Google Deep Mind Rating

After reading how the Google program crushed Stockfish in 100 games without losing once, I wonder what its rating would be?!

Given the usual performance of Stockfish, that seems unlikely. 

Where did you read that? 

In the 100 game match it scored 64% which is equivalent to 100 rating points.

In the 1200 game match it scored ~61% which is, of course, a little less.

The exciting thing for AI is it trained itself to do this, and not only beat stockfish, but also beat the top shogi and go programs as well. To put it another way, stockfish can't even beat a beginner at shogi or go.

Oh wow, reading these other comments you guys have no idea what AlphaZero is do you? heh.

https://arxiv.org/pdf/1712.01815.pdf

Just like the seals who whacked bin Laden, AlphaZero did not get to enjoy the execution of Stockfish. Really though, It did not know it was playing chess.

Alpha zero is easily the best chess program ever created, whats even cooler is that it plays/learns chess more like a human. Another interesting fact is that it only looks at 80,000 positions per second compared to stockfishes 300,000,000+, it also teaches itself, no opening books no positions or point values, it literally taught itself to play by playing itself for 4 hours... that's it. this engine is insane, Magnus is 2800 and cant beat stockfish and stockfish is 3300 and cant beat Alpha Zero so if you can go on that alpha zero is probably 3600-3700 ish. It's not rated yet but that's probably its strength. and its only taught itself for 4 hours, just imagine if it has more time to train. this thing could be 4000+ plus strength eventually. We really can't know until its played more games but right now its probably 3600-3700 ish.

technically you don't even know if your playing "chess" the creator might not even have called it chess and chess have changed quite a bit so we don't actually know if it should be called chess. Also, your chess might be different from my chess. You could be doing something entirely different from me and have just learned to call it chess, Or the other way around.

 I'm both depressed, like the day when Kasparov lost the human v computer match, and impressed at what the AI people have achieved. 

These games will change everything, it made its own openings and gives up 2-3 pawns in like every game. We have so much to learn from this, you can see games here, https://chess24.com/en/watch/live-tournaments/alphazero-vs-stockfish/1/1/1

This is the super-engine among the engines

Wow, I've been wanting to see this match since Alpha Go beat Lee Sedol, Stockfish is like a Ferrari but AlphaZero is like an airplane....Doing something fundamentally different. Think of the wright brothers, They couldn't go faster than a modern day Ferrari, but with a long enough time-line Jets are the fastest things we got (outside of space travel). I don't think anyone can imagine what AlphaZero could do if it played chess for weeks or even months....5,000 Elo? 10,000? 15,000? Chess has been solved? Period. 

I would like to see how the results change when Stockfish is allowed to use its opening book. 

I would still expect AlphaZero to win, but there's no way that Stockfish would be destroyed the way it was. It really was an unfair fight.

 i don't think Alpha zero will be able to solve chess but i think we should stop making engines now because it's getting to the point where computers might be able to solve chess and imagine a computer free to download that has the answer to any chess question to ask. GMs often use engines to help prepare games with openings and eventually because of all the small positional gains GMs would make with the solution of chess to grind out a win to the endgame because of a minor flaw ten moves into a common opening it would come to the point where people would play pretty much the same game over and over again. 

All i do is hope i die before that happens which is very uncertain. The programmers of alpha go zero could be running it and having it play against itself since it was launched and chess could be solved tomorrow (although that's very unlikely) or alpha go zero's knowledge is minute compared to the solution and it takes 100 years for a new computer to solve chess. 

I completely agree. I could even see Stockfish winning a game or two being probable given 100 games with SF using its opening book.  I’ll throw out the complete guess of 12-1=87 AlphaZero or something like that. It would be interesting to gather more data on this by introducing Houdini and Komodo into the mix too. 

This can be just hype. Stockfish did use any opening book and probably end game table base, hardware power comparison is also misleading, Alpha zero needs to prove itself vs others with fair conditions. 

I estimate = or > than 4000.

I still believe the highest possible rating is ~3600. Stockfish's rating is lowered by the fact it wasn't allowed to use its opening database.