Chess = tic tac toe...at least to computers in 10 years.
Sort:
That's why I'm playing as much as I can now, before the game is killed by technology.
Just keep playing humans...
This is a silly thread. People will always play chess as long as people enjoy chess and find it challenging. Arent there people who still play tic tac toe? Arent there still weight lifting competitions for humans even though fork lifts can lift far more?
fork lifts are different - no one knows the absolute best move in every case right now unless they let the computer take a year or so to figure it out to the n'th iteration. However, eventually it will be impossible for a human to beat a computer...they can draw potentially but it wont be possible for a human to beat the computer. Its just something to think about.
Reb wrote: This is a silly thread. People will always play chess as long as people enjoy chess and find it challenging. Arent there people who still play tic tac toe? Arent there still weight lifting competitions for humans even though fork lifts can lift far more?
A good point Reb, well made.... however, I think the word you were looking for is:
Hogwash!!
Chess will not be solved in 10 years or 50 years or 100 years. It shows your ignorance of the magnitude of the problem of solving chess if you think otherwise.
Computers will play better chess moves than humans and not lose games to them sure, but that's already true as far as you and I are concerned. Why wait 10 years, give up now.
No fork lift has ever beat me at chess.
No; it won't be solvable in 10 or 20 or even 50 years from now.
Consider this: The most recent commercially available processors have transistors which are 45 nanometers (10^-9; abbreviated nm) in size (32 nm transistor based processors are expected by 2009). A silicon atom, which is the primary material of a silicon transistor, is considered to have a radius (atoms don't really have radius'; the radius is really the distance between two of the atoms nuclei when bonded together) of 117.6 picometers (10^-12; abbreviated pm; a nanometer is 1000 picometers). Therefore, a silicon atom has a radius of 0.1176 nm. This leads to the undeniable fact that we are literally running out of space for the atoms. Then there is the quantum physics problems. Around 16 nm; conventional technology is not expected to work; largely because of quantum tunneling. Transistors will not act like transistors. So scientests have been working on developing quantum computers which take advantage of superposition and entanglement. The problem is; these quantum computers will not provide enough of a benefit for problems like this. They only provide measurable benefits for guess and check problems. With n possible answers; the number of steps (1 unit of time; generic) required to find the correct answer of a problem where every possibility is equal in time required to check is given by the square root of n. Chess is a guess and check problem. But not every answer takes the same amount of time to check. The time it would take to check one possibility would require limits on the number of moves to check. So depending on the constraints of the depth it will search; it will take a monstrous amount of time to check every possibility. To give you an example on how many possible combinations of moves there are; here is tabulation of the number of moves possible (I know there's one somewhere on this site; but I couldn't find it):
http://mathworld.wolfram.com/Chess.html
The time it would take a modern computer to find and store all this information is not something that you would live to see. In any case; the thing that really needs improvement is not the hardware; but the software. Deep Blue used software which was not exactly adept at making good moves. The reason it did so well was because of how far it looked (plys, or depth) ahead; which was possible because of the hardware. A better algorithm for the software; however is not exactly easy to create. It will take time; creativity; and hard work.
However; I do concede that eventually; if given enough time (time which we could use much more productivly); we could create a massive database of all of the non-redundant moves. However; searching this would take a very long time; even if it were thoroughly indexed. So there is no need to worry about the future of chess being threatened by computers.
Tipped wrote:
No; it won't be solvable in 10 or 20 or even 50 years from now.
Consider this: The most recent commercially available processors have transistors which are 45 nanometers (10^-9; abbreviated nm) in size (32 nm transistor based processors are expected by 2009). A silicon atom, which is the primary material of a silicon transistor, is considered to have a radius (atoms don't really have radius'; the radius is really the distance between two of the atoms nuclei when bonded together) of 117.6 picometers (10^-12; abbreviated pm; a nanometer is 1000 picometers). Therefore, a silicon atom has a radius of 0.1176 nm. This leads to the undeniable fact that we are literally running out of space for the atoms. Then there is the quantum physics problems. Around 16 nm; conventional technology is not expected to work; largely because of quantum tunneling. Transistors will not act like transistors. So scientests have been working on developing quantum computers which take advantage of superposition and entanglement. The problem is; these quantum computers will not provide enough of a benefit for problems like this. They only provide measurable benefits for guess and check problems. With n possible answers; the number of steps (1 unit of time; generic) required to find the correct answer of a problem where every possibility is equal in time required to check is given by the square root of n. Chess is a guess and check problem. But not every answer takes the same amount of time to check. The time it would take to check one possibility would require limits on the number of moves to check. So depending on the constraints of the depth it will search; it will take a monstrous amount of time to check every possibility. To give you an example on how many possible combinations of moves there are; here is tabulation of the number of moves possible (I know there's one somewhere on this site; but I couldn't find it):
http://mathworld.wolfram.com/Chess.html
The time it would take a modern computer to find and store all this information is not something that you would live to see. In any case; the thing that really needs improvement is not the hardware; but the software. Deep Blue used software which was not exactly adept at making good moves. The reason it did so well was because of how far it looked (plys, or depth) ahead; which was possible because of the hardware. A better algorithm for the software; however is not exactly easy to create. It will take time; creativity; and hard work.
