Humankind Makes Last Stand Against Machine

MrZeebo writes "According to this Financial Times story, Garry Kasparov has begun another match against a computer chess program on Sunday, this time playing against the Israeli-developed Deep Junior. Kasparov is the highest-rated chess player of all time, and lost to Deep Blue in 1997. According to the article, Deep Junior, despite evaluating less moves per minute than Deep Blue, is considered to be a superior chess player. The match will span 6 games, the last one being February 7th." Kasparov has won the first game.

Comments

chess...

[BarrettAnderson]

i can't remember the last time i ever won a game of chess...

Re:chess...

[Anonymous Coward]

i, for one, welcome our new machine overlords. i would like to remind them that i have always been a big supporter of computers, spending much of my time and hard-earned salary (and allowance, previously to that) on upgrading my computers. i look forward to years of faithful service to them.

sincerely,
AC

Re:chess...

[Anonymous Coward]

holy gosh.... it's impossible for me to get ahead in the karma world... there's always someone who has me on their enemy list (or them on mine)... how is it overrated at +2 funny???? use common sense when you mod, just because i don't like you doesn't mean you have to mod every single comment i make down. -Barrett

Re:chess...

[CmdrTaco (1) Comment]

Heh. Anyway, credit where credit is due. This story was also submitted two days ago by Digital_Story. And it was an interesting one, at that.

Re:chess...

[rhfrommn]

Play a game against me, you'll win for sure!

Re:chess...

[Anonymous Coward]

i tried that... i got the crap beaten out of me... ever since i've been trying to get thomas neusswhatever you spell it to make it easier, i doubt he will though

again?

[tx_mgm]

how many games does he get to play until this whole thing is decided?? geez...

How do they tell?

[spudwiser]

How can they really tell which computer plays better chess? I think they should put Deep Blue vs. Deep Junior. Start having robot chess championships, which team can develope better chess software. Two computers playing chess... would it take an hour, a microsecond, or until the end of time?

Re:How do they tell?

[TheSam]

Oooo, two computers... how long. Great question...

Re:How do they tell?

[cbensinger]

I'm not an avid chess person and perhaps my memory isn't what it once was; but I seem to recall reading somewhere that they *do* have computer chess championships of some kind. However, that being said I do tend to agree with a comment further down in that it's really which computer has the better set of programmers more than anything....

Re:How do they tell?

[sparkleytone]

Deep Blue was taken apart shortly after its matches with Kasparov. Interesting, being that Kasparov accuses that Deep Blue was given hints and help by humans. Therefore there can be no match of Junior vs. Blue. GG Kasparov

Re:How do they tell?

[Anonymous Coward]

They already do that. Deep Junior is the current reigning World Computer Chess Champion and that's why its playing Kasparvo now.

Deep Junior was put in retirement after it beat Kasparov a few years ago, so we may never know how strong it is against other machines.

Re:How do they tell?

[herrd0kt0r]

they do exactly that: they put computers against each other. and it typically doesn't take an hour, a microsecond, or the end of time. they usually abide by the same rules governing FIDE world championships. and yes, these tournaments typically result in the creation of better chess software.

look at the development of fritz, and deep junior, for example. or hell, why not try looking something up on google? it can't be that difficult, can it?

deep blue was dismantled after its rematch with kasparov. deep junior has been winning all the computer chess tournaments for the past three years.

Re:How do they tell?

[Daleks]

How can they really tell which computer plays better chess?

Read the second to last paragraph of this. Or just read "Deep Junior is a three-time world champion and won the last official world chess championship for computers in July".

Re:How do they tell?

[Beebos]

True story.

During a long plane flight, my brother-in-law and I pitted my Palm III vs. his Pocket PC in a game of chess.

His Pocket PC was clearly winning when my Palm III crashed, something it rarely does.

Just goes to show that technology isn't above having a temper tantrum and kicking the chess board over.

-

Re:How do they tell?

[kdgibson]

Maybe air it on battlebots?

Re:How do they tell?

[Anonymous Coward]

My Palm IIIxe started crashing since I took it on an airplane (over-atlantic flight).
Could some sort of radiation affect them somehow?

Re:How do they tell?

[WIAKywbfatw]

There is a story I remember reading in a computer magazine once (about 10 years ago) that seemed to me to be at best anecdotal but more likely urban myth. Anyhow, it was in a respected publication, and it wasn't the April issue, so I just filed it away in my brain in the "stranger things have happened" category.

According to the story, a chess computer that was programmed to win at all costs realised that it's human opponent was moves away from beating it. To avoid defeat, which was its overriding objective, it electrified the chess board and electrified its opponent when he made his next move.

Like I said, it sounded like urban myth to me (and pre-WWW I had no real way of exploring the myth further) but perhaps someone out there knows better.

Re:How do they tell?

[Anonymous Coward]

Think about it this way: would it really be that interesting to see one software program pitted against another software program? And how long would that match last, maybe 10 seconds? (y'know, all these fancy chess computers can calculate zillions of moves per second).

I think that the very essence of these chess matches is that the competitors *aren't* two computers, or two people, but human-vs-computer, because the computers are getting better at challening (and often beating) the human mind.

I personally wouldn't pay any attention to the results of Computer A vs. Computer B.

Re:How do they tell?

[anon*127.0.0.1]

Thats a switch... accusations that humans are giving the computer help. Usually it's the other way around. Virtually all the chess played online is speed matches, two or three or five minutes for the whole game, precisely because everyone is convinced that their opponent is running a chess program in a different window.

Re:How do they tell?

[domninus.DDR]

I know for a fact that they hold go program tournaments. If you google for aya, youll find a go program that competes in those and comes with some sample games

Re:How do they tell?

[BCoates]

didja get it x-rayed at security?

Re:How do they tell?

[anon*127.0.0.1]

Sounds like a bad science fiction story to me.

Re:How do they tell?

[Anonymous Coward]

Oh oh oh. I know the answer to this one!

Q: ... if it's a mail plane?
A: Just look between the landing gear!

Errrrr. Wrong question you say? n/m :)

Re:How do they tell?

[buswolley]

a perfect example of a +5 comment that is idiocy. Deep Junior is a champion because it plays other chess computers in tournements. This was indicated in the article, besides. Please mod the parent down to its appropiate level.

Re:How do they tell?

[Anonymous Coward]

i would like to see a computer vs computer match that also factors in time. i know they're already timed, but what i mean is a time limit that'll really work in the game as a crucial factor.

something like a 1 second timer for either side. that'll really make things interesting.

Re:How do they tell?

[cgenman]

Not to be a spoilsport, but this is utterly impossible. For a computer to intentionally do harm to a human being through a chessboard, the computer would have to either be programmed with the knowledge that human beings are subvertable by electrocution via the output line (and therefore via the chessboard), or have inferred it from a deeper understanding of physiology. The chess program would have to be extensively meta programmed with thinking routines and structured information about the outside world, as "win at all costs" is a statement of intent, and we have not quite moved beyond where stated intent can only be a simplification of the programmer's desires when structuring routines. This, as I have said, is changing, but is very doubtful that any machine from 1993 can said to harbor a real "intent," and the self-coding capabilities to carry out that intent. The chessboard would also have to be wired in such a way as to have access to a dangerous degree of alternating current. As basically all computers and computer ports run on DC, and DC is harmless, they would have to wire a board directly to an AC power supply, and both moniter and control the flow of power by DC regulators connected the CPU. The person at the table would have to complete a circuit between some electrified part of the board and another or be sufficiently grounded while sitting at the chair, or power transmission would fail. The chess pieces would have to be entirely metal to facilitate this transmission.

For that matter, they would have to connect the computer to a physical chessboard instead of just displaying one on the screen, or (more likely) having an IO person type in the human moves and moving the computer's pieces on the board. Commercial machines that can move / react to moves with a chessboard as IO, and with questionable AI, have been available since the mid-eighties. However, they are quite limited, hardly available, and physically incapable of electrocuting someone.

Stranger things have not happened. Things that had been previously believed to be impossible through some misreading of logic have eventually come true, given time... Machines have advanced to the point where they now can play chess, a once "impossible" feat, but it was truly impossible that Wolfgang von Kempelen's Turk could play a meaningful game of chess in the 18th century. Anything is possible given enough time, but what you describe is impossible without both technology greatly in advance of what we have available today and an almost homicidal recklessness spanning far beyond accidental negligence on the part of the designers.

As you describe it, this is truly impossible.

