Chess Will Never Be Solved. Why?

"The way to know with absolute certainty is brute force solving, so your argument is circular...you cannot ever establish that a "chess AI" is 100% perfect until you solve the game. This whole thread was done better the first time around...this is just redundant rehashing.">>>


Of course it is but it's what they enjoy. Getting stuck in blind alleys and continually bouncing off the wall at the end only seems like progress to some.

However, a brute force full solution to chess is, at the moment, inconceivable. No way to track it, store it or evaluate it so it would have to be done via successive approximations. That's a necessary but clumsy method, since a lot of room has to be left to accomodate suprising lines.

My son works in AI. He's a physicist but his masters degree was in maths. He reckons that depicting chess mathematically is impossible. By that, he means "for a number of generations".

#299
"The prospect of solving chess in a brute force manner hinges on a breakthrough in scalability in quantum computing."
++ A quantum computer is the only way for strongly solving chess i.e. a 32-men table base.
Weakly solving chess takes 5 years on present cloud engines. Sveshnikov 'practically exhausted' B33 in 1988 on his own, without engines, without table bases. So to do 499 times more with cloud engines, table bases and grandmaster assistants is plausible.

"This is an area where the physics itself isn't yet fully understood."
++ The physics is understood. Quantum computers are even commercially available for rental.
https://www.dwavesys.com/solutions-and-products/cloud-platform/ 

"consider Fermat's Last Theorem"
++ Also the Four Color Theorem etc. Provability is a higher degree of truth. Also the Riemann Hypothesis and the Goldbach Conjecture are believed to be true though not yet proven.

"The physics is understood. Quantum computers are even commercially available for rental." Yes quantum computers exist, but the physics which isn't yet well understood is the the loss of coherence which is critical to scalibility.

#297

"Do you really think things have improved so much that they can correctly analyse the theoretical value of those positions in five days?" ++ Yes. It is 5 days each, so that makes 10 days. Moreover, they have already been studying the positions for months before.

"Yes; ICCF has the same rules except it doesn't." ++ A win in ICCF could be a draw in OTB play, except it does not happen, so the ICCF results are in practice the same as if it were OTB. Perfect play in ICCF is just the same play as in OTB.

"I may well have posted three above." ++ I do not deny the existence of the long mates, I just observe that such positions with long checkmates are not reached in real play. In ICCF they can claim a win in say 400 moves in a 7-men endgame table base, but such claims do not happen.

"apart from a handful of one offs Marc has generated only pawnless endgames"
++ The 8-men table base is not yet released. I presume the longest checkmate with 8 men will still stay with a pawnless position just like it is now and just like with 6 men. Likewise I presume it will be the same with 8 men... with 10 men... So the 4 men long checkmate with pawn is the anomaly.

There are no quantum computers yet, apart from toys that have a lot of faults. Renting them out would be part of the process of trying to control the problems with them.

#304
"the physics which isn't yet well understood is the the loss of coherence which is critical to scalibility"
++ Yes, loss of coherence limits scalability, but it is not a physics problem, but a technological problem. There are extraneous influences like cosmic radiation or thermal disturbations because of nonperfect liquid helium cooling.
There are also technological solutions with some redundancy to autocorrect.
Strongly solving chess i.e. from 7 men to 8 men to 9 men... to 32 men requires 146 qubit.

Chess is indeed an exact mathematical problem. There is a finite set of 10^44 legal positions with subsets of 32, 31... 9, 8, 7, 6, 5, 4, 3, 2 men. There is also a relation 'position (FEN) B can result from position (FEN) A' according to the Laws of Chess. So solving chess is finding a path from the initial position towards either a known 7-men draw, or a dead end of 3-fold repetition.

Chess is indeed an exact mathematical problem. There is a finite set of 10^44 legal positions with subsets of 32, 31... 9, 8, 7, 6, 5, 4, 3, 2 men. There is also a relation 'position (FEN) B can result from position (FEN) A' according to the Laws of Chess. So solving chess is finding a path from the initial position towards either a known 7-men draw, or a dead end of 3-fold repetition.>>>

It isn't. I'm afraid that you really don't know what you're talking about. In order for it to be "an exact mathematical problem" it must be possible to depict chess in terms of mathematics. Specifically, in terms of mathematical equations. It's impossible at least for the forseeable future and therefore your claim is not correct.

Aside from me just pointing out that you're obviously mentally retarded, might I ask you to kindly point out where and how I contradicted myself? If you insist on making such childish claims at least you could have a go at backing them up, to give me an even bigger laugh?

Do you imagine that the first paragraph, a direct quote from tygxc, was written by me? Obviously I was quoting him and explaining why he's wrong. People here have mentioned your complete inability to understand what you read. It's perfectly true.

There must be some reason why you never manage to understand what you read. There are others here, very like you. It it the result of mental illness?

     Coolout, just look at post #307 and you will see that you were indeed mistaken. You do such things so often that it's obvious to everyone you don't pay attention and then react like a four year old when caught out.

Lol.  The amusing part is that Coolout's obvious and simple error is the result of Optimissed's years long inability to use quote functionality correctly, but that is also due to chess.com's threadbare quote functionality .

Award Coolout the loss, but no win for Optimissed, or the devs.

They stated it repeatedly...not in clear enough terms for you, though.  Having argued with Tygxc for endless eons, it is probably hard for them to imagine that anybody could mistake his very clear posting style for Optimissed's reply.  The ">>>" is also a clue, by the way.

You might want to repaste them in a text color that at least half of chess.com does not see as invisible, then ...

(it never ends, does it?)

You, being cool and all, have probably opted for a black/dark background.  So you can see white text.  The default background is white, and so your words "posted for posterity" are not readable, and it looks like you just hit enter too many times...

This is why people often choose blue text for in-line replies, by the way, which I believe you railed against (also without understanding) at some point.

I often don't bother with it. I wrongly credit people like you with the ability of seeing what I'm quoting from, since it's always completely obvious. Or it should be, if you had eyes and the ability to open them. Anyway the quote thing often doesn't work ... you can't get your comment into the white area and so it has to be on the same area as the comment you're replying to and then you have to highlight it some other way. If you were grown up, you'd be able to understand what people are replying to without imagining they have to use the quote facility. It's just one of the probably hundreds of glitches on this site that they never see fit to sort.

So win for me, loss for you.

You mean you think that it's a win for you, and a loss for Coolout.  I've never really had any problems working around your technical issues.

So far as Coolout goes, if he were actually capable of referring to how he thinks I'm contradicting myself, then by this stage of the proceedings, even that would be quite a surprise. The man is completely gone, for the most part. Same as tyg but different.

What I think is a lot more likely to be functional that what you think. That's been well proven, over time.

Confirmation bias.

If you were capable of making a meaningful point or putting forward a useful argument or comment, then I would not hesitate in applauding you for it. The fact is that usually you aren't .... so much so that it's expected that you will be extremely confused by everything. Some of the best comments I've seen you make are actually re. computer related technical issues. You do obviously know your way round them and you should play to your strength.