Friday, December 22, 2017

Entire human chess knowledge learned and surpassed by DeepMind's AlphaZero in four hours



This is good news and rather amazing.  I am sure they compared output to a look up library of all recorded games.

I always understood intuitively that the correct general strategy in chess was positional with agressive attack.  This blows the game open and allows positional exploitation to overwelm the opponent's knowledge and even talent.

Except this is likely beyond human calculation and must force you back on quick intuitive moves while searching for patterns you actually recognize. It really becomes computer against spirit body computer.  Yet it issomething that weall shouls practice in order to simply loosen up our game.  It may even be a lot of fun this way.

Speed chess must come close, but two losses in a row drive  sus back on to calculation.
.  We need to force ourselves to do this.


Entire human chess knowledge learned and surpassed by DeepMind's AlphaZero in four hours 


http://www.telegraph.co.uk/science/2017/12/06/entire-human-chess-knowledge-learned-surpassed-deepminds-alphazero/


Sarah Knapton, Science Editor 
Leon Watson 6 December 2017 • 6:55pm http://www.telegraph.co.uk/science/2017/12/06/entire-human-chess-knowledge-learned-surpassed-deepminds-alphazero/ 


Hundreds of years of chess knowledge was learned and then surpassed by Google DeepMind’s artificial intelligence algorithm in just four hours, it has emerged.

The astonishing programme AlphaZero quickly mastered the ancient game, before coming up with completely new strategies, which are now being analysed by grandmasters.

The algorithm is so extraordinary because it learns from scratch. It has only been programmed with the rules of chess and must work out how to win simply from playing multiple games against itself. 

When IBM’s supercomputer Deep Blue beat Gary Kasparov in 1997, it was because it had been programmed with the best moves. But AlphaZero has learned completely on its own. 


 

Gary Kasparov playing against Deep Blue Credit: Rex Features 

The English grandmaster Simon Williams, who runs the GingerGM site, said that the achievement was ‘one for the history books.’

“On the 6th of December, 2017, AlphaZero took over the chess world,” he said.

“AlphaZero and DeepMind then went on to dominate chess, eventually solving the game and finally enslaving the human race as pets.”

David Kramaley, who runs chess education site Chessable, added : “We now know who our new overlord is.

“The games AlphaZero played show it can calculate some incredibly creative positional bombs, the depth of which are far beyond anything humans or chess computers have come up with. 


“It will no doubt revolutionise the game, but think about how this could be applied outside chess. This algorithm could run cities, continents, universes.” 



 

Chinese Go player Ke Jie competes against Google's artificial intelligence (AI) program, AlphaGo Credit: Rex Features 


Jon Ludvig Hammer, the Norwegian grandmaster, described AlphaZero’s strategy as ‘insane attacking chess’ which was coupled with ‘profound’ positional play.

The DeepMind team eventually want to use the algorithm to solve big health problems. They believe that the programme could come up with cures for major illness in a matter of days or weeks, which would have taken humans hundreds of years to find. 

The company has already begun using AlphaZero to study protein folding and has promised it will soon publish new findings.

Misfolded proteins are responsible for many devastating diseases, including Alzheimer’s, Parkinson’s and cystic fibrosis.

The latest achievement was published online on the site arXiv.

No comments:

Post a Comment