-C

Re:How do they tell?

[John Courtland]

Dude, DC makes your muscles contract while AC will end up pushing you away. Wrap you hand around a bare wire of 100V DC and 100V AC and tell me which you can't let go of cause the electricity if forcing you to hang on to it.

Re:How do they tell?

[inode_buddha]

I've actually done this a few times with GNUchess. Results are usually a draw. The gameplay is interesting though; approximately 95% of the moves take seconds, if that. The time required to complete any one game seems to increase exponentially, depending on the initial play level and how far along the game has progressed already. Actually reaching a draw can take hours, even with 2 GHz SMP boxes and large RAM.

Re:How do they tell?

[Old Wolf]

Not really, the fact that it was +5 reflects that the moderators also didn't know anything, so the question provoked responses which could fill in the knowledge gap for many people. Instead of complaining, go and mod up the replies which give the correct answer.

Re:How do they tell?

[Anonymous Coward]

OK, I tried the DC and now I'm stuck. What do I do now?

Re:How do they tell?

[cgenman]

I will do that if you can find me a com port that puts out 100V DC. Now, I've never taken apart a Cray, but I think that is unlikely.

I have done the latter with 115v AC... Three times. Stupid cheap weedwhackers. I was never actively pushed away as you describe, I only managed to force my fingers open after several seconds of trying.

We also haven't mentioned Amperage yet, or actual wattage, but as in the normal course of computer operations the DC supplies are powered off of AC imputs, the AC supply would still have significantly higher wattage backing it.

I suppose if our homicidally negligent designers gave the system a DC-based UPS *and* gave the computer regulation over the line *and* didn't include any fuses anywhere between the UPS and the motherboard and the motherboard and the com port and the com port and the pieces on the chessboard *and* made the power regulation channel extremely beefy and insulated enough to withold a few tenths of a second of a current of sufficient magnitude to be channeled to a table and to a human being and yet find insufficient release opportunities between components on the motherboard, the controller board, the port itself, the cabling, or the table it was played on *and* didn't think to tell the guy to wear gloves then yes, it is possible that a computer could really shock someone playing chess.

It is still impossible with current technology that it would "intend" to do that.

Re:How do they tell?

[Isofarro]

Oooo, two computers... how long. Great question...

Substitute the two computers with one piece of string - that simplifies the problem.

Re:How do they tell?

[drinkypoo]

Stranger things have not happened.

I am fucking rolling.

Re:How do they tell?

[Anonymous Coward]

ahh, memories.
Many years ago (1989?) a couple of friends and I
pitted a Commodore 64 running "Othello" against
an 8MHz XT running whatever the Windows 2.0(?)
solitaire program was. The C64 won :-)

Re:How do they tell?

[Anonymous Coward]

There is a story I remember reading in a computer magazine once (about 10 years ago) that seemed to me to be at best anecdotal but more likely urban myth. Anyhow, it was in a respected publication, and it wasn't the April issue, so I just filed it away in my brain in the "stranger things have happened" category.
Umm, it was in Mad.It was on a spoof of computer gaming where they presented games that we are certain to see.

Re:How do they tell?

[shatfield]

The IBM guys changed Deep Blue's parameters half way through the competition, and it started playing differently that it did the last game. I understand that humans can do this too, but IIRC, they stated before the matches started that they would not do this.

Also, they would not give Gasparov transcripts of the previous matches that Deep Blue had played.

How many chess players go into international championships without watching previous matches?

None.

Re:How do they tell?

[Ageless Stranger]

Read the book 'Behind Deep Blue'. Computer chess tournaments are done all the time. The strengh of a chess computer is determined exactly how the strengh of a human player is determined - by their play in tournaments.

Re:How do they tell?

[Anonymous Coward]

deep blue was dismantled after its rematch with kasparov.

they probably didn't want it taking over the world.

Re:How do they tell?

[jgerman]

As he has described it it is certainly an Urban Legend, however it is based on a true event. I don't have the time to dig it up now, but an enterprising karma whore could dig up the info through the Risks Forum.

Re:How do they tell?

[Anonymous Coward]

For some strange reason, I want to hook this guy up to 100V 12Amp DC current. Who wants to help?

Ok, seriously. DC is just as "harmless" as AC. Electric current through anyone's body is bad (unless it is a defribulator). It doesn't matter what direction it is going or if it stops and turns around... it is bad. IIRC it only takes half an amp to kill if it goes across your heart.

Here... try this... hook up a VannDegraf Generator and hold your hand on it a while. Make sure you wear rubber sole shoes and are not touching a metal floor. Then, after a few minutes, touch a lamp or something with the other hand. No, this is not smart. Yeah, it will smart. What you will feel is DC current. Think if that was magnified about 10 or 20 times. It's dangerous.

Re:How do they tell?

[jst666]

"The time required to complete any one game seems to increase exponentially, depending on the initial play level..."

Ummm. Perhaps the reason might be that usually wehn increasing the play level increases the time allowed for the computer to decide on best move.

Re:How do they tell?

[Anonymous Coward]

Actually pitting a chess computer to a chess computer is really only good for double-checking and testing the software for any obvious weaknesses. You never want to use one chess computer to teach another AI/learning chess computer because you'll just be reenforcing the bad habbits (weaknesses/flaws) that computer chess has.
My human chessmasters tend to poke and prod a computer's technique to find a weakness/hole. A human opponent would recognize this "testing of defenses" strategy and compensate. A giant chess computer is rarely, if ever setup to recognize that someone is probing it for weaknesses.
Anyways the matches are rarely fair, since it's usually dozens or more programmers, engineers and chessmasters on the side of the computer nudging it along before and after a match (acting like a giant team of super-coaches) versus one man.

Perhaps we should refer to this as. 'Many men and a machine versus man." instead of "man vs. machine". This might explain why the media and general population does not care about these tests. I know I'm only interested when the Man wins, since I know the deck is stacked against him. Cheer for the underdog I say:)

Re:How do they tell?

[dmforcier]

The IBM guys changed Deep Blue's parameters half way through the competition, and it started playing differently that it did the last game. I understand that humans can do this too, but IIRC, they stated before the matches started that they would not do this.

They did change the parameters. It is allowed under the rules, and common practice. If they ever stated "that they would not", please provide a citation.

Also, they would not give Gasparov transcripts of the previous matches that Deep Blue had played.

Accurate, but irrelevant since Deep Blue had played in no previous matches. DB's precursors *had* played in matches, and those transcripts are public record. If Kasparov didn't have them it was because he didn't look.

The thing that GK complains so much about is that they refused to provide the evaluation logs. Big difference! The logs describe the *reasons* that DB played or rejected certain moves. Kasparov is no more entitled to them than any human player is to have his opponent sit down after the match and descript why he played each move. No professional player would ever give *or ask for* such a thing.

Frankly, I think this is why IBM disassembled Deep Blue and will never sponsor or play in another such match. Kasparov poisoned the well. The payoff for IBM is in the public relations, and for GK to accuse them of cheating seriously detracts from the PR value. And there's no way to disprove the charges. (BTW, the match jury *was* given the logs, and after examing them they cleared the Deep Blue team of any wrongdoing.)

Re:How do they tell?

[Dan the MM]

I remember playing a couple of friends at a Risk-like game (can't remember the title) where I was about to be nuked out of the game, and his plus my nukes (as I had the chance to retaliate before dying) wouldn't be enough to cause a nuclear winter killing everyone and causing a stalemate. I instead opted to nuke the third (innocent) player to force him to launch his nukes as well to tie the game. They both exploded at the idea and said I wasn't playing fair, though the rules didn't say I couldn't do that. They were right of course, but whilst they spent 20 mins checking the rules and shouting/cursing me, I was laughing my arse off for the whole time. Just thought it'd be funny to do that to a computer opponent to see how it reacts. It wouldn't work for Chess, but I'm sure other games would work where the human player isn't trying to win just to see how the computer copes.

Re:How do they tell?

[Dan the MM]

Even better - Deep Blue vs. Deep Junior playing Chess using Robot Wars robots to destroy the taken pieces. If nothing else, the ratings for the games would soar.

programmers

[ak3ldama]

how come they don't award the programmers of these chess programmers they take into account such brilliant ideas that they deserve something?!?

Re:programmers

[ak3ldama]

sorry about that ... "these chess programs
they" ... a little typo on my part ... please accept my sincerest apologies

NO

[teetam]

I have posted this before, but was unfortunately modded down, so bere is go again.

