|The Week in Chess Magazine
Sponsored by the London Chess Center
|TWIC Home | The London Chess Center | | Shop|
Feng-hsiung Hsu one of the three major programmers of Deep Blue has released the following open letter. He adds "It is probably historically significant as well. I believe that this is the end of the Deep Blue saga."
This open letter is about what happened after Deep Blue's 1997 match win over World Chess Champ Garry Kasparov. More specifically, it is about Mr. Kasparov's post match challenge to the Deep Blue team for a new match. You will see below that I had been doing everything within my power to make the new match happen. However, as far as I can tell, Kasparov has now effectively withdrawn his challenge. Unless Kasparov has a quick change of heart or someone changes Kasparov's mind through persuasion or enticement, there will not be a new match. Neither scenario seems likely. Therefore, I will now also use this opportunity to say my thanks and farewell to many kind people in the chess world and the computer chess world.
My name is Feng-hsiung Hsu. I am one of the three persons who designed and programmed Deep Blue, the IBM chess computer. The other two persons are Murray Campbell and Joe Hoane. Murray joined the project in 1986, a year after I started the project at Carnegie Mellon University in 1985. Joe joined the project in 1991 after Murray and I moved to IBM Research. The three of us had spent close to 30 man-years on the project when Deep Blue won the match.
At the end of the 1997 match, Mr. Kasparov challenged the Deep Blue team for a rematch. I believe that all three of us felt that we had a personal obligation to honor Kasparov's request. But the decision was not really up to us. One further complication was that Kasparov had earlier made a groundless and false accusation of cheating against the Deep Blue team and IBM itself. I personally considered the accusation an insult and I doubt that any corporate officer would think otherwise when his or her company is similarly accused.
A few months after the 1997 match, it became clear that the chance of a new match between Kasparov and Deep Blue was remote at best.
Murray, Joe, and I moved on. We gave talks all over the world. We wrote and published technical articles. I wrote a book on the whole project, from Carnegie Mellon to IBM, although it is not published yet.
Meanwhile, for over two years, Kasparov repeated his challenge for a new match. In some of Kasparov's remarks, he even suggested making the new match a world title match. Unfortunately, Kasparov also continued with his accusations, changing from one conspiracy theory to another as time wore on.
There was little that I or the rest of the team could have done about Kasparov's challenge. However, I was planning to leave IBM. Since IBM was not doing anything with the Deep Blue chess chip, if I could secure the right to the chess chip from IBM, then it would be possible to build a PC based chess machine with playing strength comparable to Deep Blue. It would be equivalent to a "poor man's Deep Blue", but with the latest chip technology, it should be every bit as strong as Deep Blue, if not stronger.
I left IBM a few months ago. Before I left, I spent a small personal fortune to get the right to the chess chip. Officially, I was getting the right to commercialize the chess chip. My real purpose was to answer Kasparov's challenge. When I said goodbye to Murray and Joe, I made an implicit promise. If Kasparov was serious about his challenge, I would see to it that Kasparov got his match.
Since Kasparov had issued his challenge repeatedly and publicly, I assumed that I only needed to find sponsors for the match. There was some interest from several potential sponsors, but the sponsors had a major question that I could not answer. The sponsors were skeptical that Kasparov would play a new match, let alone playing a title match, with the new computer. I argued that Kasparov was a proud man and there was no way that Kasparov's pride would allow him to back out. The sponsors did not seem convinced.
So about a month ago, I made contacts with Owen Williams, Mr. Kasparov's agent. In my first email to Owen, I asked directly whether Kasparov was serious about the match and whether Kasparov was indeed willing to play the match as a world title match. The initial response was not encouraging. Owen ruled out the possibility of a title match immediately and was noncommittal about the new match. In reality, I was not unhappy that Owen ruled out the possibility of a title match. I don't believe that a computer should be the World Champion. Personally, I believe that the title should be for humans only. Some of the potential sponsors did indicate that a match with the title on the line was certainly more desirable, although I believed that it might be okay if Kasparov was willing to state that he would treat the match as seriously as if it were a title match.
Owen's noncommittal attitude toward the new match was troubling. I cannot conscientiously ask the potential sponsors to spend any more time on the match proposal without at least some sort of commitment from Kasparov. I pursued Owen and kept on asking for some sort of clarification. After a long sequence of email exchanges, Owen sent me a message vaguely suggesting that Kasparov was not interested. I did a double take and asked Owen point blank, "Can I safely assume that Garry is not interested in a match any more?" Owen replied, "Garry does not want to be involved in any way". Then added, "I have now closed my file". Throughout the entire sequence of email exchange, Owen never said a straight yes or no to my question of whether Kasparov was interested in a match. But the last email has only one reasonable interpretation. After over two years of issuing public challenges, Kasparov is no longer interested in playing a new match when the answer to his challenge comes.
I don't know the reason behind Kasparov's decision. Maybe his priority changed. Maybe Kasparov felt offended somehow in my dealings with Owen. It is no longer of concern to me.
Even if Kasparov somehow changes his mind, the chance of a new match is now very slim. Match sponsor(s) would still have to sign the dotted line. I would need to find funding sources to build the machine. But given what has happened so far, it would be very hard to convince match sponsors or funding sources that the match will happen. Kasparov could change his mind yet again or come up with match conditions that are unacceptable. Anyway, I give up. I have now done everything possible to make the match happen. I have fulfilled my promise to my team members. Time to live my own life.
There are people who are interested in seeing the chess chip commercialized. I am sorry. It is unlikely to happen. The Deep Blue chess chip is not commercially viable as it is. The price charged by the chip vendor is too high. To be commercially viable, I would have to design a new chip that is not vendor specific. Without a new match, it is difficult for me to muster the energy to create the new chip. The only chance that you would ever see the chess chip commercialized would be if someday I decide to build a shogi chip. Then you might see a new chess chip designed as a by-product. I would not hold my breath though. I have some other interests that have higher priorities.
The computer chess world had treated us well and the chess world had been kind to us. There was some apprehension in the chess world when we arrived on the scene. I remembered that a few years ago when Mark Crowther started his internet chess magazine The Week In Chess, he expressed his concern that Deep Blue would destroy chess. I think I can say fairly that Deep Blue did not destroy chess. There was perhaps even a mini boom in chess popularity as a result of the Deep Blue matches. It would have been nice if Deep Blue could have done more for the chess world. I was hoping that I might be able to do something beneficial to the chess world with the new machine after the match. Well, I did what I could.
There are many people that I am thankful to. I value greatly the time Grandmasters spent with us, helping us to create a better chess machine. The IBM management provided us with the ideal research environment that allowed us to pursue the dream. Without the Computer Science Department at Carnegie Mellon, the project probably would not have happened in the first place. There were numerous people at Carnegie Mellon and elsewhere that were instrumental in getting the project going. We owed greatly our success to computer chess pioneers before us. Finally, I have to give my thanks to Kasparov himself. It is a shame that there will not be a new match, but the two Deep Blue matches were the most exciting experiences in my life, and Kasparov, our worthy opponent, played the central role in the experiences.
Thanks for reading this. Have a happy new millennium.
Sincerely, Feng-hsiung Hsu