Man versus Chess Machine - David Levy
By Vishaal
on
Sunday, December 05, 2010
with
0 comments
He was born in London, England in 1945. David Levy was the Scottish chess champ at the age of 22 and Levy 
played John McCarthy a friendly game of chess at an AI workshop in Edinburgh in 1968 and he made a bet of 500 pounds with John at that time regarding the future of computer chess. In 1977, David defeated KAISSA. He defeated chess 4.9, 4.7 later and then in 1978 won his bet from McCarthy for several thousand dollars. David Levy is the author of an unbelievable number of books. He’s written about 40 to 50 books
on the subjects of computers, chess, and a wide range of subjects and is extremely prolific. He’s President of the International Computer Games Association and he helped organize the IBM Kasparov Deep Blue Match.
This is an excerpt from his speech at the computer history museum in January this year -
In August 1968 John and I started a bet that became a milestone in computer chess history. We were at a cocktail party in Edinburgh during one of the machine intelligence workshops organized by Donald Michie who was founder and head of the first AI university department in Britain. During the party, John invited me to play a game of chess which I won. And when the game was over, John said to me, “Well, David, you might be able to beat me, but within 10 years there’ll be a program that can beat you.” And I was somewhat incredulous at this suggestion.
I’d recently won the Scottish championship and it seemed to me that programs had a very, very long way to go before they got to master level. I knew of course of John’s position in the world of AI for which I had the 
greatest respect, but I felt that he simply underestimated how difficult it is to play master level chess and I was also a bit brash and I’ve always had a tendency to make somewhat large bets. So I offered to make a bet with John that he was wrong and he asked me how much I wanted to bet and I suggested 500 pounds which at that time was a little more than a thousand dollars. Now to put that into perspective, in those days I was in my first job after graduating university and the bet represented more than six months’ salary for me.
greatest respect, but I felt that he simply underestimated how difficult it is to play master level chess and I was also a bit brash and I’ve always had a tendency to make somewhat large bets. So I offered to make a bet with John that he was wrong and he asked me how much I wanted to bet and I suggested 500 pounds which at that time was a little more than a thousand dollars. Now to put that into perspective, in those days I was in my first job after graduating university and the bet represented more than six months’ salary for me. 
So John wasn’t quite sure whether to take the bet so he called over to our host, Donald Michie, for advice. And Donald was sitting on the floor a few feet away from us and he asked Donald what he thought. And Donald immediately said to John, “Could I take half the action?” And that of course gave John a lot of confidence and so we started the bet, we shook hands, and that’s how it started with each of them betting me 250 pounds that I would lose a match to a computer program within 10 years. Later the bet grew bigger. The following year, Seymour Papert and Edward Kostrowicki [ph?] joined the list of opponents and the final amount at stake when we ended the bet was 1,250 pounds. But I had never felt that I was gonna be in any trouble and it turned out that I was right.
In August 1978 at the end of the 10-year period, I played a match at the Canadian National Exhibition in Toronto against the reigning world computer chess champion program and although they put a very pretty girl facing me to make the moves and smile at me when I was thinking, I still won the bet. In 1979, the following year, I made another bet for a further five years. This was with Dan McCracken and was for 1,000 dollars and again I won.
But by the time I won in 1984, I could see the writing on the wall so I got together with Omni Magazine and I said to them that I would like to offer a prize for the first team or the first program to beat me whenever that might be and said that I would put up 1,000 dollars if they would add 4,000 of their money to it. And so we did. We announced a 5,000 dollar prize with absolutely no time limit.
By 1989, the group at Carnegie Mellon University, of which Murray Campbell was a key member, had created Deep Thought which was a veritable monster of a chess computer and it started to score major successes in tournaments involving very strong grand masters. It even won a tournament in California ahead of a former world champion, Mikhail Tal and I knew then that my number was up. And sure enough, when I was challenged to a match at the end of 1989, I was horribly crushed by a score of four to zero, but I was reasonably satisfied because I’d lasted for 21 years since the original bet. Fortunately, Garry Kasparov was willing to take up the battle on behalf of mankind and that gave the struggle 
to improve the best programs even more emphasis. The Man Versus Machine Contest and my bet in particular elicited some interesting and provocative comments from various experts. In the human chess world, I encountered two diametrically opposing views from two former world champions.
to improve the best programs even more emphasis. The Man Versus Machine Contest and my bet in particular elicited some interesting and provocative comments from various experts. In the human chess world, I encountered two diametrically opposing views from two former world champions. Mikhail Botvinnik who was a real titan of chess who held the world championship for 12 years during the period from 1948 to 1963 with a couple of gaps said to me, “I feel very sorry for your money, David.” On the other hand, the Dutch mathematician Max Euwe who held the world championship title for two years from 1935 to 1937, as soon as he heard about my bet he wanted to take a share of my side, but I said no.
And in the world of computing, Monty Newborn made a prophecy during the 1977 World Computer Chess 
Championship, one year before I played before my first match, and this was a prophecy that was somewhat optimistic in its timeframe, but it came through more quickly than I believed possible. Monty said, “Grand masters used to come to computer chess tournaments to laugh. Now they come to watch. Soon they will come to learn.” And learning they have been. Grand masters have been employing chess programs as analysts for several years and since 1987 huge databases of games from master play have been employed to help strong players prepare for games in tournaments against specific opponents. In addition, there are databases of end game positions that have taught the chess world the truth about some positions that have been incorrectly analyzed in the literature. In one case, the configuration of pieces in the end game that had been thought and stated in all the books to be drawn was discovered through computer analysis to be a win for one side.
Category: Articles
Subscribe to:
Post Comments (Atom)
POST COMMENT
0 comments:
Post a Comment