Linux Chess Supercomputer Overpowers Grandmaster

Capt Bubudiu writes "Deep Blue vs. Kasparov is something most readers will remember but when Deep Blue was retired by IBM, a Dubai company took over with Hydra. In a $150,000 6-game challenge in Wembley UK, the games got off to a humiliation for mankind as Michael Adams, the UK Grandmaster, was mauled in games one and three, drawing game two. Adams is ranked seventh in the world and what ordinary mortals call a 'Super Grandmaster'."

Comments

"we" won?

[moz25]

The interesting thing is that in a man vs. machine fight, the tech folks can say "we won" as they assembled the machine. Is it a humiliation or triumph for mankind that it can build a machine that can defeat itself? I think it would rather be a failure for humans if mortals can defeat highly optimized machines.

Re:"we" won?

[ottffssent]

The computers are not beating the humans with math. They are not relying on superior computational capability. The computers are winning with superior algorithms. Even a fairly shallow complete traversal of the search space is many orders of magnitude away from being possible, and a machine using this approach will be consistently beaten by even middling players. Computer chess has advanced primarily due to algorithmic optimizations. The evaluation functions that a modern chess engine uses are extremely well-tuned, and while a chess engine may be backed by an enormous pre-computed opening book, this too is dependent on algorithmic advances, because the book is calculated using algorithms as well, not a brute-force search. The two sets of algorithms are different, and the opening book can benefit from hugely greater computational resources, but ultimately the search algorithm is the limiting factor.

In short, the recent successes of machine chess are due to human enginuity, to the same sort of creative processes that humans themselves use to play chess. Technology, in the machine sense, is almost irrelevant (see Fritz's victories on a dinky 8P Xeon with a few gig of RAM) when compared to the advances in understanding of the game of chess.

Interestingly, even as the programmers are developing an ever-greater understanding of chess, chess players are developing an ever-greater understanding of both the game and the way in which computers play it, though people with much greater understanding of this than I tell me that the newest algorithms are playing a very human-like game, minimizing the effect of understanding 'computer chess' on the game.

Re:"we" won?

[BewireNomali]

Exactly. There is no suspense. Machines will be better at us at these things because that is their purpose. It's what they've been designed to do.

Given enough time, machines will be better than us at EVERYTHING.

To me, it isn't amazing that machines designed to excel at chess beat the best humans. It's amazing to me that humans can still beat and/or draw games with machines designed to be brute force unbeatable.

It's like Steven Hawking beating Shaq at basketball. It's amazing, be glad that you were around to witness it. Don't expect it to happen again.

Re:"we" won?

[MyLongNickName]

Given enough time, machines will be better than us at EVERYTHING.

Except at assigning purpose. This is one thing that cannot be expressed mathematically.

And you also do not understand chess. Chess is a drawn game by default. A "perfect" player could not beat you unless you made a mistake. There are ways to play the game that focus on minimizing risk as opposed to all out win.

Take a look at players like Petrosian (world champion in mid 60's as I recall). His style was python-like. He would see to suffocate you. Then, after tying you down, would systematically destroy you. Petrosian would be much better at playing supercomputers that Kasparov every was. His style could not be brute forced with today's supercomputers... too many plies to calculate... too many fruitless branches.

But, I do agree, in a few more decades humans will never be able to score a victory against the best computers.

But who cares? It is a linear game. I do not define my worth as a human being cased on linear criteria. Kinda gets back to the "purpose" thing.

Of course, if you are a Nihilist...

I dont get it...

[__aaxwdb6741]

I dont get it. Why is it so amazing that computers beat human beings in chess? Isnt chess all about logic and calculation? Arent computers all about logic and calculation?
If both are true, then how come it is so amazing that a computer beat a human being in chess?

Wouldnt it be more amazing if a human being beat a chess computer?

Jessie Owens Outpaced by Motorbike

[nagora]

Big deal.

TWW

scared

[mnemonic_]

Actually, the time to get scared is when a chess computer becomes sentient, creates an army of robots and enslaves the organic world. Our only hope then will be the chess grandmasters, academic athletes turned heroes of mankind.

Re:It is inevitable...

[MerlynEmrys67]

Not even humans play without a database.

If you look at high end chess - you play so many moves in a certain time (20 moves in 4 hours maybe) - so to make sure you have enough time when it is interesting, you start standard moves, and until you or your opponent go outside of an silently agreed on game, the moves are fast and furious (watch the first 10 moves when 2 grand masters play) - then they slow down as the players try to figure out when to deviate from the script, then about 12-13 moves in (in some cases) the plays start taking about 20+ minutes a turn.

