Martin Meuer: "I Wanted to Do Something Different in Conjunction with Computers"

Время публикации: 28.06.2012 02:03 | Последнее обновление: 28.06.2012 14:44

Recently Hamburg hosted an International Supercomputing Conference (ISC). Jan Gustafsson held a simul in the framework of the conference, which was already reported on our website. GM faced 8 amateurs who were using the help of Houdini after every third move. Аfter the event we had a chance to ask the executive director Martin Meuer several questions.

CN: Mr. Meuer, who was the author of the idea?

M.MEUER: I wanted to do something different in conjunction with computers. Furthermore, playing against 8 amateur players would have been an “easy  pick” for Jan Gustafsson. A combination of players and chess program  created a healthy balance and a real battle for Gustafsson.

CN: What purpose does it have and what did you expect from it?

M.MEUER: This year we decided to provide more networking opportunities to our  attendees to get in contact with each other more easily. One of these  events was the simultaneous exhibition. I also expected that there  might be a lot of interest in chess in the HPC community. To be honest  I have been very pleased by the attention the match got from our  participants. Some of them asked to make this a permanent institution at ISC. So, we are really considering to invite Jan to ISC'13 next year in Leipzig.

CN: Tell us about your father's and your own work with Botvinnik.

M.MEUER: My father got to know Botvinnik when he gave a presentation in Mannheim about his chess program Pioneer/Chess Computer Sapiens (CCS).  As you know, Botvinnik tried a different approach with his chess  program Pioneer. He wanted Pioneer to really "understand" a chess  position rather than to compute the best move with the brute force  method. My father helped Bovinnik to get two work stations for his  project. Afterwards, Botvinnik became our guest at my parent's house in  Daisbach several times. He really became a good friend of our family  and invited my father to his dacha in Moscow.

I myself have been involved in the CCS project in the context of a student research project during my study of computer science in  Karlsruhe. A student colleague and I had to program the opening  library of CCS and we have been invited to Moscow for coordination  purpose with Botvinnik's staff. He led a small team of 3-4  programmers, who were all weak chess players. This made it sometimes quite difficult for them to understand what Botvinnik really wanted.

Unfortunately, the project never came to an end. Botvinnik and his  team had an office right at the Moscow Central Chess Club and I was  really impressed by the surroundings.

CN: How do you think, when will computers start to understand everything  in chess better than people? There are positions in which computers aren't very strong, so when do you think it will be completely impossible to play against them?

M.MEUER: Today's chess engines are already much better than the best chess players in the world. However, it is all due to the computational power. I really regret that Botvinnik was the last one trying a different approach.

CN: Another topical question: now it's quite popular to transmit games  with the computer analysis, but that is just a deception of the  spectators! For example, in one of the games of the Tal Memorial Aronian made a perspective sacrifice of  a pawn, but the computer showed: that's a mistake. When is it going  to give adequate evaluations?

M.MEUER: I am not at all convinced that computers will ever understand a specific position like a chess grandmaster but I have to admit that my  knowledge of chess programming is not good enough to provide you with  a well-founded answer.


Смотрите также...