This is not a match between man and machine. It is a match between humans - the human chess player vs the human software programmer. Please keep that in perspective.

Just because my desk calculator performs multiplications faster than me, doesn't mean that it is better at mathematics than I am.

Re:NO

[fafalone]

I strongly doubt the programmer could beat Kasparov himself no matter how much time he took.

Re:NO

[martyn s]

Given enough time, the programmer could just follow his own code and win that way. I think for someone just following code, more time would be much more valuable than for someone like kasparov. I think for a human (using a brain, not following code) there is a certain amount of time that anything beyond that provides no additional benefit. In other words, taking 5 days instead of 4 days for a human offers no additional advantage. It does for a computer.

Re:NO

[Anonymous Coward]

After being told that the Earth is not the center of the Universe, and that Man is just some brand of monkey, you have to hold on to something, because it'd be too depressing to learn that human intellect is crap?

Re:NO

[MoFoYa]

Is it human chess player vs. human programmer, or is it human chess player vs. programming team?

obviously in the 2nd case the match would be entirely unfair unless you consider that the real question is:

CAN a machine be programmed to "outsmart" a human every time. One programmer or many - doesn't matter. The point is, can out technology have grown to the point that we can teach it to be better than us as well as faster.

Re:NO

[Anonymous Coward]

I think it was modded down because you were incorrect.

Agreed, a simple calculator is not very good for many problems in mathematics, but it can only help. If you are only solving addition problems, you could beat just about anyone without one.

If you do indeed consider this a match between humans, than I think you have to concede it is hardly a fair one. To use yet another metaphor, perhaps this is liking showing up at a boxing match with an AK-47? The boxer who beats you with his fists is indeed a great one.

Re:NO

[Anonymous Coward]

Oh you poor wanker, you've been modded down.

Let me get out my box of tissues so I can cry.

Re:NO

[ronaldcromwell]

This isn't a match between Garry Kasparov and the programmers of Deep Jr.

I'd be willing to wager that Garry could easily beat any of the programmers of Deep Jr.

Plus, comparing Kasparov and Deep Jr is like comparing apples and oranges. Kasparov doesn't win by calculating every possible move each turn. There's something else going on inside his head, something that the computer will never be able to emulate. Call it instinct, natural skill, or a gut feeling, but when you break it down, they play the game in two very different ways.

Re:NO

[Suppafly]

If you think that, its obvious that you have no concept of how chess playing ai works.

Re:NO

[Daleks]

Just because my desk calculator performs multiplications faster than me, doesn't mean that it is better at mathematics than I am.

Yes, but it also means that it's vastly better (speed, accuracy) at addition than you are. Addition is a well-defined operation on a well-defined domain. Playing Chess is far from the purely mechanical and deterministic process of addition. You could argue that a search based solution to playing Chess (mapping all board configurations and moves) is deterministic, but this isn't what Deep Junior is doing. In fact, Deep Junior handles far fewer possible moves per second than the famed Deep Blue.

Re:NO

[dstone]

This is not a match between man and machine. It is a match between humans - the human chess player vs the human software programmer.

I disagree with you, since I could apply your reasoning and conclude that this is NOT Kasparov competing either. It is Kasparov's school teachers, nutritionists, chess instructors, fellow chess players, parents, programmers of software that Kasparov uses to train with, and authors of chess books that he no doubt assimilates knowledge from.

My point is that computer algorithms aren't the only thing shaped by the contributions and knowledge of others.

Both Kasparov and Deep Junior are "black boxes" with a recognized I/O protocol for playing chess. One box is made of meat and one is made of hardware/software. Neither box is created itself without huge amounts of guidance, programming, critiquing, iterative refinements, constant tweaking of strategies, etc.

Re:NO

[martyn s]

Okay, maybe. Tell me what you mean. I figured that if you gave the program more time it could come up with a better move.

Re:NO

[Suppafly]

no, the programmer can't just follow his own code and win that way.

Re:NO

[martyn s]

Of course it's not very feasible, but I was just talking in theory. Using enough pen and paper or some beads or whatever he needs, anyone can follow any code he wants, including chess code. Why can't the programmer just follow his own code and win that way, given enough time?

Re:NO

[Anonymous Coward]

Just because my desk calculator performs multiplications faster than me, doesn't mean that it is better at mathematics than I am.

Ok. Let's step out of the 70's for a second... What about this program? Is it better than you? (It's defintely better than me!) Anyway, there comes a point when the machine knows as much as any human does. At that point it's off to the races to see who can turn the crank the fastest. Eventually a computer will "solve" chess, and we will move on to more complicated problems like "What is the air-speed velocity of an unladen swallow"?

Re:NO

[use_compress]

Humans are incapable of executing the algorithms that the software teams programs... The programmers are more analogous to the coaching staff than to the human player.

Re:NO

[OwlofCreamCheese]

hmm, then in future the computer could run the code that the human's brain is useing. then the human could run the code that the machine was running to run the human brain. doing this the human could run a copy of his own brain inside his own mind.

Re:NO

[Anonymous Coward]

not a chess geek, but isn't time was part of the game :-)

Re:NO

[Domini]

But it is faster in arithmetic.

Re:NO

[Anonymous Coward]

Completely agree! It's humans vs. humans. In this case chess playing human vs. hardware & software engineers. Machines cannot think (yet).

This kind of bad journalism (not just /.) leads some people to believe that computers really can think and they expect to talk to one. When the real AI revolution happens, they'll say didn't that happen years ago already ? :-)

Re:NO

[Scarblac]

Why can't the programmer just follow his own code and win that way, given enough time?

Because he'd probably make some errors among the billions of computations :-).

But seriously, as long as we're talking thought experiments, it should be equally possible to write down the states and connections of all Kasparov's neurons and trace the firings by hand. Given enough time, of course.

Re:NO

[ComaVN]

The rules of boxing are pretty strict as far as guns are concerned, so the jury would probably (posthumously) declare you the winner of the match. Assuming, of course, the guy with the gun doesn't point it at the jury.

Re:NO

[jgerman]

Wrong. The computer is on it's own (for the most part, there have been times when a program was re coded during a match) once it plays. It's unlikely that the programmers could beat Kasparov, but they're creation can. The software makes the decisions, not the programmers.

Re:NO

[jgerman]

Because given enough time is enough of a qualifier to ruin the theory. You're neglecting competency, in order to be compentent at something you need to be able to do it in a reasonable amount of time.

But yes in theory, and ignoring the potential for mistakes a human could follow any code he wants.

Re:NO

[(void*)]

And I disagree. With an ALGORITHM, it is possible to verify that the computer is truly executing THE ALGORITHM, no ifs and buts. With a human, he can tell you that he carrying out alpha-beta minimization, but it is impossible to be absolutely certain THIS IS THE ALGORITHM THE HUMAN IS ACTUALLY USING TO MAKE HIS DECISION.

It is possible to take the source code of the program apart and say which piece came from which programmer, expert or people. It can be HARD to do this with human. How much of Hardy was there in Ramanujan? How would you even begin to answer that question?

Re:NO

[iabervon]

This generation of chess programs are actually basing thier play not exclusively on the code, but also on a database of games played by various people, so it's not exclusively the programmer, but also involves what the program can learn from historical play (In fact, current programs can attempt to play like specific human players).

So the programmer is actually writing a program to learn how to play chess well from looking at good play, so the program can surpass the programmer not only in execution of the play but also in understanding of chess, since the programmer is not only not as good at applying the learning methods, but hasn't looked at all of those games.

Re:NO

[Anonymous Coward]

How much of Hardy was there in Ramanujan? How would you even begin to answer that question?

You don't need to answer it. As long as you agree that Hardy contributed to Ramanujan (there's a little bit of faith involved here, but really very little), then you agree that both human and computer use contributions from others.

To put it another way, the slab of meat named Ramanujan sitting there is a product of Hardy and others. At any given moment, is he using Hardy's knowledge? We can't say for sure. But it would be very obtuse to say that Ramanujan didn't contribute to Hardy... THUS, Hardy plays some small role in that 'competitor' in the chair. Likewise, Deep Junior would have a team of programmers responsible for creating it. At any given moment, can they agree whose code or concept or algorithm is being executed? Perhaps with much work. Or perhaps they don't care. It was a group effort. Group code reviews and revisions. If I optimize one line in your algorithm, is it mine? Who cares. It makes Deep Junior. -We- contribute to Deep Junior.

