chess assessment of brain damage, end quote; However, the computer had yet to pass the ultimate test it needed to play against a human, so Shu approached a graduate student named Murray Campbell. Campbell was an accomplished chess player who had spent time helping the high-tech team. from Berlin and at the time had reasonable experience playing against computers that had a unique play style, while computers could vary widely in terms of skill, most of them shared some common traits, although they tended to be strong in terms of skill. of tactical skill due to their brute force analysis methods, they fought in slower and more strategic positions without an obvious objective, the main weakness of chess computers, however, was their materialism, both computers and humans tend to use a point. system based on the relative value of pieces to determine advantage pawns are worth one point bishops and knights are worth three points rooks are worth five and the queen nine the king is not normally given a value due to its primary importance these numbers Programs are used in chess to help the computer determine which positions and moves are most favorable, giving the king an arbitrarily high value to emphasize the need for its safety.

These difficult numbers are relatively simple for a computer to understand, but the positions of the pieces on the board are more nuanced and therefore computers tended to have difficulty understanding them, even when programmers attempted to give weight to certain patterns. , their computers still destroyed their own positions by chasing pawns on the other side of the board while their king was in checkmate, other methods, however, showed significantly weaker results, e.g. One of the first machine learning algorithms was applied to the task of learning chess. His creators fed him hundreds of thousands of Grandmaster-level matches in the hopes that he would find patterns that would lead to victory in his first match against a human opponent.

He played a relatively normal opening. then he immediately made a mistake with his queen, after some consideration the programmers realized what had typically happened when a grandmaster sacrifices his queen, the purpose is to get a winning attack, the machine learning algorithm, however , had not found the connection between attacks and sacrifices, so he assumed that the best way to win was to sacrifice his queen as quickly as possible, while the idea of ​​machine learning algorithms in chess was exciting, computers of Materialistic but powerful chess games of the time were the most promising, but they were often predictable, so a specialized strategy had been developed. to fight against these machines called anti-computer chess that focused on making moves that forced the game into slow and methodical positions in which the computers fought and it was this type of chess that Campbell had spent time learning, although the game between the new Shu and Campbell's computer era Did not record the result, the computer won.

Any doubts Shu had about his project evaporated, however, thirty thousand movements per second was clearly not the computer's limit and Shu wanted to create a speed tester to improve his skills, but despite his enthusiasm, there was a problem. . Fortunately, I didn't have the budget for it. His office colleague, Andreas Novatsik, had a solution. He proposed a particular type of board that could be obtained from the university's waste, and it turned out that Shu's mentor had a cabinet that included precisely the type of board they needed. which they stole, but while Shu and Thomas were still testing their hardware and altering their program, they received an email from a surprising person, Murray Campbell, after his loss against Shu's computer, had begun to wonder about ways in which the program could be further improved with a chess expert.

Now he agreed to help this small team and made a bold decision, according to Shu's nickname of crazy bird, they would enter their incomplete computer in the computer chess championship even though the tournament was only five weeks away since 1970 the association For Computing Machinery had maintained a tradition of computer chess tournaments used by programmers both to gauge the relative strength of their designs and to network with others in their field, and so even if they didn't have high expectations for their computer, It was often valuable to attend to enter, Shu needed to type in a name and since the computer was just a chip testing platform, he simply called it chip testing with five weeks left until the event.

Shu Thomas and Campbell worked feverishly to complete the chip test, even modifying it the night before they left. As the team expected, their program found numerous bugs and problems in their code specifically with a special defensive move called castling, if neither the king nor the player's rook have moved yet, then the king can move two spaces towards the rook and the rook. rook can move to the other. king's side effectively accommodating the king safely in the corner of the board while moving the rook to a more active position in one case, chip tess attempted to castle illegally and in another exchanged a rook for a pawn without any tangible benefit even though the incomplete .

The chip test managed to score two wins, two losses, and a draw against a host of powerful chess computers over the next two months. Shu's suspicions about the power of the chip test were proven correct after making significant changes. His search speed had increased tenfold from 30,000 positions. per second at 300,000, tripling that of modern supercomputers, these massive gains encouraged them to challenge their friendly rival at Carnegie Mellon Honsberliner and their high-tech computer, while the game itself was apparently never re-recorded, the result It was a chip test, the incomplete machine was made in less than a year with scrap parts defeated the high-tech, advanced chess computer made by computer chess specialists, in a gesture of humility, Shu stated that the high-tech It was probably still the most powerful computer, but it was clear to everyone present that the chip's testing had not yet reached its peak. full potential given this incredible success, the next moves for the chip testing team seemed obvious: continue to iterate on the design and begin seriously entering competitions, and Shu had little doubt in his mind that his team and his small computer could devastate to any competitor throughout the year.

The chip testing team continued to work on both the hardware and software of their computer, in particular Shu was interested in the idea of ​​singular extensions where a chess computer explores promising lines more deeply than they otherwise would or , in other words, a fusion of type a and type b. methods that have been in conflict for so long, the incredible power of these machines offered them the ability to look at lines whose results come later, in a sense that mimicked the way a human might look at the chess board, but it resolved that checker. much more exhaustively.

The test's first serious tournament would be the 1987 computer chess championship. At its first event, it was little more than prototype chips and scrap pieces running a basic program, and yet it was able to place above other specialized computers. This time, there would be a year of development behind. There was some concern in the project about Campbell's participation due to his association with the Hi-Tech project, but when Berliner revealed that Hi-Tech would not be participating in that year's tournament, Shu's concerns were allayed and despite Still lacking a budget, the small chip testing team turned its attention to the next event.

Shu continued to work on innovative new chips that could be used for board evaluations while Thomas and Campbell continued to improve the software. This time, when the computer chess championships began, the chip testing team was much more prepared despite some of their designs. were not ready to be deployed after simply upgrading his workstation's chip test further improved the number of positions it could evaluate per second from 300,000 to 400,000 even though the chip tests were still incomplete, the team was confident when they arrived in Dallas, Texas for the tournament while chip tests played his games shu developed a reputation for outbursts when his computer was working well and despair when it was malfunctioning at one point he jumped out of his seat as chip tests calculated his inevitable victory at the end of the chip tournament The test remained undefeated having bested some of the most powerful machines that specialists have been making for years, but this victory created a new rivalry.

Hons Berliner after seeing his success turn sour toward the token test team he had stolen from one of his chess experts that Shu later speculated about. From Berliner's perspective and talking about him, he wrote a quote: Imagine that you have a goal in your life, a higher calling, you have worked for decades to achieve the goal, you have made many personal sacrifices, you are now a widely recognized authority in the world. field you have chosen, you have achieved great success in the last few years and your goal seems to be in sight now imagine a group of students who come out of nowhere, are not from your field and have different ideas.

You have little power of persuasion over them since they are not your students. Their approach is the same. The opposite of what you are advocating, to make matters worse, appears to be progressingquickly, making it even more difficult to persuade them to follow your vision. What would you do? What can you do? End of quote now speeding past other chess computers the chip testing team had. a new goal: to turn a computer into the world's first grandmaster-rated chess machine. At the time, it was common practice for programmers to enter their chess computers in standard tournaments and earn ratings like a human player, although they were not eligible for cash prizes. international chess Organizations use a rating system called elo in which players gain or lose scores when playing against others based on their relative rankings, much like humans, computers also receive elo ratings and it was possible that a computer obtained titles from fidei, the international chess federation.

With such ambitious goals, the team decided their computer would need a new name inspired by Douglas Adams' satirical book, The Hitchhiker's Guide to the Galaxy. They named their new computer Deep Thought after the book's supercomputer designed to calculate the response to life in the universe and All this new computer would be a proper implementation of the work put into testing the chip and Shu believed it was what strong enough to compete with the grandmasters. He would soon have the opportunity to test this theory at the Fredkin Masters Open in early 1988. There both The Chip Test and Deep Thinking played side by side, while neither of them had received an Elo rating, but they received provisional ratings. of 2200 under the advice of Hans Berliner.

Berliner's high-tech computer scored 2376 and was also present while he mildly cursed. For the team, this under-assessment of deep thinking skills consistently worked in their favor. The great masters present would find themselves in complicated situations. event one of his opponents ross sprog riley commented after his game around 2200 this unforeseen advantage helped deep thought tie for second place in a field of about 30 competitors with the chiptest coming in fifth this first tournament against humans announced deep thoughts imminent dominance over the next For years, the team continued to improve their computer, but they were plagued with bugs that often manifested themselves during tournaments where untested changes implemented less than a day before the competition caused deep thinking to commit. errors, for example, a part of the deep thought code was not recognized in The Wanderer Moves a special type of capture in which a pawn is allowed to capture another pawn that attempts to pass it by moving two spaces on its first move.

The deep thinking team would be forced to revert these changes between tournament games despite this deep thinking continued to inspire fear as they played more and more tournaments and challenged high-ranking grandmasters, eventually becoming the first to defeat to one with regular time checks to help continue development. The team grew by two, so that there were now five members, the first three core members. They were still working on the computer Feng Shi Yongshuo designed the hardware and managed the project Thomas Anantharamen wrote the code and Murray Campbell acted as their chess expert, although Campbell had already earned his PhD and was told to abandon his work on deep thinking to concentrate.

