@7764
I estimate 3 cloud engines of 10^9 nodes/s as 3000 desktops of 500 $ each,
written off in 5 years hence 3000 * 500 $ = 1,500,000 $.
To weakly solve chess you need only find 1 move for black to draw against whatever reasonable moves white tries. Hence Sqrt (10^34) = 10^17 positions relevant to weakly solving Chess.
3 cloud engines calculate in 5 years:
10^9 nodes/s/engine * 3 engines * 3600 s/h * 24 h/d * 365.25 d/a * 5 a = 5 * 10^17 nodes.
It really depends on how you define the parameters of "solved". Either practical or theoretical. There will be a point where the game will be vastly different depending on the mode.
With time controls, playing unconventional yet viable lines or even other modes like Fischer Random are all possibilities imo.
@7766
"It really depends on how you define the parameters of solved".
++ Ultra-weakly solved means that the game-theoretic value of the initial position has been determined,
weakly solved means that for the initial position a strategy has been determined to achieve the game-theoretic value against any opposition, and
strongly solved is being used for a game for which such a strategy has been determined for all legal positions.
The game-theoretic value of a game is the outcome when all participants play optimally.
We are talking about weakly solved here.
@7740
"Chess will be solved eventually."
++ Perhaps. 5 years and 3 million $ is a huge obstacle.
No it isn't. If it were true, it would be a tiny obstacle and well within the budget of a higher education establishment. That's just an excuse because $3m and 5 years is a pinprick. $3m doesn't start to cover it, as I explained several hundred posts ago when I costed it more accurately.
You won't listen to proper arguments and you only wish to spam for the Stupid Party, of which you seem to be Propaganda Minister. Also 5 years is impossible.
Yeah, but have you ever seen a right triangle? I'd prefer basically anything else.
Good God, wtf 600 plus new posts, you know any interesting groups i can debate with?
Can you actually debate interestingly?
I don't know of any, unfortunately. It's really quite good because I can get on with some work.
yes, i hope so only one way to find out
OK what is it to be a slave?
to have a bond with someone, with emotions like affection and trust, trust being super underrated and undercreditted with how intense and important it can be. and to be highly influenced by that persons actions and ways of living.
Perhaps that doesn't align with most normal perceptions of slavery but I'll take your point because, for you, it's a special relationship. Quite similar to the one I have with my wife, after being with her nearly 40 years. In your case, I wonder how much the significant person's actions and ways of living are an image created by something that can be considered akin to a propaganda industry? Even so, I don't doubt that many people can and do form a special, personal bond with the subject of such an image, which helps them during the course of their lives. I also wonder, though, what would be your reaction to the suggestion that such help may not the product of the subject of the image (being held up as an example of rightness and goodness) so much as the product of the strong feeling of trust in itself.
@7766
"It really depends on how you define the parameters of solved".
++ Ultra-weakly solved means that the game-theoretic value of the initial position has been determined,
weakly solved means that for the initial position a strategy has been determined to achieve the game-theoretic value against any opposition, and
strongly solved is being used for a game for which such a strategy has been determined for all legal positions.
The game-theoretic value of a game is the outcome when all participants play optimally.
We are talking about weakly solved here.
As I've pointed out in #7961, and subsequently:
(a) You're not talking about weakly solved; the definition you're quoting is flawed.
(b) You may be talking about that definition but you're proposing to completely ignore it.
(b) You may be talking about that definition but you're proposing to completely ignore it.
As you're completely ignoring his demonstrated modus operandi.
You're both badly programmed robots, not trolls, maybe. One as badly programmed as the other.
@7764
I estimate 3 cloud engines of 10^9 nodes/s as 3000 desktops of 500 $ each,
written off in 5 years hence 3000 * 500 $ = 1,500,000 $.
To weakly solve chess you need only find 1 move for black to draw against whatever reasonable moves white tries. Hence Sqrt (10^34) = 10^17 positions relevant to weakly solving Chess.
3 cloud engines calculate in 5 years:
10^9 nodes/s/engine * 3 engines * 3600 s/h * 24 h/d * 365.25 d/a * 5 a = 5 * 10^17 nodes.
"Hence Sqrt (10^34) = 10^17 positions relevant to weakly solving Chess." is a math error. there is no reason to take the square root.
That estimate would be in line with the empirical number for checkers (on a log scale).
@tygxc your 10^17 calculation relies on a pre existing algorithm that creates the best move for ____ on any given position. im surprised but also not surprised that no one else was able to point this out.
@tygxc your 10^17 calculation relies on a pre existing algorithm that creates the best move for ____ on any given position. im surprised but also not surprised that no one else was able to point this out.
Every flaw in his arguments has been pointed out. I for one no longer argue against him. He relies on an algorithm and yet it has been made very clear that there's no algorithm which is reliable. Brain dead.
tablebases solved the endgame
That's a good point. If there is the capability of exhaustively solving chess for 7 pieces - one would think that it's theoretically possible to solve it for 32 pieces, especially if those 32 pieces all start from the same position.
tablebases solved the endgame
That's a good point. If there is the capability of exhaustively solving chess for 7 pieces - one would think that it's theoretically possible to solve it for 32 pieces, especially if those 32 pieces all start from the same position.
That's been discussed extensively. It isn't possible and never will be at current or projected computing speeds. Billions of years .... at least.
When people make the mistake of listening to tygxc, they're going to get very confused. He's talking about a weak solution, which is supposed to be those moves which achieve the game state against any opposition. He calls it a strategy, which it isn't. Anyway, he proposes considerable pruning of the search tree based on removing lines which basically aren't good chess. How are they recognised? By grandmasters, apparently. Even accepting all this nonsense, his proposal of 5 years is not possible. It would take that long to design the project.
tablebases solved the endgame
That's a good point. If there is the capability of exhaustively solving chess for 7 pieces - one would think that it's theoretically possible to solve it for 32 pieces, especially if those 32 pieces all start from the same position.
It's merely ~100 times harder for each of the additional 25 pieces - lol. And the starting position doesn't help enough.
you guys are thinking in the wrong way, define what it means to solve chess? Is it just computing every possibility? I thought chess was just a game to have fun with???
i wonder how chess could be solved and what you mean by that
the "3 cloud engines 10^9 nodes/s during 5 years " is where you have got it wrong. there is no reliable source that puts chess even close to that level of complexity. lowest I can find is ~10^30