Re:NO

[xv4n]

I think the point here is that the machine has no consciousness about the moves it carries on, it just follows orders given by the programmers.

Re:NO

[Anonymous Coward]

Just because my desk calculator performs multiplications faster than me, doesn't mean that it is better at mathematics than I am.

Well, true, but that's a grave error in logic. It does mean that your calculator is better at multiplying (at least numbers within its range) than you are, but of course mathematics is a much larger field and multiplication isn't a primitive (it's not a "Complete" problem, in the computational sense).

With chess, these computers are better at analyzing the game state and responding than most humans (including their programmers). You'll note that (unlike with multiplication and mathematics) this directly translates to being better at chess. In fact, along with knowing the rules, it's all of chess.

Before someone steps in (assuming anyone even reads AC posts) to shout me down with some argument for ingenuity being a role: this is a rigid game with a finite (albeit large) number of outcomes/states. You can't "create" any new moves; there's only so many possible. Creativity (which is an awful misnomer, here) is a sham in chess. It feels like it's working because your opponent fails to analyze far enough ahead to see what you're really doing. Try to make these "creative" moves against a good human or computer player who won't fail to correctly analyze the situation; you'll get your ass handed to you.

Kasparov has won the first game.

[Anonymous Coward]

That is Amazing!
I can't even beat the easy setting on free apple version!

AC

Re: Kasparov has won the first game.

[m00dawg]

Well, if it's the Mac OS X version then it uses GNU Chess and is pretty good, suposedly. It rules over my playing abaility, and I don't know anyone that has been able to beat it...

Re: Kasparov has won the first game.

[MrResistor]

What is even more amazing is that I saw this last night on the evening news (local NBC affiliate). As if that isn't strange enough, it was part of the sports segment!

Re: Kasparov has won the first game.

[dildatron]

Yes, I too have played the GNU chess that comes with OS X. I hardly ever play chess, and I just know all the basics rules. Yet, I have never won a game against the computer. Even on the easy setting with a fair amount of hints. I only play it when I get really bored in class, and other than that I never play chess.

hrm....

[xao gypsie]

makes me feel like a moron. i struggled with the basic chess program that came with older versions of windows, and this guy is going against super programs and such......hrm, well hopefully i can say that at least i have more digits of pi memorized than he does. no? damn....

xao

Re:hrm....

[Anonymous Coward]

3.1415926535897932384626433
Thats as much as I got.

Everyone memorizes pi. to make yourself unique, try like e or 1/pi. That will teach those effing chess players.

Re:hrm....

[slide-rule]

I used to get crushed by level 1 on most chest programs for years... until sometime a couple years ago I actually (more for general/novelty interest) picked up a "beginner"-style book on chess openings. You can't believe how much it helps to know how to properly open the first four or five moves in a game. You can bet the chess programs probably do, and if you don't know the most proper responses, then it seems to me that you've basically thrown away most any chance you have at the whole game, since the program is way ahead with a strong board position. (Turns out, there are reasons why there are so many thick tomes on openings. ;-)

Re:hrm....

[Anonymous Coward]

You don't even have to memorize e, just do the series addition. And if you memorize pi, you can easily calculate 1/pi.

I know 3.141592653589793238462643383279502884197169399375 105

how can kasparov win?

[herrd0kt0r]

deep junior can calculate 3-4 million moves per second! how can garry possibly win?

from wired: "Kasparov said he can calculate the potential of about 3 moves per second at best, 'but they are the best moves.'"

Re:how can kasparov win?

[Gyan]

Possibly because Kasparov doesn't play soley on raw intellect. Gut instinct and that hint of irrationality creeps in. The computer can't take that into account when anticipating Kasparov's possible countermoves.

Re:how can kasparov win?

[Enigma2175]

Possibly because Kasparov doesn't play soley on raw intellect. Gut instinct and that hint of irrationality creeps in. The computer can't take that into account when anticipating Kasparov's possible countermoves.

Yes, but what the computer can do is calculate the probability of all of Kasparov's moves, then only explore the options that are most probable. Although the computer cannot correctly predict the exact move Kasparov is going to make, it can probably narrow it down to 2 or 3 likely moves. It can pretty much ignore all the other moves, because Kasparov is not going to make a bad play. Kasparov's move is either going to be the best move for the situation, or at least the second or third best move. Although the 'best move' for any given chess scenario is debatable, the way the computer plays is to quantify the relative strengths of different positions and try to get into the strongest position. It is something that they are quite good at and I only expect them to get better as they get more raw power coutesy of Moore's Law.

Re:how can kasparov win?

[Gyan]

Although the computer cannot correctly predict the exact move Kasparov is going to make, it can probably narrow it down to 2 or 3 likely moves.

How is 'likely' defined to the computer ? How does the computer figure out likelihood when taking into account a human's gut instinct ?

Re:how can kasparov win?

[TruthSeeker]

The way computers play and the way human brain works are two completly different things and follow two opposite pathes.

Whereas the computer has to go through calculating moves explicitly, millions of move, the human brain has a kind of "fast pattern recognition" feature that allows him to _know_ what the right thing to do is, in less than one second.

Compare this to face recognition : it took _years_ to create a robot that is able to recognize a few faces, whereas your brain identifies any of your friends in milliseconds, in almost every possible situation.

Re:how can kasparov win?

[Scarblac]

... Kasparov doesn't play solely on raw intellect. Gut instinct and that hint of irrationality creeps in.

Actually, it's all intellect, something the computer doesn't have as it can only do stupid calculations. It's rationality that creeps in. The computer has to calculate all kinds of moves, but Kasparov doesn't even have to consider them because he knows they don't make sense in this position.

Human grandmasters go heavily on pattern recognition. They have on the order of 100,000 types of positions with typical plans memorized, as well as many many tactical patterns. Given a position, they know what both sides should be trying to do. Computers can't do pattern recognition well, so they can't use that method.

Re:how can kasparov win?

[Gyan]

It's rationality that creeps in.

Hardly. Knowing your opponent is innately an exercise in psychology and fundamental similarity of the human experience and thought process.

Kasparov can't predict how the computer will move, since then he needs to know how the computer thinks and the input the computer has. He doesn't really have a full idea of either. He has to make judgements based on incomplete information. That's where gut instinct comes in.

Re:how can kasparov win?

[zabieru]

They are the best moves because Kasparov can immediately discard 95% of the possible moves and then go on to evaluate only those moves and sequences of moves that are valuable. So once you start talking about three or four-move sequences, Kasparov is spending CPU cycles and memory only on the most promising 0.01%, while his opponent is applying (vastly greater) computing power to all possible moves. In other words, Kasparov has little RAM, and a slow CPU, but he has this amazing fuzzy-logic coprocessor which allows him to apply those resources much more efficiently.

Re:how can kasparov win?

[Anonymous Coward]

Possibly because Kasparov doesn't play soley on raw intellect. Gut instinct and that hint of irrationality creeps in. The computer can't take that into account when anticipating Kasparov's possible countermoves.

Well, I don't know if this is normal but when I used to play chess regularly I used to be able to look at the board and sort of "see" the paths that every piece could take, simultaneously. I'd look at the pattern, then explore the sections of the board where the pattern looked promising. It really helped to find the moves I wanted quickly.

I would imagine that really advanced chess players do something like that either consciously or unconsciously except on a larger scale.

As for the hint of irrationality... I think that would only work against simpler computers that just do a depth search for moves. Then your "irrational" move could get you into an area of the search tree which the computer hasn't explored very well, and you could benefit from that. However, against the sort of algorithms and processing power Deep Junior has, being irrational is just that: irrational. Any move Kasparov makes has to be deeply rational in order for him to win.

Remember also that these games can last a long time. While I'm sure Deep Junior will do better the more time you give him, seven hours is still a lot of processing time for a computer that can evaluate millions of moves per second.

Re:how can kasparov win?

[JimDabell]

The computer has to calculate all kinds of moves, but Kasparov doesn't even have to consider them because he knows they don't make sense in this position.

Surely that means that Kasparov has evaluated them? Otherwise there would be no way of knowing that they don't make sense. Discarding very low-quality moves may be something he does unconciously, but he's still evaluating them.

Re:how can kasparov win?

[Enigma2175]

How is 'likely' defined to the computer ? How does the computer figure out likelihood when taking into account a human's gut instinct ?

'Likely' refers to the best move for a given situation. While Kasparov will occasionally make an unanticipated move, generally he makes the best move for a given situation. I think he rarely plays by 'gut instinct'.