In his work as a researcher, he was passionate enough to continue helping the team in his free time. Andreas Novotzik, the man who helped acquire a board for chip testing, had also returned to work on the deep thoughts testing feature and was joined by Mike Brown. who helped with deep opening book thought that even the Berliner would be forced to eventually admit that he had been surpassed by this newcomer in 1989, wrote in an article for Chess Report citing from May 1988 a profound reflection that has attracted creation of a team of students at Carnegie Mellon University. A deep thought that attracts a lot of attention is a dual processor chess machine in which each processor can make 450,000 positions per second and the two working together, these figures make it the fastest chess machine ever created, approximately five times faster than high technology. end of the quote, he continued to praise its advancements, specifically its unique extension technology, citing it as the main reason for its achievements, although this perhaps downplayed how radically different deep thinking was from other chess computers as time went on. and deep thinking continued to dominate computer chess. world it became obvious that other challengers were unlikely to appear against him in the near future and it was also evident that the human grandmasters would soon have to deal with deep thoughts as a serious opponent, but the team had long been interested in something even higher which shu had referred to as one of the oldest holy grails of computing shu and his team were determined to create a computer that could defeat the world chess champion as the capacity for deep thought increased, so did example articles appeared that fallaciously linked deep thoughts to the military, suggesting in apocalyptic terms that they and similar programs would eventually be used to coordinate tactics and strategies on the battlefield.

This anxiety was likely driven by contemporary fears that increasingly powerful computers will become an existential threat to humanity. Encouraged by films such as Terminator and War Games, both released shortly before Shu began his work in chess computing, the emergence of this field and deep thinking had become representative of this anxiety and therefore also had been the question of what it meant. It would mean when a computer would defeat the best human chess player in the world. While Shu was completing his work on his thesis, it was clear who precisely this human would be. Gary Kasparov Born in Azerbaijan under Soviet rule Kasparov attended Russian chess schools and quickly rose to He reached the rank of grandmaster in his early teens, traveled to attend international events during his youth, and earned the title of undisputed world champion at the age of 22.

Since 1948, the Soviet Union had dominated the highest levels of chess and all but one undisputed world champion. He had been Russian ever since, but this notoriety made Kasparov a target as Shu continued to improve on his deep thinking. Computer production companies were monitoring it, including the International Business Machines Corporation or IBM, known unofficially by its nickname Big Blue. IBM was one of the largest computer corporations in Russia. The world was among the first to commercialize computer technology and produce devices for all levels of use, and its mastery made its products some of the most recognizable in the world.

IBM executives and engineers knew that Shu would likely seek funding to continue his work after earning his PhD, and so eager to acquire his talents before others, they began courting him even before he completed his thesis work. However, Shuu only had one demand for them: they gave him the opportunity to create the ultimate chess machine just a month after defending his thesis. Shu left Carnegie Mellon and Hans Berliner behind and joined IBM, however, his team could no longer exist as it had even without transitioning to IBM, the other team members were receiving job offers from companies or focusing in his education and in the end only Murray Campbell followed Shu to work on deep thinking, sacrificing his status as a researcher at the university to do so the same month that Shu joined IBM.

The company arranged a confrontation between his and Kasparov's new computer on October 22, 1989 at the New York Academy. Art, it would be a small match with only two games to give each player a chance to play with black and white and only 90 minutes on the clock for each of them. Kasparov would start the match with black, while a deep thought had the game with white. One saw Kasparov. Taking a rather passive approach to the game in adherence to the principles of anti-computer chess by depriving any obvious target of deep thinking, his powerful computational abilities would become irrelevant and in doing so, deep thinking was bullied into the corner of the board forced to dance. his pieces back and forth as Kasparov maneuvered his to deliver a devastating final attack, the second game was noticeably different now that Kasparov was confident in his abilities compared to those of deep thinking, he was more willing to take aggressive positions and the series of inexact moves that The deep thinking done in the early stages of the game only emboldened him in controlling the center of the board.

Kasparov was free to make threat after threat as Deep Thought desperately tried to protect his pieces by neglecting their development while his knight and bishop languished towards the back of his position and his king was vulnerable in the center of the board while most of the Aides knew that Kasparov would probably defeat. He deeply thought that Shu had expected a stronger performance, but this failure only strengthened his determination. The positive press Kasparov received after these landslide victories encouraged others. great teachers to challenge deep thinking, but these men would not see such encouraging results. deep thought would attract one of them and defeat the other two;

However, moving to IBM created a surprising problem as the company was sharing deep thoughts with a wider audience and people began to complain that its name was too close to the phrase deep throat and had been introduced by several people by that time. accidentally as such, so you would have to choose a new name privately, your computer would continue to be known as Deep Thought for some time, but publicly everything you created would be called by the name by which the computer would be known from So Deep Blue After these events, the newly christened Deep Blue team retreated further into their offices to quietly progress on their new computer, but Kasparov was still training and improving to continue his Domination of the chess scene despite the threat that computers represented for their status.

Kasparov quickly adopted new technologies soon after becoming world champion. He had partnered with the new company Chess Base, which created for him an up-to-date database of chess games. Kasparov illustrated the prevalence of this type of preparation in an anecdote about being approached by an interviewer and photographer at an event who he wrote quote the photographer wanted some photos of me on a chess board to accompany the story the only problem he didn't have a board of chess with me all my preparation was done on my laptop a compact that really expanded the definition of a laptop it should.

It weighed about 12 pounds, yet it was much lighter and more efficient than traveling with my paper notebooks and a stack of encyclopedias of openings. The advantages would accrue when the Internet made it possible to download the latest games almost as soon as they were played. Instead of having to wait weeks or months for them to be published in a magazine, his willingness to embrace these new technologies helped him stay ahead of other older grandmasters who were less likely to adapt, Kasparov would even star in his own TV show. consumer chess in 1993 called Kasparov's Gambit published by Electronic Arts for which he provided voiceover and short video clips praising or admonishing the player for their moves.

I'm not so sure you're on the right track. Over the years, Kasparov gained a reputation for crushing chess. machines and companies sometimes paid him to do it at public exhibitions. Kasparov believed that participating in this way was not only good for his image but also helped advance the field of computing and so he was extremely receptive to the challenges of a While playing, he came to recognize how these programs were changing by writing a appointment. One of the interesting things about following the evolution of PC programs from one version to the next was that he could always detect the DNA of the program, so to speak, there would be new ones.

The code added new search algorithms and optimizations for a new generation of processors, but for lack of a better word, the damn things hada style. I joke about programmers treating their machines like children or at least pets, but there's no doubt that their creations follow them in some way and that these characteristics are passed on from one version to the next, like green eyes or hair. red. End of quote, even as Kasparov was improving, the newly christened deep blue team was working hard and for shu the computer had come out of a spotlight. to an obsession with sacrificing all of his time to dedicate almost all of his waking hours to building deep blue.

New team members joined and left frequently, but there were three who would see the project through to the end. Shu and Campbell remained the main members and joined. by an IBM programmer named Arthur Joseph Holm who would help rewrite Deep Blue's code, but he wouldn't be alone. IBM hired several grandmasters to occasionally train with Deep Blue so the team could adjust its program and discover its weaknesses. Point Shu was no longer the leader of the team and each member had full autonomy within his field to create his part of Deep Blue and progress was going well by the end of 1994.

IBM finally approached Kasparov to organize the first meeting between him and Deep Blue. Kasparov was immediately receptive due in part to his quest for funding after earning his undisputed champion title in 1985, began feuding with the international chess federation and eventually created the grandmasters association to help give grandmasters greater influence on the FIDE organization and later Kasparov even created his own competitive organization in 1993 named the Professional Chess Association or PCA after Intel dropped sponsorship of his new organization. Kasparov was desperate to find another one before it collapsed and IBM was quick to start talks with him, making negotiations much more amicable and he accepted the match with a prize pool of five hundred thousand dollars.

Kasparov wanted the split to be all. or nothing, but ibm was able to convince him to accept a four to one split, with ibm's involvement returning to research in the company the date of this match would be february 1996, which gave shu and the team little more than a year to get deep blue ready, but deep blue's most glaring weaknesses would not be revealed until the 1995 International Computer Chess Association Championship, attended by the world's strongest and most expensive computers. Think 2 that the computer the team used in place of Deep Blue when they finished creating their hardware and coding performed admirably, but they managed to draw games that should have been wins due to Deep Think Two's inability to recognize repeated positions, while the A stalemate can occur when a player has no legal moves, there are some other tie situations, the most common is the repetition rule, if the same position appears on the board three times, the game is declared a tie since neither of the players players is willing to advance in the game, while deep blue would do so completely. overshadowing some deep thinking in terms of computing power, the team recognized here how important it would be for their final build to include the ability to recognize repeated draws Beyond this, it was also still desperately struggling with omnipotent moves, in many cases this additional rule would cause the chips to fail basic tests during 1995, the team was unsure if they would have the complete deep blue in time for their match, but when the final batch of chips arrived two weeks before the event, everyone was working as needed , the looming deadline meant there was no time for the team to make changes to adapt to the hardware, but their machine still dwarfed the power of any other.

