no other player in the world ever can beat bobby fischer age 63..NOT EVEN MAGNUS =NOBODY
Will computers ever solve chess?
Sort:
Fixing_A_Hole wrote:
waffllemaster wrote:
...
A) Bigger than the moon.
B) Bigger than the sun.
C) More atoms than our solar system.
D) More atoms than our galaxy.
Answer is none of the above- More atoms than in the universe.
Answer is none of the above, aka all answers are wrong. Take answer C "C) More atoms than our solar system." since this is wrong the answer must be "less atoms than our solar system".
Since you add "More atoms than in the universe" this must mean there are more atoms in our solar system than in the universe. Since the universe contains the solar system this must mean there are negative atoms...
This is a scientific breakthrough! We're going to be in Nature, all the newspapers and television shows and possible even on the radio!
Apr 19, 2017
0
#764
The only way chess won't be solved is if we humans destroy our habitat(s) and ourselves before extremely high density storage units are developed and algorithms to solve the mathematics involved.
Everytime I was an episode of the 1960's Star Trek series I'm not only amazed at how backward the electronics look compared to today, I'm amazed that those of us who watched it when it was new thought the depictions of electronics were certain a century or more in the future. So don't underestimate computer progress. But do fear attempts to create computer consciousness: sentient computers will be our doom or trigger our conversion into Borg-like creatures: Resistance is Futile.
Apr 19, 2017
0
#765
What I find silly about mechanical devices outpacing humans in chess is that it means absolutely nothing in terms of how well we can play chess - though we can use it to help us.
We have pitching machines than can be modified to pitch the ball faster than Walter Johnson, yet it detracts nothing from Johnson's ability and 100-year-reputaation as the fastest pitcher ever.
We have machines that could hit a baseball twice as far as Mickey Mantle's record 565 ft. home run but that doesn't diminish the thrilling memories many of us have of seeing him play.
The same is true, not only of other sports, but of spelling, knowledge of foreign languages, etc. So what if we've created non-sentient, brute-force machines that outperform humans. We are STILL the more complete total package.
Runel0 wrote:
Nope. Supercomputer against himself sometimes wins sometimes lose.
When a supercomputer plays against himself he will always make a draw but if he plays against other computers he will sometimes lose or win
vickalan wrote:
ahah
Once you guys solve chess, we have the next game ready for you to solve.
Come visit us at the chess-variant forum!
Apr 23, 2017
0
#769
the_johnjohn wrote:
Runel0 wrote:
When a supercomputer plays against himself he will always make a draw but if he plays against other computers he will sometimes lose or win
Nope. Supercomputer against himself sometimes wins sometimes lose.
You mean another supercomputer
RookSacrifice wrote:
the_johnjohn wrote:
Runel0 wrote:
When a supercomputer plays against himself he will always make a draw but if he plays against other computers he will sometimes lose or win
Nope. Supercomputer against himself sometimes wins sometimes lose.
You mean another supercomputer
What's the difference between a supercomputer playing against himself (a supercomputer), and a supercomputer playing against other computers?
There is a big difference.
Apr 29, 2017
0
#772
There is a difference: the horizon of each computer is different (although there will be a large overlap). One computer might explore one line a little further, another will explore a different one a little further. So they make different errors due to the horizon effect.
This being said, as the horizon of a computer changes each move, they can also, as has been pointed out, make mistakes when playing against themselves.
The horizon is ultimately controlled by the software (the software can be coded to wait for any specified depth of analysis). There might be a difference in speed of different computers, but that still doesn't explain the difference between a supercomputer playing against himself (a supercomputer), and a supercomputer playing against other computers.
Apr 29, 2017
0
#774
a question like this can not really be answered....any answer given will be speculative....
May 8, 2017
0
#776
Will computers ever solve chess?
1. With or without opening book?
2. With or without ending tablebases for the horizon effect?
3. With or without the slightly rough edges?
For example:
K + 2 N vs. K = Draw........most times 
K + 2 N vs. K + P = Win...sometimes
Perhaps a question that someone should be asking is this:
Chess was used to evolve computers at the dawn of the computer age, can chess do it again?
May 11, 2017
0
#777
game_designer wrote:
Will computers ever solve chess?
1. With or without opening book?
2. With or without ending tablebases for the horizon effect?
3. With or without the slightly rough edges?
For example:
K + 2 N vs. K = Draw........most times
K + 2 N vs. K + P = Win...sometimes
Perhaps a question that someone should be asking is this:
Chess was used to evolve computers at the dawn of the computer age, can chess do it again?
I think you don't quite get the notion of "solve".
It is absolute. Opening books in this context can only be lists of positions that are won, drawn or lost, with no ambiguity, A table base for chess in general (a 32 piece tablebase) is a good way to think of it, but a weaker notion of solution involves only generating enough of the tablebase to be able to always generate a move that is definitely best.
In the game of draughts (checkers), this allows you to reduce the complexity almost by a power of 1/2. But in chess it is not quite so useful, as most pieces move in all directions (draughts only move forwards until they are promoted). Moreover, the game of chess has such a much larger number of possibilities that raising it to the power 1/2 would not suffice to make it feasible to compute.
May 11, 2017
0
#778
@Elroch
"It is absolute."
Good luck with that.
Opening book, is it used or is it purely computer logic, that is what I was asking.
Ending tablebases, working backwards from a 32 piece tablebase, I understand that.
Lookup tablebases, fudge to reduce very large memory overhead, find the position in the lookup tablebase, see the result, WDL, DTM, DTC, blah, blah, blah, once position is reached computer crunches actual moves, kinda like peeking over the horizon, I get that.
Slightly rough edges? Need a tablebase or lookup for that, no way around that.
I was asking if it would be possible without any of all that, purely computer logic.
Total positions? Legal positions, triangulation, transpositions, etc.
Will certainly have Zobrist hash key collisions, would probably need to use a bit board composite primary key to identify all possible positions.
"Solve"
I think I do get it man.
Good luck with all that, have fun.
It has been solved, the answer is 42. You don't have to take my word for it, just Google "Answer to life, the universe, and everything" http://bfy.tw/1nsj
a human brain is capable of solving chess , unlocking the time and resources is the key..
My money says if Bobby Fischer competed and played for the world at 65 years old, he wouldve been rated 2900+ standard over the board.
The federations snuffed him with poison because he wanted to win world again. fischer is proof that 2900+ is possible on a standard basis..