ALPHA ZERO or CHESS GOD

What happened few hours ago, shall we call it a doomsday for Chess engines or something else, An artificially intelligent machine crushes the best chess playing entity in this universe($tock fish) like a black hole would crush anything else near to it, this Goddamn machine is impeccable like a human child adapting to learn to walk by himself, and has the capacity to learn chess by playing himself, and to some extent it also incorporates human intuition, well this is all of what I know about this ,, but the question remains, Will it be able to retain the reputation, will it start a new era of digital chess, Will it revolutionize the computer chess and give us new horizons of playing chess or is it gonna settle down by slightly improving the current chess super machines ??

https://chess24.com/en/read/news/deepmind-s-alphazero-crushes-chess

What do you understand in term of "human intuition"?

This is intuitive I reckon , learning from playing itself , the program remembers what he has done before, and take that knowledge onwards,

Now wondering if OTB chess gonna be dead, if/when the Al plays/solves chess 

Which one are you calling the chess god? Alpha zero was allocated significantly more computing power in the games.

It's interesting that it learned by itself, but it only learned a little. It's like a baby learning how to crawl, while Daddy (Stockfish) watches.

Humans have been playing chess for about 1500 years, and most likely will be playing chess for another 1500 years, regardless of what chess engines do.

A chess "God" ?

Does it have supernatural powers ?

If one of my chess pieces falls from the board and gets cracked, can AlphaZero fix it ?

Yes Google is incorporating that feature as well

lol. A good word for that program until its promoters calm down.😜

stockfish was evaluating 70 millions moves per second whereas alpha-zero 80k moves per second. This shows that by looking at smaller number of moves, alpha-zero evaluate current position better and predict next move better than stockfish

Hi, 

There are interesting implications from this match. Some people say the comparison wasn't fair because stockfish was cramped. IMO the point about this experiment is different :

1) Alpha zero is proven to be scalable & trainable for similar but different problems

2) Training was internal for problem with fixed rules

3) Training time is quite fast for practical purposes (but on expensive hardware for now)

Irrelevant. Chess is about winning games with fair play.

In that case, there should be no time control and the one who will be poor in position should spend all his life in finding the next move and don't play ;P that will be fair play as well

That's one way but not ideal. Another way is to submit their engine in a tournament with rules of fair-play, such as chess-com-computer-championship.

Maybe next year they will submit an engine. Would be fun to watch.

Even if they do not submit it, many engines are coming on this methodology.  You can check on GitHub, some guy has modified his convolutional neural network chess engine to reinforcement learning methodology. So, we can test this methodology in our labs as well.

AZ won't be able to run on commodity hardware though. Maybe they can interface with it but that would be an interesting experiment nonetheless

Yes, will be cool to watch the technology emerge. The day when they submit their program into a tournament with neutral directors, and rules of fair-play will be a big deal, and fun to watch.

I have not seen a definitive description of how Stockfish was or was not limited.  If it was playing without an opening book, then the demo seems pointless, and this would be a crucial thing to know.  If its clock were set to fixed time per move, that makes some sense for a demonstration, but it is still important to know about if we're going to assess what happened.

The full paper is linked in the following article: https://www.chess.com/news/view/google-s-alphazero-destroys-stockfish-in-100-game-match

The paper had the following information:
Each program was given 1 minute of thinking time per move. strongest skill level using 64 threads and a hash size of 1GB

Page 6 also gives statistics from 12 specific openings played; 100 games in each opening, 50 as white and 50 as black. There is no indication if Stockfish was using an opening book there either.