Almost 50 years ago, Claude e Shannon had stated that 1 million positions per second was an optimistic number before Deep Blue could perform any optimization. looking for 100 million and this was only a third of its theoretical maximum speed, while major improvements were still needed for deep blue to reach its true potential and although the team knew that they were nowhere near ready, there was no longer time to even Kasparov would not even wait for IBM and Deep Blue's opponent was anxious and confident. The match between Deep Blue and Kasparov was highly anticipated even by those who were not interested in chess. a significant advantage, but this would use more standard timing controls.

Each player would begin the game with two hours on the clock and an additional hour would be added after the 40th move. This prestigious method of play was the final domain of chess that computers had not yet invaded. and Kasparov accepted the public's title as a defender of humanity against the scourge of machines, despite his positive opinions towards computer technology, they would play six games over eight days on two days' rest starting on February 10, 1996. For a high-level event like this, six games were insignificant compared to the dozens of games played, for example to win the title.

Kasparov played 48 games against Anatoly Karpov over the course of 5 months. This short format worked against Kasparov with only six games left in the match, he would have little time to play. He learned to play against Deep Blue, although he believed his victory was inevitable. Kasparov had not been waiting idly for the match. He knew that he would have difficulties if he was not properly prepared, especially due to his unique chess style. As players improve, it is common to develop different playing styles unique to them, they prefer and perform better in particular positions over others and this chess personality can be extremely distinct.

As they discover their personality, they will tend to study opening lines and tactics to help them reach these favorite positions, further strengthening their personality in some players. They will favor aggressive strategies, finding tactics that involve frequent exchanges, while others prefer more passive strategies, slowly developing their pieces and accumulating an advantage over time while creating impenetrable defenses. One of the most extreme examples of this is Mikhail Tall, a Soviet grandmaster and undisputed world champion from 1960 who was known as the Magician of Riga in his everyday life. Alto was indifferent and playful, but when on the board he adopted a strategy of brutal ruthlessness, above all, Alto was known and feared for his ability to manipulate his opponent's psychology, making the game that much more complicated. the analytical skills of almost any player and make sacrifices that made his opponents cringe without knowing if he was bluffing, sometimes even he wouldn't know that one of his most famous quotes says that you must take your opponent to a deep, dark forest where two plus two are equal. five and the path leading is only wide enough for one, in contrast, Kasparov was a much more stable player whose calculating mind made him one of the most dangerous opponents in the world, in particular he was known for two things, the first It was his precision.

Kasparov preferred the sharp. positions where the winner is the one who can refrain from making mistakes the longest and he analyzed his opponents by guiding their games towards these positions and encouraging them to make mistakes or simply waiting for them to do so themselves while continuing to generate pressure. and the attack second, and perhaps most feared were his preparation skills. Grandmasters typically do not enter games against other grandmasters without first researching their opponents and discovering their strengths and weaknesses, in particular they will examine the favorite openings of their upcoming opponents and invent new strategies to fight them.

This is why Kasparov was famous and often He came so prepared that he already had an answer to everything his opponent tried. The anxiety he inspired in his opponents was sometimes so extreme that other teachers abandoned their favorite openings out of fear. that Kasparov had studied them and discovered something new to counter them when he teamed up with Chess Base and was able to recover his opponents' publicly recorded games even faster. His mastery of chess only grew stronger, but against Deep Blue Kasparov's greatest strengths would prove much less useful as mentioned. Previously, computers were strongest in sharp positions where their positional evaluations were most powerful and the processors handled in deep blue were far more powerful than anything a chess computer had used before, despite all its skill and experience, it was It is unlikely that Kasparov would surpass the deep blue in such positions. a situation kasparov was an aggressive player but computers were particularly known for their defensive capabilities in places where humans would have difficulty kasparov's preparation would also be of limited use the deep blue team had created a smaller computer on which they could test their software which they called profound blue junior and when kasparov asked ibm about the games he had played, they agreed, however, the entire computer had never played a game against a grandmaster, so there was no way to know for sure how deep would play blue, the day finally came and it continued.

February 10 Gary Kasparov sat across the table from Fengxiangshu, who would perform whatever moves Deep Blue chose. Next to him sat a moderately sized crowd of spectators and journalists eager to see how the world's most powerful chess computer would fare against the world chess champion. In the front row sat Kasparov's mother, who had Raised after his father's death when he was young, she was always present at his most important games. Behind him was a cordoned off room filled with snacks. These smoking rooms are standard for high-profile games like this. One of the players is allowed to get up and leave the table whenever he wishes and the extra space allows them a place of isolation to take a moment of respite while their opponent makes their move.

Kasparov was known for pacing back and forth while he stared at the board during The first game of Deep Blue had white pieces, meaning he would go first in some respects. This was an advantage for Kasparov. Black is considered to be at a disadvantage because he started a move behind and the previous world champion, Anatoly Karpov, was famous for stating that he would try to win with white and draw with black, as each of them would play each color three times. This would give Kasparov the opportunity to adopt a defensive strategy and observe Deep Blue's capabilities.

Finally, the players were ready to make their first moves. began with a king's pawn opening generally considered the safest option and Kasparov responded with his favorite, the Sicilian defense, while the deep blue team had prepared for him to do so, they were still somewhat surprised, usually the Sicilians lead to extremely sharp positions where the deep blue stood out, but this opening move indicated that Kasparov was not afraid of his potential or at least not so afraid of giving up his favorite opening. The deep blue team had scheduled him to play c3 in that position because of how uncommon it was at the highest level.

The grandmaster hired by IBM to help prepare Deep Blue had had little time to construct an opening book and so, in the case of well-known openings such as the Sicilian, such tactics helped them save development time, although the queen of Kasparov was gone. open on move 3, often a weakness in the middlegame, both sides were able to develop unsurprisingly until move 10, where Kasparov advanced his dark-square bishop by aggressively restricting Deep Blue's rook, at this point Deep Blue was off schedule. Now he would begin to calculate for himself and the real match would begin, Deep Blue began to advance his knights on the board into aggressive positions, taking away space from Kasparov by counting his experience.

Shu wrote a quote when I played the knight's 13th move to b5 for Deep Blue to attack his queen. Gary was surprised he stared at the board for a moment then he raised his eyebrows and gave me a furtive glance I couldn't help but smile momentarily when I saw his expression I don't know how he interpreted my fleeting smile but he spent a long time on his response end of the date finally Kasparov decided that folding his queen was the best option perhaps in part due to his fear of Deep Blues' attacking capabilities, but this gave Deep Blue opportunities.

On move 17, Deep Blue attacked Kasparov's knight pinning it to the lady and when he inevitably captured the horse. Kasparov was forced to punch a

hole

in his king's defenses or move his rook for one night. A bad exchange in an already uncomfortable position. He chose to weaken his king with a powerful position. Deep Blue began to exchange parts. pressed his advantage by gradually eroding the defensesKasparov's first exchanged his knight for Kasparov's bishop making the material completely equal but only three moves later he made a move that surprised everyone present pawn to d5 by doing so the pawn was no longer protected by the knight or queen, in essence , was sacrificing his known pawn for his unmatched materialism.

This move was unthinkable for most computers and Kasparov was later surprised by saying that it was a move he himself would have made, recognizing the new attacking opportunities. Kasparov retreated his king to the corner position of the board and defended it with his rook while simultaneously opening it for a counterattack on the open file. Despite this threat, Deep Blue saw no legitimate attack on his king, so he took the time to move his knight away from the king to grab a pawn, a move that seemed almost ridiculously reckless. Kasparov saw this greedy capture as an opportunity and so made one last desperate attack on Deep Blue's king, leaving him just one move away from achieving checkmate, but in doing so he left his king with even fewer defenders and Deep Blue was defeated.

They rushed to take advantage of this. Deep Blue had calculated that he had enough time to capture the side pawn before moving his knight again to the attack and so he set out to destroy what was left of Kasparov's defenses by controlling his king. and once again recognizing the futility of his position, Kasparov accepted the inevitable and finally resigned. Deep Blue had taken the first game. The spectators were dumbfounded at all levels of understanding. At the lowest level, viewers were shocked to learn that a computer had won a game against the best chess player. player in the world, while those with chess experience were surprised by the aggressiveness with which the deep blue played, something uncharacteristic of a computer, Kasparov and shu stood up to shake hands and tell their experience.

Kasparov wrote a quote. He was a little shocked, as were the viewers and commentators. Shu, who would have been aware of Deep Blues' winning evaluations on his screen, seemed a little confused, almost apologetic at the time of his big win, still a little dazed by how well he had played the machine. . He asked a thoughtful question immediately after giving up the two way. Grandmasters could begin what we call a post-mortem of an entire game. Where was I wrong? I asked, but Shu wasn't a great chess player and he was probably a little dazed because he couldn't remember enough of Deep Blue's analysis of the game. screen to answer so it was a bit of an awkward moment for both of us fung xiung shu I would also write about this moment quote this question it took me by surprise I could barely play chess and the greatest mind in chess was wondering where he could have improved his play I tried answer his question the best I could from what I remembered from Deep Blues analysis feeling like I wasn't watching the game at the same level as him and Deep Blue Gary gave up after a few minutes and quickly left the building without speaking .