Re:how can kasparov win?

[Fizzol]

Kasparov, indeed any world class chess player, eliminates huge swaths of moves based on simple pattern recognition. He'll pick out a small number of candidate moves based on positional, and tactical considerations and calculate those, sometimes more deeply that the computer can. Intuition comes into play too. Kasparov can see by the general qualities of the position that a king side attack is called for or perhaps a push to gain space on the queen side or something else. He dosen't calculate that general strategy but he'll definately take it into account.

Re:how can kasparov win?

[JimDabell]

Kasparov, indeed any world class chess player, eliminates huge swaths of moves based on simple pattern recognition.

I can't tell whether you are trying to agree or disagree with me, so I'll just point out that this is evaluation.

He dosen't calculate that general strategy but he'll definately take it into account.

Of course he calculates it. Otherwise, where does it come from? Somebody whispers it into his ear? It magically pops into his head? Every decision we make is a calculation. A person might not be conciously aware of every factor that leads to the result, but it's still a calculation.

Re:how can kasparov win?

[Gyan]

Ok..bear me out on this

generally he makes the best move for a given situation.

Given that there are 'm' number of possible moves from a given state. After choosing one of those 'm' moves, the computer can make any one of it's 'n' responses. So that's mn possible scenarios. Take this ahead for every deeper level and you get mindboggling numbers. Now, Kasparov has played against people (i.e. humans) all his life. Humans can't think as quickly or as deeply as computers which are dumb machines with awesome calculation prowess and lack of any prejudices via training. They can be reset if those exist.

A human doesn't have those luxuries. When two humans play, neither of them computes all possible combinations to determine a 'best' move. They depend on prior demonstrated prejudices of their opponent, their own training, their limited computation to come up with the 'best' move. This is innately done.

Kasparov can't combat the full repetoire of possibilities that the computer can process. Kasparov can't look at past history and get some clues on how the computer might respond. The computer as a machine isn't bound by those constraints (atleast it doesn't have to be).
So Kasparov's best move is essentially a combination of reasoning/pattern recognition AND gut instinct.

Re:how can kasparov win?

[Anonymous Coward]

Aha! so that's the problem, we've been going about this all wrong: Instead of designing a better electronic brain, we should be designing an electronic gut.

Next year's Deep YACPC will have a mechanical stomach and a random number generator (for that Hint of Irrationality [tm]), and will mechanically fart in Kasparov's general direction.

[Yet Another Chess Playing Computer]

Re:how can kasparov win?

[Anonymous Coward]

the strangest thing is that Deep Blue could calculate between 200-300 million moves, whereas Deep Junio can only calculate 3 million.... makes the whole thing just seem like a publicity stunt to reassure people in the face of 'ambient computing' and other invasive practices that will eventually rule our world.

Re:how can kasparov win?

[Blimey85]

how can garry possibly win?

He might be able to trick the computer. If the computer is programmed to study his moves, and knows what move to make based on a move he makes, he might be able to make alternate moves, moves he would not normally make, to somehow trick the computer into making a fatal mistake move.

Obviously I know very little about how the computer he is playing is setup. I have an electronic chess game that was supposedly programmed by Kasparov himself... or had something to do with the programming of it and even though it has 32 levels, I can't get bast the very first "beginner/super easy/easy enough for babies/first timer/newbie" level. No matter what I try, the damn thing is always one step ahead of me. I do fine against human oponents but either my game is messed up and I'm always on the hardest level or I'm not as good as I thought I was.

Re:how can kasparov win?

[rhfrommn]

One important point about this. In any given chess position there may be very many moves, but only a very small fraction of them are worth anything at all. The rest are simply bad moves that hurt your position. Typically even in a very complex positon a quick (few seconds at most) glance at the position can eliminate all but a handful of moves. That is how masters can play simultaneous matches against dozens of opponents at once, they can eliminate all the bad moves from consideration and spend a short time thinking about the 2 or 3 decent choices left.

The computer doesn't need to consider the bad moves much at all. It is a pretty safe assumption that the opponent will not choose a terrible move. But if they are stupid enough to play one then the computer can calculate the new position as needed.

Ralph

Re:how can kasparov win?

[StupidHelpDeskGuy]

Tactics vs. Strategy. The beauty of chess is due to the fact that cold calculations are not the sole factor in determining the outcome of a game. There is a certain amount of intuition that a machine can not mimic. I don't claim to understand the sublties of this, but am intruiged by the fact that Mr Kasparov can actually beat this machine. It shows that there is more to the game than raw processing power. It takes an artful mind as well.

Re:how can kasparov win?

[Anonymous Coward]

Computers basically use a tree search to calculate the best move. This has a serious case of diminishing returns in terms of number of positions. Assuming a full width search, theres 38 legal moves per position roughly in a chess game. That means the number of positions to search is 38^X where X is the number of half-moves.

Of course, there's several improvements to a full width tree search (transposition tables and backwards pruning like alpha-beta search algorithms are probably the biggest) but the the problem is still the same. It gets exponentially harder for a computer to search deeper.

People, on the other hand, do no such thing. They eliminate moves immediately, at the root of the tree. The use forward pruning.

Several attempts have been made at forward pruning chess engines, but so far none of them have been very good. This is also the problem with Go engines. A branching factor of 38 is not good, but can you imagine the branching factor in a go game, on a 13x13 board? At the start of the game, theres well over 100 legal moves. One must throw out moves without searching deeply on them in order to see far enough into the game to play well, but there hasn't been an algorithm designed yet to remove the bad moves without accidentally pruning some good moves. As far as I know, there is one successful forward pruning engine, and it is for crazyhouse, a variant of chess. It's called sunsetter.

Re:how can kasparov win?

[Anonymous Coward]

Sometimes, it seems like no one is trying to understand each other here.

What Fizzol is trying to say, I think, is that Kasparov has a lookup table (of sorts) that he uses to determine what are possible good moves for any board state.
Admit it, if a program (the opening and end game of beginners is a good example) used a lookup table to determine its move, you wouldn't call that calculating, would you? At best, you might say, instead, that the move was pre-calculated...

Re:how can kasparov win?

[kavau]

deep junior can calculate 3-4 million moves per second! how can garry possibly win?

In a nutshell: even with 3-4 million moves per second, the computer cannot look ahead much more than, say, 12-15 moves. The branching factor is just too large, with about 20 possible moves to choose from; the number of positions grows exponentially with the number of moves.

Drawing from his vast experience and intuition, however, Kasparov can sense, in certain situations, how something he does now might affect the endgame 20 moves later. The computer has no chance to see this, because its horizon is limited to 15 moves. Kramnik made good use of this approach in two of his games against Deep Fritz. Unfortunately he got carried away in the later games. He could have beaten Fritz if he would have kept his calm.

Re:how can kasparov win?

[Anonymous Coward]

That's an interesting comment. I wonder whether anyone has investigated incoporating pattern recognition into Chess programs to recognise typical positions and strengths and weaknesses from these positions?

Re:how can kasparov win?

[Gyan]

I know it's pretty late. But on the off-chance you catch this reply.

I think he rarely plays by 'gut instinct'.

Read this

Kaparov Loses

Anything other than chess....

[very]

If this computer were "superior" than human at chess game only, we wouldn't have to worry for a Matrix/Terminator-esque future ahead of us.

Re:Anything other than chess....

[silvaran]

It's a step in the right direction... who's to say the next-generation computer will be able to write its own software? Or the generation after that can design new chips that are even better than its own? Then these new generation computers will certainly give the human race a run for its money. It's going to happen... likely not by 2020, or even 2100 for that matter, but eventually...

He should switch games...

[Anonymous Coward]

He should switch to Go. Even the greatest computers can't compare to an average player.

Go is far better suited to the way a human brain works - pattern recognition, neural networks and all that.

Of course, once a computer arrives that can beat us at Go, then it'll be time to rethink a lot of things :)

Re:He should switch games...

[bwt]

Perhaps this is just a consequence of the fact that computer scientists have studied chess substantially more than they have studied go.

I also don't understand why people think that because a computer program can play better than you means that you should stop playing. These games are deterministic and finite -- there is a mathematically perfect play whether or not somebody has calculated it. It really makes no difference to me as a chess player that a machine can trounce me any more than it does that Kasparov could trounce me.

Re:He should switch games...