So yes - openning databases are known quite deeply by the best players - a computer using a database is only fair.

Beating a supercomputer is easy..

[Thomas DM]

I bet I can beat every supercomputer on Earth.. If you just allow me to pull the plug ;)

Yeah, but...

[chriswaclawik]

In my mind, there will always be only one true grandmaster.

I'm still waiting for the day where a supercomputer can win a rap battle against a human...

A far better contest is compression.

[Baldrson]

Compression is a far better basis for intelligence competition than chess, the Turing test or even SAT verbal analogy tests.

Marcus Hutter's AIXI paper provides a proof that if an agent is a good model for human behavior, and the universe is computable, that the most intelligent program is the smallest program that losslessly compresses the set of observations of the universe.

I've formalized a prize competition based on this criterion as the C-Prize, modeled after the Methusela Mouse Prize. The big difference is that instead of lifespan the metric is intelligence. Here is the currently published C-Prize criteria:

Since all technology prize awards are geared toward solving crucial problems, the most crucial technology prize award of them all would be one that solves the rest of them:

The C-Prize -- A prize that solves the artificial intelligence problem.

The C-Prize award criterion is as follows:

Let anyone submit a program that produces, with no inputs, one of the major natural language corpora as output.

S = size of uncompressed corpus
P = size of program outputting the uncompressed corpus
R = S/P (the compression ratio).

Award monies in a manner similar to the M-Prize:

Previous record ratio: R0
New record ratio: R1=R0+X
Fund contains: $Z at noon GMT on day of new record
Winner receives: $Z * (X/(R0+X))

Compression program and decompression program are made open source.

Explanation A very severe meta-problem with artificial intelligence is the question of how one can define the quality of an artificial intelligence.

Fortunately there is an objective technique for ranking the quality of artificial intelligence:

Kolmogorov Complexity

Kolmogorov Complexity is a mathematically precise formulation of Ockham's Razor, which basically just says "Don't over-simplify or over-complicate things." More formally, the Kolmogorov Complexity of a given bit string is the minimum size of a Turing machine program required to output, with no inputs, the given bit string.

Any set of programs which purport to be the standards of artificial intelligence can be compared by simply comparing their Artificial Intelligence Quality. Their AIQs can be precisely measured as follows:

Take an arbitrarily large corpus of writings sampled from the world wide web. This corpus will establish the equivalent of an IQ test. Give the AIs the task of compressing this corpus into the smallest representation. This representation must be a program that, taking no outside inputs, produces the exact sample it compressed. The AIQ of an AI is simply the ratio of the size of the uncompressed writings to the size of the program that, when executed, produces the uncompressed writings.

In other words, the AIQ is the compression ratio achieved by the AI on the AIQ test.

The reason this works as an AI quality test is that compression requires predictive modeling. If you can predict what someone is going to say, you have modeled their mental processes and by inference have a superset of their mental faculties.

Mechanics The C-Prize is to be modeled after the Methusela Mouse Prize or M-Prize where people make pledges of money to the prize fund. If you would like to help with the set up and/or administration of this prize award similar to the M-Prize let me know by email.

The computer did it?

[glwtta]

Here's the thing, all the computer "did" was run electricity through a circuit - an electric heater does the same thing.

The hardware and software engineers who built and programmed that computer were the ones who achieved the victory - the computer has no understanding of chess, nor in fact any capacity of understanding.

Now if they designed a general purpose AI that then learned to play chess and trounced a great-grand master (or whatever they are called), that would be a computer defeating a human.

Hydra is not the ultimate chess entity

[Redshift]

Hydra performed very badly in the recent Freestyle Chess competition run by ChessBase - competitors were a variety of grandmasters and amateurs assisted by databases and computers. I other words, any form of cheating was acceptable, all that mattered were the moves on the board.

The two Hydra machines did not even make it into the final sixteen. Moreover, the eventual winners were a couple of amateurs using pretty ordinary PCs running over-the-counter chess programs. On the way to the title they beat a selection of computer- and supergrandmaster-assisted grandmasters.

On this evidence the "strongest chess entity on the planet" is a team consisting of a New Hampshire database administrator + a soccer coach + 3 ordinary PCs.

Links:

Hydra knocked out

Final result

Winners debriefing