For anyone outside of their entourage, news sources quickly discovered this victory and overnight the match became global news with late-night shows in the United States making room to give updates on the rest of the games. During the night, people from all different nations were paying attention, even those who were not interested in chess and invariably this competition was portrayed in a particularly dramatic way, Gary Kasparov stood as the sole defender of humanity against the invading force. of the machine, meanwhile the message boards were dominated by discussions about the game and the website set up by IBM for the match was almost completely inaccessible due to the massive influx of traffic, but tension had been building for Kasparov for years.

He had watched computers increase his playing strength since he acquired his world champion title and was now recognizing that his position was more fragile than he himself had acknowledged for many years. I would later write a quote about the entire history of chess, even the best players have been shielded from the kind of incredibly complex tactical game that computers handled almost trivially in 1993. You knew your human opponents had roughly the same limitations as you when it arrived. the moment. To deal with anything that came up on the board, that perceived balance went out the window when you were up against a strong computer, I played chess well, but also differently, psychological asymmetry and physical factors were an issue, but the new sensation of always wondering if your opponent might be seeing something you could barely imagine was very unsettling, it created a terrible tension in complex positions, a feeling of fear that at any moment a gunshot could ring out in the dark in response, you double-checked. and triple your calculations instead of trusting your instincts like you used to. against a human opponent, after a lifetime on the chess board, you had no choice but to become a creature of habit, and all those habits were interrupted when you played against a machine.

I didn't like it, but I also wanted to show that I could overcome these disadvantages. and to prove that he was still the best chess player in the world, human or machine, end of quote, but there were still five more games to play and now no one was sure what to expect, not even the players, when the second day began Questions arose as to whether or not Kasparov would change his tactics to counter Deep Blue, but the preparation was not limited to Kasparov as they had in their other tournaments. The Deep Blue team sat down to alter his program, specifically trying to refine his opening book to avoid strange tactics.

Since Kasparov they knew that changing the program in the middle of the match was risky but with a poorly prepared computer they did not believe they had any other option. Again Kasparov sat at the table this time in front of Murray Campbell and the game began with Kasparov playing. White had the opportunity to dictate the opening of the game in its first move. They chose what is called the ready opening, where White advances the knight to the king's side. A rather flexible choice that leaves Kasparov's plans unclear, but his strategy was quickly revealed. The list was transposed to the queen's pawn opening and the position Kasparov fostered was slow with few attacking options his objective was clear Kasparov had abandoned his sharp attacking strategy to play anti-computer chess the next 24 moves continued calmly with no major threats and both sides made peace.

They exchanged to simplify the position as Kasparov had planned on move 24. They both had a rook, a bishop, a queen and five pawns. However, Kasparov had managed to exchange his pieces in such a way that he opened the deep blues' defenses as much as he had done. Blue had done to him in the previous game and now he had a clear positional advantage although the deep blue pawns were all lined up they occupied dark squares and Kasparov's bishop was on the light squares which meant he could move with impunity recognizing his bad position blue Deep began repeating moves trying to force a draw by moving his king back and forth, but Kasparov saw the potential for victory by maneuvering his queen around the open side of the board.

He began a joint attack with his bishop on Deep Blue's king, now with little defense at the end of the game on move 48. He was able to rest a pawn away from Deep Blue and on move 66 he captured another leaving him with two pawns. less and his king exposed after Kasparov forced a queen exchange, it was clear that there was no hope left and Murray Campbell resigned for the deep blue. blue on move 73 deep blue and kasparov were now tied on one point each kasparov's plan to create relatively boring strategic positions had been a resounding success the deep blue team raced that night to find out what had gone wrong and they quickly found the problem in chess strategy a featured piece is one that can no longer be attacked by enemy pawns and featured knights are incredibly powerful, however, Deep Blue was considering that bishops and featured knights were worth the same when featured bishops are much weaker if this had been discovered beforehand.

It is likely that Deep Blue would have been able to force a stalemate, this created growing anxiety in the team. They knew their computer wasn't ready for the competition, but with such an egregious bug present in the code, what other unforeseen problems could arise on the third day? of the event was a day of rest for both sides, but after the second game Kasparov did not worry, he wrote a quote now that he knew that Deep Blue was just a much faster version of computer programs, I understood it well, I relaxed a little, he was very strong, yes, but he was not stronger than me and he had clear deficiencies like with a human opponent if I could target his weaknesses and avoid his strengths he would win the final match date the next day they returned to the table for the third game and once again shu would sit in front of Kasparov, Deep Blue played the king's pawn opening and in a repeat of the game, Kasparov responded with the sicilian following the same opening as the first game, Kasparov found his queen in the center of the board, just like in the first game, but this time. was much more certain that both sides exchanged pawns and developed their pieces in somewhat expected ways, but on move 8 Kasparov made a bold pawn decision on d4 faced with the decision to recapture with the knight or the dark blue pawn he chose the pawn. , but this created a weakness with no pawns on the adjacent ranks.

This d4 pawn had no others to help defend it, a situation known as an isolated pawn. To maintain it, a stronger piece must be dedicated to its defense if it is ever threatened, but if those pieces are attacked on the turn, it is often a simple matter to capture the pawn and on move 11 Kasparov seemed to have an advantage with his better developed pieces, although Deep Blue's king was much safer after castling next to the king. on move 12. Kasparov finally launched a series of exchanges. ending with an exchange of queens, but these also gave Deep Blue another pawn to defend the isolated d4 pawn and modern computer analysis reveals that by allowing this, Kasparov had made a serious mistake, but still believed that Deep Blue had exploitable weaknesses.

Among them, the now-exposed C-3 palm, although this was largely according to Joel Benjamin's plans, Deep Blue noticed something that Benjamin had not noticed when he prepared the line, this series of moves missed a piece of Deep blue, so he deviated from the plan and narrowly avoided disaster. On move 20, both sides had two rooks, six pawns and one minor piece, Deep Blue with a bishop and Kasparov with a knight, although bishops are often the favorites in endgame scenarios. Kasparov was optimistic about his chances; despite this confidence, Deep Blue began to do what computers know best: defend. obvious weaknesses he began to draw his king out from behind his defensive pawn wall, a common tactic as pieces become scarce and the king's power is required, although Kasparov was able to drive him away with his rooks, he quickly realized that he could not There was no way to adequately attack Deep Blue's vulnerabilities and was forced to admit that its defenses were impenetrable after 39 moves.

Kasparov and Shu agreed to tie even though the score was tied at one and a half points each and even though he had lost what he felt was a chance to win. Kasparov was comfortable with the results, White is normally considered to have an advantage, and he had White in two of the three remaining games. Now, with this in mind and his understanding of the Deep Blues' playing style, he felt that he could finish the series in dominant fashion, but The Deep Blue team would fight this assumption until the end. The fourth game began the next day, as Kasparov opened with Reddy once again.

Deep Blue transposed him into the Slav defense, which gave him stronger central control but slowed down his development, almost guaranteeing a defensive game on his side that eventually built into what is called the silent variation, which refers to the games methods normally found in him, by reaching the middlegame Kasparov had a notable advantage and by move 19 had managed to double his rooks into an open file, an extremely powerful attacking pattern, but Deep Blue was not only facing problems in the board after making move 21. Shu got up and went to use the bathroom, but when he returned, the referee did not let him enter the stage until Kasparov made his move due to a quirk of the rules.

When Shu was finally allowed to take his seat again, he noticed that Deep Blue had gone into sleep mode when Kasparov got up to take a drink.short break. xu tried to wake up the computer and enter Kasparov's move, but when he did, he made a mistake and the computer crashed. Kasparov again saw several team members hovering over the computer trying to get it working again and when he tried to complain to the referee about this strange breach of decorum, his mother yelled at him from the audience to shut up despite this. problem, the team was able to revive Deep Blue and the game continued at this time of the game.

Modern computer analysis shows that Deep Blue had once again managed to bring the game to an even position. Both sides were able to find clever ways to grab pawns. each other, but although deep blue was able to take advantage, there was no obvious way to turn it into a victory. Kasparov had spent almost all of his time and had only a few minutes left to play to advance 40, where each side would be receiving an extra hour on their clocks in the frantic position, the board began to shrink even more as they found ways to capture the each's pawns, but with no clear path to victory, both sides finally agreed to a draw after 50 moves, Kasparov and Deep Blue now each.

