TalkChess.com

corres wrote:It should have to perceive that the match circumstances during games between A0 and Stockfish is an unimportant thing. The essence is the team of DeepMind have proved they can make a powerful system based on neural networks which can solve such complicated tasks like winning go and chess games against very strong opponents.

and chess games against very strong <crippled> opponents...

that statement is more accurate...

pilgrimdan wrote:

corres wrote:It should have to perceive that the match circumstances during games between A0 and Stockfish is an unimportant thing. The essence is the team of DeepMind have proved they can make a powerful system based on neural networks which can solve such complicated tasks like winning go and chess games against very strong opponents.

and chess games against very strong <crippled> opponents...

that statement is more accurate...

Only if Stockfish was "crippled" to such a degree as to no longer be a very strong opponent. Was that the case?

pilgrimdan wrote:

corres wrote:It should have to perceive that the match circumstances during games between A0 and Stockfish is an unimportant thing. The essence is the team of DeepMind have proved they can make a powerful system based on neural networks which can solve such complicated tasks like winning go and chess games against very strong opponents.

and chess games against very strong <crippled> opponents...

that statement is more accurate...

I wonder why you have trouble accepting that the Stockfish that lost the 100-game match is an amazingly strong engine.

I have not seen this attitude among any of the engine developers (it is possible that I missed one).

Evert wrote:They don't give a rat's ass about what "the chess programming community" thinks, they're not trying to build the best and greatest chess playing program. They're trying to build a general AI that can learn to solve complex puzzles (in this case, play games) by re-inforced learning.

Exactly! All this talk about faking games and hiding stuff to impress us is nonsense, they don't even read this or another chess forum and they didn't really care for opening books or endgame tablebases.

The crux of the issue is that the development of Stockfish took several years in the making, and it's a state of the art engine that uses a recollection of the best classical methods of computing for chess and any other board game of its kind.

Stockfish was a great choice because it's open source, very strong, and pretty much every old idea is implemented in it, and either kept and optimized for a chess program, or ditched because it didn't work.

They came with their idea, and replicated that in 4 hours of training.

The result doesn't really matter, and even if their approach just scored 49-51 in favor of Stockfish, the achievement would have been as remarkable.

zullil wrote:As far as I know, Stockfish has only been defeated by engines that use alpha-beta (and, of course, by Lyudmil and Pablo).

Pablo has defeated Stockfish by closing the position and beating it (or rather, a lagging GUI) on the clock.

About Lyudmil's victory, the jury is still out on that one...

zullil wrote:

pilgrimdan wrote:

corres wrote:It should have to perceive that the match circumstances during games between A0 and Stockfish is an unimportant thing. The essence is the team of DeepMind have proved they can make a powerful system based on neural networks which can solve such complicated tasks like winning go and chess games against very strong opponents.

and chess games against very strong <crippled> opponents...

that statement is more accurate...

Only if Stockfish was "crippled" to such a degree as to no longer be a very strong opponent. Was that the case?

i really don't care that Stockfish loss in such a maner...

what i care about is why do it with non-optimal conditions...

can somebody explain that...

syzygy wrote:

pilgrimdan wrote:

corres wrote:It should have to perceive that the match circumstances during games between A0 and Stockfish is an unimportant thing. The essence is the team of DeepMind have proved they can make a powerful system based on neural networks which can solve such complicated tasks like winning go and chess games against very strong opponents.

and chess games against very strong <crippled> opponents...

that statement is more accurate...

I wonder why you have trouble accepting that the Stockfish that lost the 100-game match is an amazingly strong engine.

I have not seen this attitude among any of the engine developers (it is possible that I missed one).

I do not have trouble accepting that Stockfish lost the 100 game match...

the 'trouble' I am having is why do it under non-optimal conditions...

can you explain that...

pilgrimdan wrote:

syzygy wrote:

pilgrimdan wrote:

corres wrote:It should have to perceive that the match circumstances during games between A0 and Stockfish is an unimportant thing. The essence is the team of DeepMind have proved they can make a powerful system based on neural networks which can solve such complicated tasks like winning go and chess games against very strong opponents.

and chess games against very strong <crippled> opponents...

that statement is more accurate...

I wonder why you have trouble accepting that the Stockfish that lost the 100-game match is an amazingly strong engine.

I have not seen this attitude among any of the engine developers (it is possible that I missed one).

I do not have trouble accepting that Stockfish lost the 100 game match...

the 'trouble' I am having is why do it under non-optimal conditions...

can you explain that...

No, I cannot explain why that is such a big issue for so many people when it is really just small potatoes.

I liked this post by Ed, which explained his earlier "resistance" to the paper very well:
http://talkchess.com/forum/viewtopic.ph ... 634#745634
He is not troubled by minor gripes with the test conditions that only people who have spent many many hours with chess engines could possibly have (a category of people that likely lack representation in the Deep Mind workforce... to answer your real question). But he has trouble believing that it was possible to rival in 4 hours of computation time what mankind took 50 years to do (but it's actually closer to 70 years).

pilgrimdan wrote:

zullil wrote:

pilgrimdan wrote:

corres wrote:It should have to perceive that the match circumstances during games between A0 and Stockfish is an unimportant thing. The essence is the team of DeepMind have proved they can make a powerful system based on neural networks which can solve such complicated tasks like winning go and chess games against very strong opponents.

and chess games against very strong <crippled> opponents...

that statement is more accurate...

Only if Stockfish was "crippled" to such a degree as to no longer be a very strong opponent. Was that the case?

i really don't care that Stockfish loss in such a maner...

what i care about is why do it with non-optimal conditions...

can somebody explain that...

There was nothing wrong with the conditions.
No book?
Neither engine had a book.
Core count?
The technologies are not commensurate, and giving SF 1000 cores would not add much Elo due to SMP scaling. Set the cores to anything you like and run the experiment. The count is clearly arbitrary.
RAM for hash?
They could have spent hours tuning SF, but the choice they made was a good one. At most it cost a couple Elo and would not have made any significant difference in the outcomes.

pilgrimdan wrote:

corres wrote:It should have to perceive that the match circumstances during games between A0 and Stockfish is an unimportant thing. The essence is the team of DeepMind have proved they can make a powerful system based on neural networks which can solve such complicated tasks like winning go and chess games against very strong opponents.

and chess games against very strong <crippled> opponents...

that statement is more accurate...

people are focussed on the wrong metric - it's not the settings, it's not the Sf version used, it's not the time control, -- just look at the games and the moves played by A0 - some of the move played appeared to come from another planet.., consider what they also did with AGo and it's impressive and its believable, it was clearly unexpected and apparently for many people , not believable - something most have been drastically wrong to the point where SF 8 was crippled - why use the word crippled , because the results were over overwhelmingly in favor AO such that SF had to be crippled... it was not done the way many of us would have to like seen it be done , but it is what it is ..I also understand why people feel the way ...it's called emotion. SF is open source , everybody gets it for free, hundreds of people have volunteered either time and money ( time, computers and electricity are not free) and it was and still is the strongest free open source engine available, SF brings a lot of positives to the chess community and people appreciate that it so much such that it has become the favorite son of many chess enthusiasts and rightfully so and then big bad google comes with big bad money, hires a bunch of really big bad smart people, spends a few million on big bad hardware and then blows away the white knight SF , and people get upset ...

for me personally, I want to see some AO games played the way we would want them to see them play - but not because I think it would then be even and SF would have its revenge - but purely from standpoint that it would think it would be interesting and fun to watch...anyway I feel your pain too as the keeper of SF-McB - lets just hope there are some more games we can get to watch and analyze..