All the best, Joe Lux
Thanks for the link. I’ve been interested in computer chess ever since Blitz grew up only 90 miles from me around 1980.
Good article.
This reminded me of the Futurama episode “Mars University.” Two robots (Oily and Fatbot) are playing Chess.
But, while they can’t solve it today, the computers of ten years from now? Twenty? (And consider that twenty years from now the term “CPU” may be something you only find in history books. Instead, you’ll have the “NCA” for Neural Cognitive Array or somesuch…)
This grab-bag of partial truths is not a good advertisement for the rest of the article.
Who wrote the article? Surely not Kasparov.
He mentions poker, which is a game not only of cards, but also of money management. Can we say that poker, requiring additional skills, is superior to bridge? Once, in a chess club where certain people occasionally chose to play bridge, I told one of the bridge players that chess was a superior game because its lengthy period of development had rendered it so. The bridge player became audibly upset, and a master-level chessplayer assured him I was joking.
Steve Dann once noted that I would make remarks such that people would take them seriously. I think he had in mind my letter (probably not published) in which I asked the editor of Chess Horizons whether the opening classification in Chess Horizons then using the letters R, E, D was intended to be significant. (Yes, there was such a classification in pre-Informant days.)
But seriously, folks, isn’t chess a better game than bridge? (Say yes.)
Almost, but surely Bughouse is a better game than bridge.
From the article- “Today, for $50 you can buy a home
PC program that will crush most grandmasters.”
How many will care when the first computer achieves a 4000 rating or crushes all grandmasters?
For chess the most crucial role for a computer will be a tool to teach.
For $50 I want to buy a home PC program that will explain the moves/variations it choses. And why White is slightly better, White is clearly better, or Better is. Or how about the moves I was thinking about?
Will there be a home version of HAL in our lifetime or some reasonable facsimile?
Which partial truths? I see two. “Nor will they ever come to pass,” based solely on the notion that one should never say never. And also I’m not condifent that the technology hasn’t been conceived that might someday solve chess. It may be a matter of infrastructure and the state of the art not being up to the task of the conception.
Or, tongue firmly in cheek and not intended to be taken seriously, the notion that a computer could be “unbeatable” at chess goes back at least as far as the Star Trek episode “Court Martial” - we just have to wait for the reality to catch up to the fantasy. 
That aside, I’m sorry I conveyed the impressison that the quote advertises the article. I was intending to convey that the notion of, “mate in 38,884 moves,” reminded me of that episode of Futurama where mate is declared at start of game. (Which I also enjoyed - the episode and Futurama generally.)
And I should tip my hat to gotfuturama.com/ for the screencap of the episode.
Given the amount of time computers will be required to “solve” chess, one is reminded of an old sci-fi story where a computer was given the task of solving an ultimate question. After billions of years, with all civilizations gone and no one to give an answer, with the universe in its last seconds before entropy, the machine finally solved the problem and gave its answer: “Let there be light!” I cannot remember the author, but it was a great story.
Unfortunately, the present chess algorithims still have a glitch and make some of the same mistakes that that they did back in the 1980’s. Wetware is still better than software.
“The Last Question”, by Isaac Asimov.
Jim
There was a cartoon along the same line, with a guy standing at his blackboard, filled with equations which ended with:
= LET THERE BE LIGHT!
Carbon will never compete with silicon in pure number-crunching.
But the capacities of carbon far exceed that of silicon (for the time being!) in pattern recognition.
Style is a funny (and very human) thing in chess. Think how distinctive the styles of Carlsen, Topalov, Anand, and Shirov are. They all work closely with similar chess software, yet they all have radically different styles. And when each player is in form, they’re extremely close in strength.
Only the first part is true.
The technology to solve chess already exists. What is lacking is money and/or time – more money spent on hardware would mean less time. Admittedly, ludicrous quantities of one or the other would be required.
Go-as-you-please checkers was solved by humans a long time ago – a forced draw. Chinook has apparently solved three-move-restriction from the initial position – again a forced draw. Checkers as a whole is not yet solved.
BTW, the “looked-down-upon” was totally unnecessary. I am not alone amongst chess-players in highly valuing checkers.
Ah, I think we agree then… And, as you allude to, just because a game is solved doesn’t mean that it isn’t still fun for humans to play. 
I don’t like checkers for only one reason… I always seem to lose. 
And I only mostly lose at chess. :mrgreen: (Only kidding - I like checkers once in awhile, even though I always seem to lose. ) And I certainly don’t “look down” on checkers - I wonder how many chess players do?
Ah, yes, Isaac Asimov. I should have remembered; temporary brain freeze or a “senior moment”. He wrote so many stories about computers and robots.
When I taught in a school we tried to interest special ed students in a variety of games to stimulate motivation and improve their concentration. The only game they liked and took to was chess. They found checkers, go, and other board games to be easy to learn but boring to play. They liked the variety of the pieces and that no two games seemed alike. Chess also appealed to their more tactile oreintation to learning. As one kid said, “Chess is funner than checkers.” I still respect checkers because when you really get into the game you realize how deeply a strong player has to calculate.
What’s all this talk about chess being “solved” (even at a practical level) in the near future? I do not underestimate the power of the computer, but chess (six types of pieces, 64 squares, exponentially greater branching) is vast compared to checkers (two types of pieces, 32 squares). Let’s see the math.
Granted, there are certain practical issues for go-as-you-please chess (Anand-Kramnik at Corus aside, the Petroff can be sterile, for example).
But the Open Sicilian? The Grünfeld? I find it hard to believe that certain positions in the transition from opening to middlegame in such openings could be “solved” (even at a practical level) in our lifetimes. (Even if the given position is intuitively dynamically equal…)
Depends on what you mean by “solved”. I expect that today’s high end computers could already calculate and store all of the possible winning permutations and thereby become unbeatable. However, I doubt rather strongly that any human could even begin to store all of that information. And if a computer is eventually able to come up with an “equation” that yields all possible winning moves, instead of storing all those permutations, that equation would also be too complex for a human to remember and utilize. So, just as cranes have not eliminated the World’s Strongest Man competition, computers will not eliminate human chess tournaments.
Really?
6-piece tablebase requires something on the order of one TB of storage. Yes, this is trivial: traveling GMs could put this data on an external hard drive. (Edit: just found this cool service.)
But now do the math: 6 to 7, 7 to 8, 8 to 9… Let me know when you start to have storage problems. (We’ll ignore the significant calculation time for the moment.)
There are only so many particles on Earth.
OTOH: the number of positions in chess has been estimated at a relatively small 10^50. So a narrow critical path (as in checkers) might reduce the number of possible games by a factor of (throwing numbers around for amusement) 10^100 or so.
I played the Greenblatt program in 1977 and made pessimistic predictions about the future of computer chess. Wrong wrong wrong.