He had two points, but more importantly, Kasparov had failed to secure a victory in one of his two remaining games with white, but these games were already starting to run out. kasparov shu recalled speaking with frederick friedel, founder of chess base and friend of kasparov, who told him that after the fourth game kasparov had entered his hotel room stripped to his underwear, lying on the bed and staring at the ceiling for a while. prolonged period of time during the rest day the deep blue team was beginning to get desperate trying to find an opening line that would help give them an advantage in their last game with white while kasparov tried to calm his nerves too early, however, the two Competitors returned to the table for their fifth game again, Deep Blue moved his king's pawn, but to the surprise of both deep blue's team and the assist team, Kasparov responded with e5 by moving his own king's pawn, he had given up his opening line. favorite to use some other strategy although what he had planned only his entourage knew his strategy became clear after a few more moves he had prepared the defense petrov often opening the king's pawn results in a sharp game, but by playing the petrov the deep blue team knew that kasparov was probably trying to draw the game, by kasparov's own admission petrov is quite slow and easy to play and many find it boring.

Pieces are often exchanged while pawn structures remain somewhat stable. Kasparov's plan with Petrov was obvious. The anxieties of playing against such a powerful and unknown computer had exhausted him and he was interested in keeping the game simple so he could get to the final game. he fresher and ready to get a win with white, so he would simply keep the position even and exchange pieces to reduce the complexity of the game. Sticking to obvious and familiar positions, this strategy caused significant consternation on the deep blue team throughout the week they had. been fine-tuning and preparing deep blue to play against the sicilian defense and kasparov's change of tactics meant that deep blue was much less prepared, which is perhaps why they were so surprised when on move 23 kasparov offered a draw , this was quite early in the game to make such an offer and after some deliberation, the Deep Blue team chose to reject it knowing that they could make the same offer later.

They believed that Deep Blue could still have had chances, but quickly regretted this decision they saw as Deep Blue. He began making a series of inaccurate moves immediately afterward, first allowing his pieces to become trapped hindering any potential for an attack in the near future and then beginning to move his pieces back and forth unable to find a proper way to advance his game. position. Kasparov tightened his grip on the game, the team watched in agony as they recognized the problems that could have easily been fixed if only they had a little more time to test each move, the deep blue inching closer and closer to a brutal defeat , their pieces were united in the center of the board, totally ineffective.

Finally, Kasparov was able to promote his pawn to queen, forcing an exchange that left him with three pawns and a bishop against Deep Blue's five pawns on move 47, seeing no way out for Deep Blue. , the team resigned because of this Kasparov had won his last game with black setting the score at 3 points to 2. Now all he had to do was tie his last game with white, but beyond this, the deep blue team recognized that Kasparov had an even more powerful advantage. He had learned how. to manipulate and dismantle Deep Blues' chess, although it was no longer possible for Deep Blue to win the match, he could at least draw if he could win the last game, but Deep Blue's team's hopes were not high this time, they had prepared the game.

Slav defense and for the first few moves everything seemed to be working as planned after Kasparov played the ready opening again but then Kasparov sprung his trap instead of playing the normal move order for the opening he transposed his moves changing the order and taking the deep blue. from his opening book correctly suspecting that the deep blue team had not prepared for this sequence, this meant that all of the deep blue team's planning had been wasted on move 10, deep blue was beginning to deviate from the opening plan and make serious decisions. Strategic errors He first failed to move his bishop to the correct square and then began to advance his knight ineffectively along the king's side on move 12.

Modern analysis shows a 1.2 point advantage for Kasparov, normally an advantage of one. point is enough to reliably achieve victory at the highest levels of chess. After chasing away Deep Blue's night, Kasparov continued to take control of the board, first driving Deep Blue's bishop back and blocking any attacking opportunities in the middle. game and then continue to invade Deep Blue's queenside, strangling any opportunity for multiple attacks. At times, Kasparov saw opportunities to make crushing attacks, but he knew this was unnecessary as long as he could avoid a sharp position. Deep Blue couldn't steal the game after a few peace exchanges.

The last remaining Deep Blue pieces were trapped in the corner of the board and after a pawn sacrifice by Kasparov were completely lodged on it, practically useless on move 43, the deep blue team admitted defeat and resigned. Kasparov was elated and was immediately more willing to speak to the current media now that his victory had been confirmed. was already assured, there were doubts about a rematch as early as the following year and he made it clear that he would accept that opportunity, a view that was partly encouraged by IBM's offer to finance his faltering chess organization when pride in the estimated elo rating of Deep Blue of 2700, very close to Kasparov, claimed that its capabilities were wildly inconsistent, saying that 2700 perhaps, but 3100 in some positions and 2300 in others despite the deep loss, IBM's investment in the chess computer of Shu had already achieved his goals as a result of his newfound notoriety.

Not only had IBM become a well-known company in the United States and abroad, but its shares had increased by $3.3 billion in the week following the match. Shortly after, IBM signed a contract with Kasparov to play a rematch next year. next, but Shu's team felt very different. they knew that deep blue had not been prepared as they would have liked and now kasparov had revealed many of the weaknesses that they might otherwise have overlooked even while playing deep blue they were considering methods to improve it beyond the practical considerations that this week had given them something perhaps equally important determination throughout the match kasparov had made surprising decisions giving up sacrifices for peace that he otherwise would have made for a simple reason while confident that he could defeat a human with his typical way of playing, He knew that a computer was a very different opponent, one that was unfazed by brutal tactics like the ones Kasparov enjoyed, no matter how strong his calculations may have been, recognized that using them against Deep Blue invited defeat, and had switched to slower strategies. just in time to gain victory at that moment.

He didn't know how sketchy Deep Blue was and after the rematch was scheduled for the following year, he had assumed that it wouldn't be possible to substantially improve his skills against them, but over the past 12 years, Feng Shi yongshu's work had been underestimated time and time again. again and in each case this overconfidence had ruined his opponents before the 1996 match with Kasparov shu had theorized numerous potential improvements to the software and after the match he also had ideas about new chips, but not to increase the search speed, as Claude and Shannon had stated in their 1949 paper, faster search speeds provide exponentially diminishing returns and it was clear to the team that it was not the search speed that limited deep blue, but rather they resolved that they needed improve your evaluation capabilities.

To recognize his weaknesses, IBM hired chess grandmasters to visit him and play against him so they could offer their insights and opinions, and all of this was used to fine-tune his evaluations, helping him recognize opportunities and how to achieve more powerful positions, gradually its strategic capabilities improved. The team recognized that if they could improve this aspect of Deep Blue's game enough, then they could eliminate any obvious Kasparov counterplay, but IBM had their own designs after losing to Kasparov, his stock had increased enormously. What if Deep Blue won as the year went on? Kasparov began to get nervous before IBM had been relatively forthcoming and every time he had questions or wanted to argue with their executives, they quickly answered him, now they were almost completely silent after he signed the contract for the rematch, no.

There were more serious discussions. Regarding the financing of Kasparov's chess organization, it was clear that while IBM had cooperated before they were now completely adversaries, this problem was exacerbated by the fact that, as in 1996, the rematch would last only six games, which would give him little chance of discovering the deep blues. Playing style and weaknesses are a much more important aspect for Kasparov than Deep Blue. As the team continued its effort to improve its testing function, they created a miniature version of Deep Blue to test software changes that they called Pico, which ran at roughly the same speed. of other chess computers, but when they tested them they received a surprise.

Shu wrote a quote: One day, Murray Campbell mentioned that he had played some games against the best commercial chess programs and asked me to guess the result since Pico really had no advantage in search speed and the new evaluation function was not yet available. it was tight i answered around 50 50. then murray dropped the bomb pico had won all the games i wanted to see the evidence murray showed me fragments of the memory games the reason for pico's dominance became obvious in our quest to improve chess knowledge deep blues, we had created something far beyond what had been done in commercial chess programs, in every fragment of the game I saw, the relative deficiency of chess knowledge of commercial programs became very clear and I could tell with precision which hardware evaluation features were at play in each game's final quote, however, deep blue was still plagued with bugs, for example, the team was once again struggling with the passing rule that manifested itself in a strange way every time the opportunity for a step on the board appeared.

Blue would somehow assume that a ghost queen had appeared around the corner drastically affecting his evaluations, plus it seemed like the problem was in the hardware, not the software, which was much harder to fix other times due to bugs in the code. . a random move a significant amount of their dwindling time was spent discovering these errors and devising solutions that would not significantly harm their speed or evaluation quality to address some other problems, they devised a method to handle the numerous errors that appeared instead of lock the system. the entire deep blue system would auto-terminate the program so you wouldn't waste so much time witha hard reset, this meant that even with deep blue in an unstable state, he would hopefully still be able to play full games and this time there weren't many more opportunities to look for problems, although his time was still quite limited, while his evaluations had improved drastically , they had still found time to improve Deep Blue's search speed and after optimizing their program and altering some hardware, they were able to double Deep Blue's maximum. search speed from 100 million positions per second to 200 million, but this time the deep blue team was not only relying on their computer's ability to play chess.

Miguel Yesgas, one of the great masters who helped tune Deep Blue, would later write a quote Of course, we also incorporated some tricks for Gary For certain moves there was a delay or some moves were performed immediately In some positions We bet Gary would make the best move and if he does, let's respond immediately this has a psychological impact as the machine becomes unpredictable, which was our main objective end of the appointment meanwhile, ibm was planning their event, which they intended to make much more publicized that before, during the preparation they gathered all the media attention they could by creating a campaign to publicize it and posters with the iconic image of Kasparov's face. on a chess board while journalists began to clamor for seats in the auditorium knowing that the public interest would be even greater than the previous year with so little cooperation and communication from IBM Kasparov was beginning to feel nervous and so he He contacted the company to ask them to show deep blues games so he could prepare, but the company had been clever when they drafted the contract and match rules in response to his requests.

