What do you think everybody has been doing all this time?
Will computers ever solve chess?
Sort:
May 13, 2016
0
#403
FirebrandX wrote:
johnsmithmyer wrote:
will computers rewrite opening theory and chess and chess theory? yes! is chess a question with a solution? no
It's been my experience that computers are merely extending existing theory rather than coming up with anything revolutionary. In fact, it's somewhat surprising just how sound pre-computer opening theory has been in general.
I know I've posted this before in this thread, but "solving" chess isn't required to reach a point where the computer playing simply never loses. The evolution of software will reach that point quite soon when combined with a large opening database. In effect, the positional and tactical considerations of a given position do NOT require calculating every atom in the universe, which is why mapping out every move of the game isn't important to playing a theoretical perfect outcome
Yes, a theoretically perfect game is much more possible. But actually solving chess, or proving that a game is perfect, will likely never happen. Not that this difference should matter to fans or players.
s23bog wrote:
Trying to win.
Of course, you want to make an engine that plays the best moves, i.e. tries to win. The objective mapping part is also in progress, all positions with 7 pieces on the board have been solved. It takes more than a bit of time and effort to extend this to 8 pieces or more, all of this is well known, storage space, time, etc...
Re counting. How would this work with a much simpler problem, tic-tac-toe for example? I can't really see how it would help.
My 2 cents.When the computer wins while playing the world champion the game is solved.
My 2 cents.When the computer wins while playing the world champion the game is solved.
Of course it would be possible to count positions without storing them, but it would be a useless effort. In the end you'd have a nice number but wouldn't have come an inch closer to solving chess. I don't understand the logic behind any of your suggestions, I'm sorry
gerberk wrote:
My 2 cents.When the computer wins while playing the world champion the game is solved.
Then we would've solved the game years ago which obviously isn't true
Once again, you guys are trying to re-invent wheels that already were rolling long ago:
Example to keep you entertained:
http://kirill-kryukov.com/chess/nulp/results.html
This discussion is like a bunch of guys sitting around drinking beers saying:
"Hey, think we can colonize Mars?"
"Yeah, sure."
"How will we do it?"
"I think we'll need a rocket."
"Yeah. A rocket...deep insight, Bob."
gerberk wrote:
My 2 cents.When the computer wins while playing the world champion the game is solved.
The problem is that a game being 'solved' has a set definition and is a very different thing from simply having computers exceed humans. What if Computer A beat the World Champion, but then one year later a newly-developed Computer B defeated Computer A? Did Computer A 'solve' chess one year ago even though it lost today?
s23bog wrote:
Well, to avoid duplication, there would have to be temporary storage.
One of the first things that children learn to do is to count. So, let's count. Stop looking at the problem from the eyes of someone who knows everything, at look at the problem from the eyes of a child.
Ok, so you're trolling. Time to leave this thread behind...
Probably but who cares , who wants to play with an unemotional opponent.
s23bog wrote:
And curiously, you point to a page with incomplete results. Maybe it isn't as easy as you make it seem.
Tip: I'm the guy saying it isn't easy and that weighing in decades after the fact and tossing out random thoughts you've had in 5 minutes of ruminating is disingenuous.
But hey, if you want to count, go for it. Let's see you put forth something...anything really...with actual results more complicated than basic multiplication. Up to now, you've put in nothing but conjecture.
Show us something more than a wide-brimmed hat...or should we just end this facade, and say "what, fish for lunch again?" and call it a day?
s23bog wrote:
RoepStoep wrote:
s23bog wrote:
Well, to avoid duplication, there would have to be temporary storage.
One of the first things that children learn to do is to count. So, let's count. Stop looking at the problem from the eyes of someone who knows everything, at look at the problem from the eyes of a child.
Ok, so you're trolling. Time to leave this thread behind...
No, not trolling. But you are welcome to leave.
No, only a troll could propose reverting to a childlike state of counting positions as a serious solution to solving chess
In the end, it must be counted. An idea, however beautiful, must be checked for accuracy, by counting the possibilities. Counting the tactics.
You can count less and less, but that is an illusion ( that one's on the right track doing that): by not counting what 'seems obvious to the human eye', we are missing tactics that a computer, by its very nature and functioning, cannot.
So by counting less we are missing some ideas, at least some of the time.
May 13, 2016
0
#418
You enterred college? How did you do that when you spell entered enterred?
We do NOT need to play every chess game to PROVE chess is a Draw with BEST moves. Take this example; is N plus 1 always n+1. We would sure believe it and state that it is true. We did not take n things, add one more and recount them for all number to say that it is true. Do you agree ? But infinity plus one is not infinity + 1. if it were than infinity minus one would be less than infinity. It's not, it's still infinity and so is infinty plus one. In all these cases ; reasoning is how we know. Not be doing everything.
Best play is ALWAYS a DRAW. Here is why...
This is probably a bad attempt at trolling...
You're probably right.