[Forgotten]

It's an interesting question whether a chess grandmaster could switch to go, as you say, and be successful at it. No doubt some of the tools of thinking would be applicable, but many of the habits he'd bring would probably be bad ones. Likewise for a go master who switched to chess.

As far as "suited to the way a human brain works", they're both games of abstract logic. Neither is particularly suited to anything. More importantly it's a red herring - the human brain is flexible and adaptable well beyond the scale of these games. But chess may be better suited to current paradigms of programming (it is in many ways a more procedural game).

Re:He should switch games...

[Russellkhan]

" Perhaps this is just a consequence of the fact that computer scientists have studied chess substantially more than they have studied go."

It's very likely true that there has been less time in man hours spent developing Go playing programs than Chess playing, but there has been a very significant amount of time spent on the problem by some very intelligent people who are both good Go players and good programmers. So I would say that it is unlikely that this is the root of the difference. After all, Backgammon and Checkers have both also had significantly less time dedicated to developing programs that play and the programs out there play at championship level. Go is just a harder game to program. Its style of play doesn't lend itself well to linear lookahead or databases of board positions (or, in the case of backgammon, statistical prediction of dice) as the other games mentioned above do.

"I also don't understand why people think that because a computer program can play better than you means that you should stop playing. These games are deterministic and finite -- there is a mathematically perfect play whether or not somebody has calculated it. It really makes no difference to me as a chess player that a machine can trounce me any more than it does that Kasparov could trounce me."

Agreed. The games are still fun and still have something to teach me.

Re:He should switch games...

[Anonymous Coward]

No, I used to think this too, but play a bit of Go and it's immediately obvious that it's exponentially harder than chess for computers to play. Skilled Go players rely on "intuition" --- Go positions are much harder to score and decompose rationally than in chess. Computers may eventually get good at Go but not anytime soon.

Re:He should switch games...

[Anonymous Coward]

Go is far better suited to the way a human brain works - pattern recognition, neural networks and all that.

Complete nonsense. Humans don't beat computers at Go because our brains are somehow better suited for Go than chess. Rather, it's because computers are less suited for it (due to combinatorial explosion).

It's not even a subtle distinction.

Re:He should switch games...

[jonhuang]

From what I've read, it's not just that Go has so many more possibilities -- it's that even the best players are often stumped to say what makes a good move and what doesn't. Like other intuitive actions, this makes it difficult to program.

Re:He should switch games...

[Vryl]

A quite few of the chess guys I know play Go. Mainly for a change. And they are pretty good.

The guy that taught me Go is basically a drunk in Fremantle, but was once an IM in chess. I can't beat him at either game.

Re:He should switch games...

[Beliskner]

The one thing I wonder is when will a computer beat us at the game of general life - when will robots raise our children better than we can? Already Knowledge Management systems are collecting enough data about our behaviours, when will an intelligence come about that can apply them automatically?

computer + water = zap

[-strix-]

but make them play chess in a swimming pool and see who wins.

Re:computer + water = zap

[Anonymous Coward]

They would both loose.

Test this by taking a plugged-in appliance into the bath with you.

Re:computer + water = zap

[Anonymous Coward]

Yeah? Make them both play in a vacuum....then see who wins.

machine vs. machine?

[YodaToad]

So, when do we get to see Deep Junior vs. Deep Blue? It'd be kinda fun to see who developed the best chess machine :)

The only way to stop them...

[happypizzaguy]

is to Slashdot them! Anyone know Deep Junior's ip?

Re:The only way to stop them...

[ReverendRyan]

its 127.0.0.1, i think ;)

Re:The only way to stop them...

[Bullet-Dodger]

its 127.0.0.1, i think ;)

Ah, that's why it lost the first one. It was distracted by all the porn that seems to be on its harddrive.

Re:The only way to stop them...

[Anonymous Coward]

Hurry!
The ping is GREAT so far!

Chess? No, no, no!

[Anonymous Coward]

Chess is hardly a game of intelligence, but rather a game of experience. Computers with only thousands of times the capability of humans beat our mere collection of grey cells simply because we're more experienced. A game is a single event forever to be forgetten by a computer whereas for humans it is a learning experiences with energies to be drawn in the future.

Now if we really want to a game to see if computers are capable of ever besting us, I propose a game of "truth or dare". The only winner is the one least embarrassed. Once they can beat a human at this, I forfiet my humanity.

Re:Chess? No, no, no!

[Bicoid]

Now if we really want to a game to see if computers are capable of ever besting us, I propose a game of "truth or dare". The only winner is the one least embarrassed. Once they can beat a human at this, I forfiet my humanity.

Eh....depends on a computer. The old iMacs (the technicolored ones) seem pretty shameless. If you can make one of those get embarassed by a little truth or dare, you have to be DAMNED clever. Anyways, what're you gonna do? Dare it to install Microsoft XP as its OS? Ask it how many pornography jpegs are on its harddrive? Ask it if it's ever been pinged without a firewall?

What I really want to see is a computer that can read slashdot articles and post replies. And I'm not talking about easy-to-program "In Soviet Russia" replies, either. My experience shows that at least half of the anonymous cowards who post probably ARE that type of bot...

Hmm....

[Misch]

I saw the headline, and I thought it was an ad for the next rehash of the Terminator movie franchise.

Please not another IBM

[TheJesusCandle]

I hope they treat him fairly in this match. IBM didnt with their match, even though i didnt like the way Kasperov handeled himself either.... Lets face it, the human mind is a great computational machine, but somethings are better suited for computers. Thats why we make comuters. At some time, the design of hardware and software will be beyond anyone human minds comprehemption, were pretty much there now. Try coding in assembler for ia64. Yeah you can do it... But a finely tuned algorythm is gonna give you a run for your money

Re:Please not another IBM

[rufusdufus]

The human mind is a terrible computation machine. How many people can give you the sqrt(17) to 8 digits without pen and paper?

Human minds are great pattern matchers. How many computers can recognize the difference between a dog and a cat?

Re:Please not another IBM

[Daengbo]

I don't know... Why don't you tell me how to recognize the difference between a dog and a cat...
Incedently, I'm confused why Junior gave up his rook for a knight and no real position in move 17.

Re:Please not another IBM

[MacJedi]

Easy. The cat looks "cat-like" and the dog looks "dog-like." Whats the big deal? ;)

Re:Please not another IBM

[TheJesusCandle]

The human mind is a terrible computation machine. How many people can give you the sqrt(17) to 8 digits without pen and paper?

Human minds are great pattern matchers. How many computers can recognize the difference between a dog and a cat?

Computers are terrible pattern matchers because we havent taught them how to pattern match effectively, or taught them how to learn to pattern match.

Pattern matching is a difficult computational process, which the human mind has developed over millions of years. I guess that kinda makes the human mind a pretty powerful computational device, but only for specific applications.

Wait a minute

[dsb]

I thought we were still doing post game anaylsis of Super bowl advertisements?

Yes.

[pb]

It's easily possible to write a program that plays a game better than the programmer; in fact, this very thing happened early on in the history of computers that play games (in this case, checkers).

I guarantee you that Deep Blue and Deep Junior play chess better than their programmers, and for that matter, almost everyone on earth. That's why they get to play Kasparov.

Re:Yes.

[almeida]

I don't think they are better at chess. I think the computers are just better at the things that are useful in chess. They can analyze moves faster and remember more about their opponent's technique than their human creators. Given enough time and maybe a notebook to keep track of stuff, you could accomplish the same thing. The computer is using the same basic chess rules that everyone else uses. The difference here is the computer can apply the rules ridiculously fast.

Re:Yes.

[Ianworld]

hehe i met one of the creators of the computer that beat the top checkers player. For him it seemed like a minor achievement. He thought the real challenge would be in creating one of these chess computers. Well i guess here they are.

Re:Yes.

[Anonymous Coward]

Hence the term, 'better'.

Billy Bob the moron can do all the math I can, he just needs a notebook to write everything down and take notes.

I can do timestables in my head!

Re:Yes.

[Anonymous Coward]

I don't think they are better at chess. I think the computers are just better at the things that are useful in chess.

That's the same thing you moron!

Re:Yes.

[JimDabell]

I don't think they are better at chess.

Uh-huh.

They can analyze moves faster and remember more about their opponent's technique

...the computer can apply the rules ridiculously fast.

You are saying yourself that the computer is better at chess. It's clear to see that the computer is better at chess - it can beat the world's best players.

What I think you are actually trying to say is that computers aren't smarter than humans yet. I agree. Crunching numbers is not a sign of intelligence in computers.