IBM cited a precisely worded rule stating that all public deep blues games would be available to it. However, study games between him and the grandmasters they hired were private and therefore they were under no obligation to provide them to Kasparov. By ensuring that Deep Blue did not play public games before the match, they could completely negate one of Kasparov's greatest strengths: his preparation. However, the Deep Blue team had access to all of Kasparov's games, allowing them to prepare specifically for him and thus create an advantage for Deep Blue before the first game was even played. However, some of his demands were greater, such as when he asked IBM for a version. of deep blue against whom he could practice this request was also denied the match would begin on may 3, 1997 at the equitable building in manhattan just a 30 minute drive from deep blue's laboratory when the day finally came that the team from deep blue was unsure again Time was against them and they felt that if they had six more months they could confidently challenge and defeat Kasparov, but as their computer was, they still felt they had a reasonable chance that the prize pool would have increased for the winner between Kasparov and IBM to take. 700,000 and the loser would take 400,000, but for both Shu and Kasparov the money was irrelevant.

The only thing that mattered to both sides was victory. Finally, May 3 arrived, but there were already countless problems. Kasparov immediately noticed a number of shortcomings in his accommodation. First, there was insufficient climate control for the playing area, especially given the intense, hot stage lights pointed at him, and this would only get worse over time. He had to fix the air conditioning so that he could sit comfortably in the place. His suit table, his personal smoking room, was also several minutes' walk from the board and when he complained, they cordoned off a new area adjacent to the playing space, but this was far from comfortable or private. .

Kasparov's mother told him, that reminds me. In the 1984 world championship match against Karpov, you had to fight against Karpov and also against the Soviet bureaucracy. Here we are, 13 years later, and you have to fight a supercomputer and also a capitalist system using psychological warfare, although of less direct importance and less extensive, deep blue. The team received equally strange treatment. First, an accommodation mishap deprived them of their comfortable space to plan their opening preparation and during the week the observation room to which they were assigned was cramped and uninviting, unable to accommodate more than six people at a time.

Those without seats were allowed into the auditorium to take pictures at the beginning of each game, going chaotically in groups, but there were so many photographers that they were crushing each other trying to get good angles and at one point a reporter hit. a security guard IBM's planning team had so severely underestimated the number of journalists that a new room had to be rented to accommodate them all, as the major television stations were prepared to provide hourly updates on the games while they were in progress. finally developed after months of campaigning for ibm and days of interviews and photographs gary kasparov and feng shiong shu went on stage, shook hands and sat opposite each other at the table for their first game that they would play outside the auditorium in front of a small crowd of reporters approved while inside the auditorium, a small panel of hosts would analyze the game for the audience.

The team making the moves for Deep Blue would rotate as the games progressed. This time, Kasparov would start the match with the white pieces as he hoped to start with the black pieces. He had done so in 1996 in order to evaluate Deep Blues' abilities, but as a target he felt forced to achieve victory with little knowledge of Deep Blues' new capabilities. He mirrored his play in the 1996 match and again started with the flexible opening ready, but his next moves were surprising instead of trying to occupy space in the center, he first chose to immediately fianchet his two bishops.

A pledge is a powerful but safe attack pattern in which the pawn in front of the knight moves one square forward and the bishop moves to the now empty square. This accomplishes two things: it creates a safe nest for the bishop and at the same time places it along the longest diagonal of the board, allowing the player to safely apply pressure on the central squares, although giving up space. The reason for this was obvious if Kasparov couldn't have a defensive game with black then he would force one with white this was combined with another idea one of the principles of anti-computer chess was to make waiting moves allowing the computer to make mistakes this he wrote a quote instead of the normal expansion in the center pushing my king's pawn two squares I timidly advanced it a single square avoiding contact with Black's forces it was intentionally passive almost a waiting move a throwback to the old David Levy trick to see if the computer could be tricked into weakening its own position when it ran out of concrete targets and lo and behold, it might end.

Deep Blue's tenth move was to advance his h-pawn, already creating unnecessary vulnerabilities for his king as he saw the opportunity. Deep Blue began to invade Kasparov's space and Kasparov continually refused to take advantage of it. and offer an objective on move 13 Kasparov had not moved a single one of his pawns two squares creating a tight position on his side of the board while deep blue enjoyed the benefit of space, however on move 14 Kasparov had protected his position with So careful that there was still no clear way for Deep Blue to attack him, what's more, Deep Blue had castled the king's flank but had now demolished his own defenses leaving him vulnerable to attack while Kasparov's king was safely behind. of his pieces and the pawn game continued with Deep Blue constantly preventing.

Kasparov made a significant advance in his position until move 22, when Deep Blue made a surprising pawn move to g4, stunning the commentators, people began trying to analyze the move, speculating that it might have been a mistake or some strategy. complicated computing that could only be seen by the powerful computer that played him, while Deep Blue allowed his king's defenses to be completely eroded, Kasparov was weakened in turn, but not in a crippling way, so after Kasparov offered a pawn exchange and the d-file opened, Deep Blue tried to take advantage of this with his rook. While Kasparov closed other ranks on move 32, however, the deep blue king's vulnerability began to manifest itself, practically Kasparov's kingside pawns had advanced all over the board and were now putting pressure on the deep blue king and the However, the board joined the attack or threatened to do so imminently.

Deep Blue still had chances and the position was completely equal, but on move 34 they made the decision to start a queen exchange after this exchange Deep Blue completely lost control of the game. Move 40 began a rapid series of peace exchanges that left Kasparov with little material, but as they came to an end, Kasparov's final tactic was revealed. The exchange had left Deep Blue with no way to prevent Kasparov's pawns from advancing up the board and threatening to promote a fair position. A few moves later, the Deep Blue team quit. Kasparov had won the first game convincingly.

Kasparov stood in front of the crowd smiling and talkatively elated by his victory, but secretly there was great concern because he was reluctant to address Deep Blue's finishing move. He had taken an inordinate amount of time to move his rook to Kasparov's back row and by all accounts this move seemed worse than pointless. He had spent some time that night with his entourage studying the move trying to see if Deep Blue had detected any tactics that he wasn't clear after a short time that it was still a losing move? But they wondered if they could study why he made this move to understand his thought processes.

Now Feng Shun Shu was nervous. IBM had spent all its goodwill with Kasparov to gain every possible advantage and it was reasonable to assume that Kasparov could reject any offer of a rematch no matter how much money IBM offered and now he had lost the first game embarrassingly exhibiting some of the worst flaws in chess by computer, but neither he nor the deep blue team was willing to give up hope and they still had opportunities to make changes throughout the week at the beginning of the second game gary kasparov was still confident, but the myth around the match was already being built While Kasparov's anxiety over the rook's move had been resolved fairly quickly, Frederick Friedel wrote an article for Chess Space exaggerating the mental tension.

The move had caused Kasparov and others to talk about the game in reverent tones discussing how some of the moves of Deep Blues had actually been great with different justifications when game 2 began, many expected Kasparov to continue using anti-computer strategies as much as possible and at first this seemed to be the case when Deep Blue played the king's pawn opening . Kasparov responded in kind, showing that he was still unwilling to play his favorite Sicilian defense in favor of a closed central position after Deep Blues showed that in the first game many spectators, including the commentators, had given up.

There was no chance for Deep Blue to win the match and he had already started to berate him for allowing Kasparov to keep the closed position where computers usually have trouble, but Shu wasn't worried about Deep Blue continuing with Rui Lopez opening a series of well-explored moves. Popularly known as the Spanish Torture due in part to the way in which White's bishop restrains Black's knight by pinning it to its king. Play continued in a conventional manner for a while with Kasparov advancing his pieces much more aggressively than in the first game. Why had he changed his strategy? he wrote after seeing how aggressively the machine played when given the chance, we decided that using the same passive anti-computer strategy with black would be too dangerous with white.

I could control the pace of the game much better and wait for my opportunities with black. It would be safer to play a known opening even if it were in Deep Blue's book, especially if it was a closed opening where it would be difficult to find a plan. The downside to this strategy, like all games, was that it wasn't my style either. He was playing anti-computer chess he was also playing anti-Kasparov chess end of the quote entering the middlegame the position was hopelessly closed a common characteristic of Rui López and another reason why he is known as thespanish torture neither side has many obvious opportunities besides just vaguely improving their position while Kasparov's pieces were largely trapped on his side of the board, this was enjoyable to him because of the way it would frustrate a typical computer , but as the game progressed, Deep Blue's play became strange on move 16.

