Author: Christopher R. Dorr
Date: 14:56:53 01/05/99
Go up one level in this thread
Personally, I find this a very interesting discussion. Professionally (I'm an IT manager at a bank) I think this topic will be very relevent in the next few years. On January 05, 1999 at 14:53:18, KarinsDad wrote: >IMO (for what it is worth), I think Microsoft could do a good job creating a >"better chess program". I also think that most people here will disagree with me >on this topic, hence, I think that if you created a poll question on this, most >of the people would probably indicate that Microsoft could not create a much >better chess program. > >My reasoning for why MS could do better is as follows: > >1) Knowing several Microsoft engineers and listening to them, I have come to >realize that there are a lot of bright development engineers there. Obviously, >there are exceptions to this rule, but I think that they have the talent to >build a better mousetrap (or chess engine). No doubt that MS has some of the best minds in the world, but for something as advanced as this, I believe that it would take more than just smarts to make any serious advances in computer chess. A good friend of mine is a geophysicist at a school down south. He was working on some modeling software for an oil company last summer. One of the guys he was working with was a mensa-member hired-gun programmer brough in specifically to assist on this project. Apparently, he was a great programmer, but he didn't help much because he had no specialized knowledge of the subject. It would have taken him quite a long time to acquire this specialized knowledge. In arcane subjects like these, I don't think having good generalized programming skills and some serious smarts is enough to get to the point where you can start breaking new ground. > >2) There is a lot of information on the Internet currently (and in books and the >ICCA journal, etc.) that would enable them to catch up fairly quickly, possibly >as quickly as 9 to 12 months. > Perhaps they could catch up in, let's say a year. That's when they *start* to become really productive. Figure several months for design, coding, testing...I don't think they could start making a major 'Microsoft Impact' on the field for 2-3 years at the earliest. And much of the info available on the internet is probably a bit behind the times. Team IBM, ChessBase, Ed Schroeder, and the rest probably aren't real keen about putting their ideas out for public perusal. >3) Microsoft has a lot of resources, money, high speed equipment, and talent. >Just look at what IBM did with Deep Blue. With Microsoft's money, they could >purchase the source for Deep Blue, reverse engineer it, and improve upon it. >They could also hire some of the original Deep Thought/Deep Blue engineers and >possibly even other engineers from other chess engines. But that brings us again to the question 'What is the best possible chess program that can run on current hardware?' Is it 2600? 2800? 3000? How much better will this 'perfect' program be than the best that exists right now? Unless team MS comes up with some profound new idea never before ideated or implemented, then they are stuck with this limit. And I don't believe that this 'profound new idea' will be discovered by anyone but somebody with years and years of chess programming knowledge and experience. > >So, Microsoft has the resources to create a better chess program. There is no >doubt (in my mind) of this. However, they will not ever do it since there is no >market share (from their perspective). Hence, my thought is that we all know the >answer to the question. We may disagree on it due to likes or dislikes of >Microsoft, but the bottom line is: 50 talented engineering programmers with a >lot of high speed equipment can reverse engineer and research current technology >in any computer subject and do a good (if not great) job of improving upon it >within 2 years. I agree that nobody is ever going to pub a billion dollars into this. MS has no reason to even think about such things. But I think that even if they did, they wouldn't make much of a splash. Their product would certainly be better in some ways, but it would be earth-shattering, and certainly wouldn't change the state of the art. > >KarinsDad > >PS. I think that the current level of chess program technology is actually quite >sophisticated. However, it can be improved upon (as is seen by the fact that the >programs get just a little bit better every year). > > >On January 05, 1999 at 03:45:13, Lawrence S. Tamarkin wrote: > >>I recently met a former Microsoft department head, who is now a chess coach and >>using many of the availble playing programs, databases & tutorials. What he >>told me I found truly amazing! He said that the currently availble programs >>were from a programming point of view rather unsophisticated and trivial in >>design. He suggested that if Microsoft (or other large software maker), decided >>to get involved in creating these things, the results would (or at least could), >>blow away the currently existing things in the marketplace. >> >>Honestly, I think he is wrong about this, but I certainly would like to see a >>poll question about it, to see what other's think. >> >>mrslug - the inkompetent chess software addict! >> >> Chris Dorr
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.