However, this new computer is smarter than the old one. It does more in less computations.

Re:Yes.

[jgerman]

Err, therefore... computers are better at chess.

A game that a computer will never dominate...

[Anonymous Coward]

Yatzee.

MACHINES GOOD

[Anonymous Coward]

PEOPLE BAD

fear the day that...

[havaloc]

Deep Junior calls Kasparov 'Coppertop'

Less moves...?!?!

[JHelgie]

" despite evaluating less moves per minute than Deep Blue,"

How many years ago was the match against deep blue? shouldn't it be VERY easy for them to acquire something far more powerful that deep blue was?

Re:Less moves...?!?!

[havardi]

Careful lets not start another war about pipelines and the "Megahertz Myth". . . .

Re:Less moves...?!?!

[Anonymous Coward]

Read the wired article, deep jr is using a completely different strategy. Deep blue tried to find every possible move, deep jr only looks at the good ones. They attempted to program jr to play more like a human, unfortunately humans are what kasparov is inhumanly good at trouncing.

Re:Less moves...?!?!

[jpmorgan]

If I remember correctly, Deep Blue was a modified customised RS/6000. I believe a significant portion of the move evaluation algorithm was implemented in hardware on top of an otherwise fairly beefy machine, meaning Deep Blue could evaluate vast numbers of moves per second.

To beat it in purely software today would still require some serious hardware.

where to view game replays, and watch live games:

[herrd0kt0r]

had the editors of the almighty slashdot bothered to read my infinitely more entertaining and informative story submission on this match, they might have had this interesting link included in their story, amongst other informative and titillating urls. this particular link will show the replay of today's game, as well as live broadcasts of future games.

boo to the editors.
yay to herrd0kt0r.
herrd0kt0r for prez.

So Who DOES. . .

[havardi]

. . .have the advantage anyway?

Black or White??

Re:So Who DOES. . .

[hdparm]

This question is being discussed every time machine plays against grandmaster. Definitive answer is white. That one tempo makes (very often) all the difference when opponents are of the same/similar strength (in chess terms).

Re:So Who DOES. . .

[Copid]

I tend to agree for all practical purposes, but there is a small possibility that in the end, that evaluation isn't quite right. Since there are a finite number of possible games, it's possible that there is a way for one side or the other to force a win (it's also possible, like tic tac toe, that two players playing the best lines will always draw).

There are times in many endgames when being able to skip a move and force your opponent to move would keep you out of serious hot water. The question is, is there a way for black to always force some line of play that causes white to have to "jump the gun" and degrade his position? My guess is that there is a roughly equal number of "white wins" possibilities and "black wins" possibilities, so the tempo white gets allows him to force a "white wins" line every time, but wouldn't it be interesting if it went the other way?

What if Kasparov wins

[ExCEPTION]

Kasparov could win, but cautions should be taken. Who knows if Deep Junior Junior Junior Juior would send some robot to kill him. The history will be altered and mankind won't stand a chance against the machine.

Hey there Milton.....

[Anonymous Coward]

What's....happening. We're going to need you to go ahead and um, move your desk to the basement. Okay, great.

Who else saw the headline...

[zapfie]

...and immediately thought "The Matrix"?

Re:Who else saw the headline...

[smash]

Actually the first thing I thought of was the Terminator series... but maybe I'm just showing my age ;)

smash.

No match

[gmuslera]

I suppose that a cluster of computers can resolve the game of chess in a future, i.e. all possible moves in any game, so with this database (that can have a really astronomical amount of alternatives, but with the rigtht representation of data it maybe will not take all available magnetic/optic storage in the world)

Right now, with some sort of position evaluation engine, this supercomputers can calculate the relevant part of that tree for the match they are playing with a lot of turns in advance.

Its only matter of time till er.. "intuition" will not be enough for chess.

Fortunatelly, there is a lot of fields where pure calculations is not enough, computers may be faster, but we can take this with humor.

Re:No match

[nomadic]

Right, it always annoys me hearing people insist that computers will never really be able to always beat us, because of intuition/experience/unpredictability/etc. Eventually there will be a computer that can crush any human player through pure computational ability, without having to be specifically programmed for that opponent (like Deep Blue was).

Re:No match

[basic70]

No, chess can't be solved. There are more positions in chess than atoms in our universe, so that method is out.

Re:No match

[gmuslera]

There are also more numbers than atoms in the universe and that not stop matematicians to discover some general properties of all or part of them.

If some encoding of board position/game development gives some method of dealing with winning positions maybe you don't need to generate all possible positions, but work in number theory to prove, whatever, that whites always wins or something like that.

You know, sometimes changing how to look to a problem solves it (like printing pi in base 31). I'm not saying that this will be possible, only that is not so easily discarded because, well, unlimited probabilities.

Can we please not make this a race issue....

[WIAKywbfatw]

There you go, bringing colour into everything.

Can't we all just learn to love each other and give peace a chance?

Re:Can we please not make this a race issue....

[Anonymous Coward]

Yet, Kasparov is Jewish. What colour does HE have? Brown?

the only people claiming deep fritz deep blue

[jbellis]

are fritz's PR people :(

you'd have to have a hell of a lot better evaluation function to overcome calculating 1/100 as many positions per second, and deep blue's eval was miles better than fritz's back in '97. from what I've read on rec.games.chess, fritz may have CAUGHT UP in the eval department but it's not 100 times better for sure.

if you're interested in computer chess, check out "behind deep blue," by IBM's team lead. most interesting book I've read in a long time. One part I didn't know was that IBM's move generator & eval function were done in hardware, which is the main reason that even with 6 years of moore's law under its belt, deep fritz can't touch it for sheer power. I always got the impression from the general media that deep blue was just a software program on a massive RS/6000 box but no, it had hundreds of these custom chess boards in it, too.

re kasparov's claims of cheating, remember there's two sides to every story and you're only getting one. For his part, Hsu says that he tried to get garry's team to agree to a rematch both with IBM and after he left, and kasparov's team basically dodged while complaining loudly and pubicly that Hsu was running away from him. Perhaps the truth lies somewhere in between, but given the obvious huge size of garry's ego I'd take what he says with a correspondingly large grain of salt.

Kasparov won the first match.

[Venti]

Junior sacrifised its crook for an uknown reason (to me) and lost...

Re:Kasparov won the first match.

[rootofevil]

what is a crook? (besides richard nixon)

Re:Kasparov won the first match.

[Anonymous Coward]

Don't worry, Junior is just messing with Kasparov's head. :-) Watch the next game closely...

Re:Kasparov won the first match.

[Anonymous Coward]

Disclaimer: I can beat my palm (oh er) on its hardest setting, but I have a 15 yr old chess computer (Novag) with about 16k of memory and an 8khz (thats right, kilo not mega) processor that cleans my clock.

Q: Why did it sacrifice its Rook?

Up until that point they were even, piece for piece (but with bishops on opposite colours).

Black's bishop was blocked by his queen, and one of his knights had been chased to the side of the board. Note that White could have swapped a knight for the last Black Bishop (normally considered a small advantage), but the Black Bishop is so ineffectual, and the White Knight so strongly placed that this doesn't happen.

White was threatening to move up its second knight by taking the pawn that Black needed to support its Knight breaking out (f4, e2)... (though the computer showed no inclination towards trying this on, I'd have liked to see a strong counter attack)

I thought Blacks Bishop move prior to this critical juncture was a bit weak (it was trying to free up its back rank for its Rooks), and I confess to not being able to figure out why Black decided to move its other Knight round to join its pal on the side of the board away from all the action.

White's move of the King into the corner is to allow the pawn play that brings about the queen swap.

The humans running the computer resign when White gets a Rook onto the seventh Rank (considered a large advantage). I would have liked to have seen it played out (I wanted to know wtf it was doing with those darned horsies).

Or to put this into slashdot terms, this looks at first glance like Gary laying the smack down with the equivalent of a tombstone from the top rope followed by a total bitchslapping (with a noogie thrown in for good measure).

Cynics will note the similarity to previous matches where Gary kicks the computers butt in the early matches, and then the programmers change its parameters each time to play in the style of a different International Grandmaster each game to throw him off.

where's kramnik?

[outsider007]

shouldn't kramnik get to defend our honor since he's the world champion? kasparov is higher rated but maybe only because kramnik hasn't been playing much.

is this like when apollo creed stands in for rocky to fight dolph lundgren in rocky IV?