He played a4 a move described by Ieeeskas as theoretical In other words, it was a surprisingly humane move The commentators in the other room were starting to change His views where they had once been mocked by deep blue now recognized that the position he had created for himself was preferable and Kasparov squirmed in his seat and sighed heavily as he thought, as he later said, that the Spanish torture was underway and I was on the rack. Kasparov continued shuffling his pieces behind his pawns and at one point brought out his queen to recapture a knight on Deep Blue's side of the board and then removed it again just as quickly, but on move 26, Deep Blue again performed a strange pawn move on f4, another human decision that was surprising for a computer.

Now Kasparov was starting to get suspicious, but still. He would continue playing even though his concentration and will faltered, he resolutely continued with his anti-computer strategy by making weak moves that ideally would have made a computer play badly, but deep blue would only strengthen his strategic position over time, further restricting the movement of the pieces. by Kasparov. and even as Kasparov traded down, he felt his position increasingly unstable by move 36, his distrust intensifying when Deep Blue rejected a material advantage that would give Kasparov a position that could have allowed him a comeback, something that any other, more materialistic computer would have found irresistible.

Deep Blue then approached the now open column with his queen and her rook. Kasparov desperately tried to defend, but he was already exhausted playing for a host who was hostile towards him in a playing position he actively disliked after making a helpless check with his queen kasparov resigned, deep blue had not only defeated kasparov, but had crushed him even before the end of the match, however, kasparov's mind was elsewhere and he had become consumed by a thought that ibm and deep blue were cheating, this idea had grown as the hours had passed focusing. about some moves in particular and how it seemed impossible that Deep Blue could have played anything more than he expected, speaking later in an interview he said quoting game 1, the machine did not pay any attention to the safety of the king, we have to remember that because In game 2 and other games, the machine will be very concerned about the safety of the king.

Game 1 and Game 2 were played with different end date computers, but due to IBM's secrecy he did not know why Deep Blue had ignored their king. security so badly and didn't know it for some time shu later wrote in game 1 deep blues' move 22 g4 that resulted in his g-file pawn being exchanged with Kasparov's h-file pawn was controversial, some analysts considered it justified because deep blue won better peace activity as Kasparov did not get a clear winning position as a result it was difficult to say for sure that it was a bad move the move was the result of a mistake we had introduced just before the rematch an auto-tuning run for the evaluation The function pointed out to us that the wait for a king security term class should be increased when we increased the weight of the deep blue terms seemed to work better and we left the new weights on What we didn't realize was that in extreme cases the new weights reached the maximum allowed value and became saturated;

In other words, deep blue no longer distinguished between a very bad position and an even worse position. Discarding the g-pawn, therefore, meant nothing to deep blue as long as it eliminated Kasparov's h-pawn. end of the quote regarding his unusually strong skill and his close positions, joel benjamin talked about how far he went. i quote what kasparov didn't realize is that i spent months forcing deep blue to play closed positions so that if he ever got into one, to handle it better we spent a lot of time on that specific problem end of quote even before the next game, However, Kasparov's mental anguish would intensify during the day of rest as he walked down the street for lunch, one of his confidants pointed out a disturbing observation he had had. made about the second game kasparov could have forced a stalemate deep blue had several different moves for his king when kasparov made his desperate check with his queen and from the options he chose the wrong king to f1 kasparov then could have played the queen to e3 and If he had played correctly, he would have been able to force a perpetual check and repeat moves and, according to the repetition rule, turn his losing position into a draw.

He had been so distracted by the idea of ​​foul play that he had given up before discovering this opportunity and how they had done it. recently passed move 40 he had another hour on his clock plenty of time to find this line if he hadn't surrendered the deep blue team wasn't sure how to proceed and prepare shu wrote the seed of gary's defeat in game 2 was clearly his desire to play anti-computer chess which led him to reach horrible positions that were supposedly good against other computers but turned out to be dismal against deep blue but this was an analysis based on pure chess merit our amateur psychological analysis of Gary's mental state led to a very different situation Conclusion he had a very strong will most people would even say he could be stubborn at times he came to the match with a set plan and that was apparently to play anti-computer chess most likely he wouldn't change his plan unless it turned out to be a complete failure end of the appointment the next day Kasparov appeared on the board completely distraught and clearly unable to concentrate on some aspects the imperfection of the deep blue worsened his situation he could no longer say with any level of confidence if he saw something tactical he couldn't or if he was simply making a mistake, what's more, he couldn't help but wonder if there was actually a human choosing the moves, perhaps aided by deep blues evaluations.

For the third game, Kasparov once again had the pieces. White and However, blue had the black, as the opening progressed, it became very clear that even with White's advantage, Kasparov was playing scared by closing the position, retreating and rejecting the peace exchanges that would otherwise have occurred. performed, although the deep blue was not always playing the best moves, Kasparov. He would hesitate at every opportunity he discovered after 17 moves. Kasparov finally made the first capture of it to open the board and allow attacking opportunities on the b-file. As he did so, he began to offer other peace exchanges, but Deep Blue rejected them and took the game.

In the mostly closed and disordered position, Kasparov would ideally be able to take advantage of this, but for him, not knowing Deep Blue's true capabilities, every perceived mistake was a potential tactic, in particular Kasparov noted that instead of pushing his rook towards the back ranks of Deep Blue, he opted instead to exchange the queen, an unusual decision for him on move 35, through a careful peace exchange, Kasparov was able to give Deep Blue three vulnerable pawns as potential targets, but without no dark square bishop to take advantage of them and no way to maneuver their rooks into position. side the vulnerability of these objectives was mostly illusory after 48 moves it was agreed that the game would be a draw the score was still tied but kasparov had tied one of his games with white leaving him at a significant disadvantage in the match after the third game kasparov won He stood up to speak in front of the audience, but when the press asked him about the game he had just played, he quickly moved on to talk about the second game again and said that quote with an awkward smile.

Something really incredible happened, but the reporters had detected the suspicions of him and one directly. He asked him if he believed there had been human intervention. Kasparov refrained from directly answering the question, but made an indirect reference to Diego Maradona's famous Hand of God cheating incident in football. Kasparov later claimed that this referred to the fact that he had not done it. I didn't fully understand what was happening at the time, but Joel Benjamin immediately and reasonably interpreted this as an accusation of foul play. The two began a brief discussion on stage with Benjamin explaining certain computer chess concepts and professing Deep Blue's abilities while looking directly at each other. on kasparov, but kasparov completely avoided eye contact with benjamin, all this uncertainty, along with deep blue's massive hardware upgrades, had completely changed the nature of playing against him, kasparov understood that he had always been powerful, but now With the confusion associated with it, he stated that he no longer felt like a computer, it felt like playing against a black

hole

and it was now absorbing him by stripping him of his methods, his skills and most importantly, his confidence.

Kasparov withdrew once again to talk to his small group of confidants and analyze the game but with the match on the line ibm was not satisfied with giving them their privacy according to Miguel Yesgas ibm had replaced Kasparov's English speaking bodyguard with one who spoke Russian to be able to eavesdrop and listen to strategies so that the deep blue team could prepare against them, but this report is apparently uncorroborated and according to him it had no impact on the match, what's more, IBM was now concerned about the concerns of Kasparov, as well as in post-game interviews, Jeff Kisseloff, the reporter hired by IBM to write about Kasparov's side.

The party had its press pass revoked after IBM edited one of his stories to make it more favorable to the company, but accidentally posted the unedited version online. According to Kisselof, he was accused of hacking into their servers to publish the incorrect version of the article. The lack of communication between Kasparov and IBM, as well as between groups within the company, and Kasparov's demands seemed to change regularly, but even if Kasparov wanted the full records, there was a problem: self-terminations meant that in some cases it was done impossible to play the games, which meant there was no record of how deep he had reached his conclusions, this only increased Kasparov's suspicions and ibm seemed to be doing everything it could to promote them, offering to give him records and then rescinding the offer the next day or just handing him the records for One Play that night with his worries unresolved and his imagination creating more and more ideas about IBM's foul play, it was unclear whether or not Kasparov would show up for Game 4 despite his severe reservations and then that his manager spoke at length with him, Kasparov reluctantly.

He appeared the next day and game 4 began with deep blue as white pieces and Kasparov with black. deep blue opened again with e4 and this time kasparov played the kero khan defense, a passive but flexible response, the rest of his opening was unconventional, however, with a novel. and passive stance and again kept his pieces protected on his side of the board as deep blue expanded outwards on move 5. Modern analysis shows a 0.8 point advantage for deep blue, but again he would struggle to know what to do with this advantage and would soon squander his attacking opportunities when he closed the position on move 9.

Deep Blue quickly swapped all the bishops on the board to double Kasparov's pawns and weaken his kingside, so both sides decided to castle. queenside, but Deep Blue saw opportunities on move 9. on the king's side, so it exchanged its central pawn to isolate the double pawns, weakening them further over time. Deep Blue advanced his position, but then, strangely, once again began to push his pawns in front of his king, completely demolishing his defenses, just as he had done in the first game. To gain attacking opportunities, Kasparov was left baffled and the commentators were not sure what to make of this move, as it turned out that this move was caused by exactly the same king defense error as in the first game, while some had Thought Deep Blue's position was reasonably strong, this was refuted as the game continued and afterward. negotiating the position was now equal with two rooks and three pawns each on move 41 after threatening a repeated draw, deep blue was able to force Kasparov to exchange rooks, which quickly led to a clearly stalemate position and the twosides agreed to a draw.