However; I do concede that eventually; if given enough time (time which we could use much more productivly); we could create a massive database of all of the non-redundant moves. However; searching this would take a very long time; even if it were thoroughly indexed. So there is no need to worry about the future of chess being threatened by computers.
Or fork lifts. LOL
May 6, 2008
0
#11
i think that there are only a couple good moves for each situation (look at a vote chess game.) Usually all the "super programs" (fritz rybka shredder) come up with 1-2 different equally good moves when used together. that being said 80^2 is still a pretty big number as is 80^3 and 80^4 (40,960,000).
Thank you Tipped
To add on to the comments posted by Tipped:
If you consider all of the possible positions of the pieces on a chessboard, and all of the legal sequences of moves required to reach them, the possibilities outnumber by an order of magnitude the number of atoms in the universe. Therefore, no computer can ever possibly analyze and compare every possible position to play a "perfect" game of chess.
This is how chess engines play chess, they choose a position, analyze it and assign a score. Then, after a given amount of time, they play the move that has acheived the highest score. If it's impossible for the computer to analyze every position, there is no way it could ever be 100% perfect.
Theoretically, the computer can use various algorithms to weed out obviously bad moves and positions, but the number it would have to analyze is still astronomical (literally). So, while computers will probably approach very, very good chess, and perhaps even get to the point that no one can beat them, they could never be theoretically perfect, because there could never be a way to know for sure.
Also, because computers only do what they are told, there will always be a way to defeat a computer by playing by a strategy that plays to its weakness. Granted, you and I may not know what that weakness is, but the programmers who made it will know how it likes to play, and there will always be an "anti-computer" strategy for any given engine. They may be able to adapt if they're programmed very well, but they sill can only follow set sequences on instructions.
If you are serious about chess = tic tac toe, then maybe you should use this equation since it fits your idea better
Chess = Rubik's Cube... it's still not suitable but more accurate in your perception
I agree that chess is still fun and human vs human will keep it all the better
I said it before, you can't just make a more powerful computer to solve chess (there is not enough matter). You need to find some pattern or mathical equation that can make the problem smaller. That is the only hope of solving the chess problem we currently have.
take the sudoki example,
the theory is really fast, also I bet is very easy to program an app to solve it ( a lot of IF-THEN-ELSE) and recursive functions and nobody says SUDOKI ITS OBSOLETE!!
please check this out and tell me if will not be cool for a lot of years...
http://demuxer.blogspot.com/2008/03/cubo-rubiksudoku.html
does it really matter? i mean who wants to play against a computer that has 'solved' chess anyways. people 100 years from now will still play chess against each other. just because some computer 'solved' the game, doesnt mean any human will care.
and LOL at comparing chess to tic-tac-toe.
Can chess be solved? I don't think that anything is impossible, so it's very likely that some super-computer or savant of savants may figure it out, but who cares anyhow?
Like Reb pointed out, solving chess won't affect the participation rate in chess; if anything I see it as being a newsworthy event which may bring more to the game (a la Kasparov v Deep Blue").
If the solving day ever happens, I'll applaud the engineers and programmers for using such efficiency in their code and design to accomplish such a feat, and hope to see the efficiency used in other applications besides chess.
As for me, I'm looking forward to many many years of good fun and study with chess :).
dylan wrote: No fork lift has ever beat me at chess.
lucky... my uncles Bobcat destroyed me the other day. it sorta just crushed my king under 3 tons of metal. i tried the dutch defense, but it just proved flimsy...
Tic Tac Toe is a simple game...but really just a smaller board and only 2 moves. Imagine a 1000x1000 size tic tac toe board. Equally hard for a computer to solve (just for arguements sake) but no one would doubt that a computer can do it.
I think we'll see the computing power able to play chess perfectly at some point - and I think it will be sooner than you think. 10 years was an arbitrary number. The reason this will have an impact on the game is pretty obvious. Do you deny how much better most average chess players are today. Much of it is the result of computers helping to analyze positions and plan openings / move sequences far beyond where most normal joe blow chess players can think (2 moves ;-) Sure no person will likely be able to memorize everything needed to play perfectly but it will be interesting to see if it changes how play even at the advanced levels.
Will chess suffer - prob. not - it will certainly change the game though. When we have software that can pound everyone but a few of the elite into the ground already I am just curious how much of a difference it will make when chess is actually solved.
P.S. Chess is exactly like tic tac toe - just has significantly more move choices...its still a finite game...which will be solved.
It is almost hard to realize that soon chess will be nothing more than a tic tac toe game to a computer ... just as easy to play perfectly as they calculate every possible move to game completion in less time than a human can possibly do it.
What will happen to the game when it is "solved" by the next great fast computer. There is nothing different between chess and tic tac toe just that the board is much bigger and the pieces can each move differently.