Now there were two games left with an even score. Kasparov was frustrated at letting the game reach a stalemate, but decades later, in his book, thinking deeply, he wrote a quote: I missed a good attacking opportunity, but I still had a clear lead until the end. I finished the game only to discover that the machine had a series of incredible drawing maneuvers that I could never have predicted, even today, looking at the game after the 36th move, I can't believe that I couldn't win that position and, what is still Even more incredible, that the position might not even be objectively winning in the end quote above all this was a demonstration of a computer's ability to see much deeper than any human, even if it didn't always interpret the positions correctly due to coding errors, There would be a two-day break before returning for the The last two games with the score tied and each side playing one game with black and another with white, the stress of each game increased, but as Kasparov weakened from exhaustion, Deep Blue He only grew stronger as his team corrected the mistakes that had hurt his game.

In the first four games, Kasparov last had white in game 5, meaning a draw was unacceptable, but his passive strategy often made it difficult to force checkmate opportunities, unable to construct a better plan; However, he once again chose the ready opening and fianchetted his bishop. and when Kasparov advanced his pawn to attack Deep Blues' bishop, he chose to exchange it rather than retreat, weakening Kasparov's hypothetical defenses on the king's side, so that although the back rank was clear, castling was still dangerous, since before Kasparov would play much more passively. His bishop tried to recapture it and retreat immediately, but Deep Blue was advancing and on move 12 his pieces were much more active and developed than Kasparov's.

However, Kasparov was able to force Deep Blue's pieces back and exchange the last of the central pawns. leaving both players, the kings noticeably exposed, deep blue opted to castle on the queenside, which modern analysis reveals to be a huge mistake; However, Kasparov missed the opportunity or abandoned it for fear of having some complicated tactics. This fear was reinforced by his earlier pawn push to h5 a. confusing maneuver iescas talking to shu noted that perhaps one or two of the top 10 players would play it, but only after a drink or two, although this weakness gave Kasparov opportunities, he was too afraid of the deep blue to take advantage of this the way I instinctively felt.

Although he should do so on the 19th move, the game had become noticeably sharp and complicated, a position that favored Deep Blue, and now both sides had castled on the queenside, sharpening the game even further. Deep Blue opted for a rook exchange followed by a queen exchange offer which Kasparov accepted. Now both players had doubled pawns on the g-file without any obvious avenue of attack and as they exchanged their last pieces it became obvious that the game would almost certainly result in a stalemate on move 49, the two sides accepting a draw now that the score was tied there was only one game left, deep blue had the advantage of the white pieces and kasparov was too afraid of deep blue to take advantage of the opportunities it had given him mentally kasparov was defeated but he would still rush to the board one last time The day of By the sixth game, the auditorium exhibited an atmosphere like that of a sporting event, but Kasparov was clearly exhausted when he sat down opposite Joel Benjamin; in fact, I hadn't even decided if I was going to try to win or play for a draw, but playing for the win was an important decision given that I had the black pieces again, it started with Karo Khan's defense and Deep Blue wasted no time taking the center while Deep Blue advanced his horse to the king's side.

Kasparov again played defense for his opening. moves, but then, on move seven, after taking seven minutes for what the commentators saw as a rather obvious move, he decided that he would try to scare away the deep blues knight with h6. The spectators were bewildered. The line they were playing was well known and well explored. and they knew that if Deep Blue sacrificed his knight on e6 he would have a significant, if not overwhelming, advantage if he could play the next position well, but that could only happen if Black played h6 first, so why had Kasparov allowed this vulnerability which he later wrote about? quote says, of course, he was aware of the shots of the horse on e6.

I was also aware that it would be a killer move if Deep Blue played it against me in the sixth game. I just knew it wouldn't be like that. Machines are not speculative attackers, they need to do it. See the return on your investment in your search before investing material. I knew that Deep Blue would decide to withdraw his horse rather than perform the sacrifice, after which my position would be fine. End of quote. This was not a ridiculous assumption. Most computers were encrypted. not play the night sacrifice because its creators knew they couldn't play the position well enough, but then Deep Blue did the unthinkable for a computer: it performed the sacrifice without him knowing.

The Deep Blue team had assumed that Kasparov could play this opening and so on. They extended the Deep Blues opening book to include this particular file. Kasparov expressed surprise at him, but he continued playing, however, his defensive line was hopelessly blocked and with few opportunities to counterattack. First, Deep Blue moved his bishop deep into Kasparov's position to control his king and was forced to do so. move it, meaning he could no longer castle while Deep Blue continued to move his pieces to the king's side. Kasparov laboriously dragged his king to the queen side, but there his defensive pawn structure had already been severely damaged and Deep Blue's dark-square bishop had a comfortable position. line of sight directly across the ponds, so Kasparov removed his pieces to create a safe pocket for his king, but Deep Blue was able to force Kasparov to sacrifice his queen in exchange for a bishop and a tower.

He had had enough on move 19, after less than an hour at the table, Kasparov resigned, stood up, shrugged his shoulders at the reporters and walked to the auditorium for his final press appearance for the match. He had lost the shortest match of his professional career in humiliating fashion when he stood in front of the still supportive audience. Some even chanted his name due to the negative publicity surrounding Kasparov's accusations of cheating on Fung Xiong Shu and his team were ordered not to smile in case it would damage IBM's reputation despite support from the crowd. News stations chastised Kasparov for his poor sportsmanship and praised IBM for its win.

A few days later, Kasparov contacted IBM CEO Lou Gerstner to ask him about a third best-two-of-three match. but although Gerstner was superficially receptive, Kasparov acknowledged that there would be no third party. IBM had already achieved what In the week after the first game in 1996, the value of IBM stock increased by approximately $3.3 billion. This time the value of its shares increased by approximately 11.5 billion dollars. The company probably wanted to make sure they remained the ultimate victors. This was against the will of feng shingshu who was still interested in continuing with his work despite having technically achieved what he set out to do but now for him despite his triumph the victory was marred by the bad press that had surrounded the match and the ill will that had grown between his team and Kasparov after this event the careers of both Kasparov and Feng Shiong Shu would continue quietly Kasparov returned to chess to resume his dominance of the scene and maintain his world champion title by crushing his competition as he always had, but much of his interest in chess had waned, he had won every major tournament there was to win, and after becoming world champion and achieving the highest elo rating in history, he felt that his work was complete three years after his loss to Deep Blue, he would lose his world champion title to vladimir kromnic and after another five years of dominating chess tournaments he finally retired from professional play, however, he would have no career after his match with Kasparov served its designed purpose, it was separated into two halves, one of them was sent to the National Museum of American History, where it remains to this day, while IBM housed the other half on site for years along with others. server racks before donating it to the Computer History Museum.

Feng Shiong Xu left IBM shortly after the match, as Kasparov described it as the deep blue team. They had engineered their own obsolescence by defeating me, however before leaving, Shu was able to make a deal with IBM that allowed him to use his research on Deep Blue to make more chess machines when Xu approached Kasparov for a rematch with a reinvented version. by Deep. blue kasparov declined apparently being ready to leave that part of his past behind shu began working for compaq but moved on quickly and by 2002 he had retired completely from creating gaming computers and soon accepted a job in microsoft's research division in Beijing in 2003, where he was put in charge of the platforms and devices center and would go on to have a successful, albeit private, career in computing, as Kasparov became more interested in politics and became an outspoken opponent of the Russian government. , although sometimes he still talks about chess and gives lessons.

In addition to writing books on topics of both his long-standing interest in chess and his new causes, especially supporting freedom of expression despite his loss to deep blue or perhaps because of it, Kasparov has become an even more vocal advocate. vocal of the advancement of technology and its ability to improve the general standard of living by encouraging governments and businesses to embrace technological innovation Today, chess computers have surpassed any human chess player with programs such as the open source Stockfish and the ai alpha zero neural network, transforming the way the game is played far surpassing the abilities of deep blue that had defeated kasparov, although they are not usually as powerful as deep blue, these new programs have evaluation algorithms much more robust and sophisticated computers that are based on the idea of ​​singular extensions, in the end it seems that claude e shannon was right when he said that the weak but intelligent type b computers would inevitably achieve dominance;

These computers, however, did not make the game obsolete; while computer chess tournaments have gained some popularity; the main interest of players and spectators remains in purely human competitions; the use of computer-assisted analysis in These tournaments are strictly prohibited and the ideal match remains a mental struggle between two players without help from others, including chess programs, and where Kasparov had accused Deep Blue of playing chess suspiciously like a human. Today, chess players will accuse each other of playing suspiciously computer-like. Still, players are expected to use programs to aid them in their preparation, but even as the capabilities of both computers and humans continue to grow and progress by working side by side for mutual improvement, the two halves of blue deep are found in separate museums, isolated from each other. another will never again gather for its intended purpose deep blue waves rise high it's alone all content and black we're just waking up deeper skies hang over dim stars shattered by design it's happened david hello streams deeper waves rise around me you