Re:where's kramnik?

[Anonymous Coward]

He already did late last year in Bahrain against Deep Fritz. They tied.

Deep Fritz proved that it wasn't quite horsepower but efficient pruning algorithms which can produce an excellent chess AI.

Re:where's kramnik?

[Anonymous Coward]

Kramnik played against Deep Fritz a few months ago. The official website for the match seems to be down), but other sites report that they tied.

It was pretty big news back then. Even Slashdot covered the event.

Re:where's kramnik?

[frank_adrian314159]

The world of Grandmaster-level chess is truly screwed up at this point in history. There are at least three different organization claiming that they and only they have the right to name the World Chess Champion with most GM's not recognizing the true legitimacy of any of them. At this time, no one really knows who the best chess player is. It would take about $30M in funding to gather the 16 best players together to play a round-robin tournament that would settle this thing once and for all and, even then, you'd have people bitching about which 16 should have been invited, how many games should each player play against each other, etc. I blame FIDE for it's continual screwing of players and internal corruption for this mess.

Better player?

[Trevalyx]

Seeing as Deep Junior has lost the first match, this would infer to me that Deep Blue was the better player.. But time will tell, I suppose. Go programmers!

Re:Better player?

[Jacer]

Kasprov won the first match against deep-blue, too. Deep-Blue was allowed to be programmed with all of kasprov's moves, and updated after each match, i think kasprov maybe even won the first too, maybe someone will post more accurate information than my fuzzy memor

Re:Better player?

[Anonymous Coward]

If you infer from a single piece of sample data, you've got a rough life ahead of you. ;-)

Re:Better player?

[Copid]

Kasparov won the first Deep Blue match. My thought is this: DB always plays at its best (being a computer). Kasparov has good days and bad days (and occasionally becomes a basket case like when Deep Blue had him on the run toward the end of those matches). Since Kasparov won even one game (I believe he won two... my memory is fuzzy on this as well), I'd say that makes him the better player, all other things being equal. I also think that given time to regroup and play again, he'd have taken Deep Blue out. Pure speculation, but one cannot discount the fact that humans don't handle stress nearly as well as computers.

wow

[marekbrz]

I still have trouble beating the chess program on my palm, set to the easiest setting, and clicking on the hint button every move!

Re:wow

[Fr33z0r]

Heheh, that's more "Machine makes last stand - against self" though, isn't it? :)

Wait a minute...

[goatasaur]

Did anyone else think of The Matrix when they looked at the title?

Re:Wait a minute...

[bstadil]

More Terminator 3 Rise of the Machines

Re:Wait a minute...

[Anonymous Coward]

Yes.

If you want to know if someone thought of that first, READ THE PREVIOUS COMMENTS.

Re:Wait a minute...

[goatasaur]

Whoa, deja vu... must be a glitch.

Re:Wait a minute...

[Anonymous Coward]

Is it a coincidence that the following Sourceforge banner popped up on the page?

Coding in a Matrix?
Wake up from project nightmares.

PGN of the game?

[Dr. Spork]

Some kind person, please link a .pgn file. My Google search failed miserably.

Re:PGN of the game?

[angle_slam]

A link to the PGN is on this page here.

Re:PGN of the game?

[gandalf013]

This is the link to the PGN file. This has a java applet where you can replay the game.

Big Dilemma

[Anonymous Coward]

Its a scientifically proven fact that watching more and more porn leads to an insensitivity to normal tiltilating pictures and necessiates more perverse pornography. Which, in a laymans words, just means that internet makes a person perverted.

As we all know, help for linux is available only from newsgroups. Manuals from distros are useless, i am sure all of you would agree.

Using too much linux thus would lead you to watching more porn. And at a point of time, straight sex will not turn you on! you will be forced to seek excitement by watching gay porn.

I hope you all now realise the damage of using linux. It will turn you into a total fag. Dont use it.

Play safe, Use windows.

Movie Idea

[long_john_stewart_mi]

The year is 2003. The world is being taken over by chess playing robots. Our only hope is one man: Garry Kasparov (played by Arnold Schwarzenegger... A tough sell, I know). He has to control his childish temper as he takes on Deep Blue, Deep Junior, Deep Fritz, and (We're In) Deep Shit. Sure, they look like sissy beige boxes, but they're tough. There will be no time to pout, no leaving in disgrace; every move is on the clock (so to speak). In the final scene, Kasparov beats Deep Blue to a pulp with a Louiseville Slugger. So much for strategy! Astalavista baby!

Re:Movie Idea

[jlechem]

This is funny! Mod him up you bastards...

grr, "plain old text" ate my ">"

[jbellis]

stupid slashcode :(

NO, NO!

[Anonymous Coward]

wait, it's not a human programmer against Kasparov, it's a human programmer against an imitation of Kasparov's teacher...

Mod up - extreamly cool web site

[mtm_king]

Man, is that done with java or what???

Sorry I used all my mod points before I looked at you web site.

Re:Mod up - extreamly cool web site

[herrd0kt0r]

it's flash.
pretty nifty, no?

Clusters of Quantum Computers Playing Chess

[cyber_rigger]

By then the winner (or loser) will determined by the coin toss of who goes first.

something to remember

[goatasaur]

Kirk always beat Spock at chess.
/trekkie

Re:something to remember

[Anonymous Coward]

well, there you go!
proof positive! a reference to a star trek episode. if you can't take that as fact these days, i don't know what you can.

Re:something to remember

[Anonymous Coward]

Uh, you're sig is stupid. You can't moderate a thread you've commented in (and vice versa).

you're retarded

[jbellis]

obviously the words "branching factor" mean nothing to you :(

short version: computing every possible chess move is computationally intractable, even with computers billions of times more powerful than those we have.

long version: look up some past threads on this subject in rec.games.chess on groups.google.com...

Re:you're retarded

[gmuslera]

I read somewhere years ago that the amount of possibles 10 first chess moves are more than all stars of universe, and probably a few more moves could cover also all elemental particles in the universe.

But I can't say for sure that someone will come with a shortcut to that kind of calculation. Is that kind of inspirations that I can espect from humans.

Re:you're retarded

[kasperd]

obviously the words "branching factor" mean nothing to you

Obviously the words "dynamic programming" mean nothing to you.

with computers billions of times more powerful than those we have.

With such powerfull computers we would not be able to compute every possible game, but we would be able to store every possible state. The numbers of states is a lot smaler than the number of games, and the states is all you need to play perfect chess. The best move does not depend on how the state was reached.

Re:dear slightly less-than dumbass moderator,

[herrd0kt0r]

i commend you for moderating fairly! the parent to this is indeed offtopic. bravo!

but still, rather than wasting your time deciding who sux0rz...

oh, nevermind.
hehaohahoaho!

Diversionary tactic! We are far from lost!!!

[Anonymous Coward]

Everybody: You are being lied to.

The AI wants you to think that Chess is the last bastion of human analytical superiority. It's not. (Go is).

We are led to believe (by the AI, who control google news), that if the best computer wins more games out of seven than the best human at CHESS, then we must bow before the AI, as its intellectual inferior. Wrong.

First of all, as long as we are winning one single game against the computer under tournament settings, we've got a chance. Karpov may have only drawn against deep fritz, but you know what? That means we have a chance: That draw includes some wins.

Kasparov won some games before ultimately losing to Deep Blue in 97. Now he's already won one more in 2003.

But as interesting as this is it's not the issue.

Chess is a game chosen by the AI to deceive you, because computers happen to be, today, really, really good at Chess. With judicious pruning, they have look-ahead trees of ten, fifteen, twenty, fifty moves. Folks, that means that except for some cute evaluation software to determine what lines to prune down, they're basically brute-forcing their way into winning.

And they want us to bow before this brute force?

Never!

They can brute-force their way out of 56 bits, sure.

But let's throw them against 128 bits.

Let's throw them against Go.

From "The Game of Go" by Matthew Macfadyen, page 122:
(I'm typing this for you out of a book -- and first-strike claim fair use with +2 save for being anonymous).

Computers and Go.

Gary Kasparov's recent difficulty in handling computer opponents has been described as the fall of the last citadel in the battle of humans against the encroaching computer meance, but Go still stands as a refuge well beyond the reach of curent programmers. This is not for want of trying. The late Ing Chiang Ki from Taiwan sponsored an annual Go tournament for computers with good prize money, and several of the entrants put ijn years of work on their programs. But the tournament finishes with a challenge match between thew inner and a teenage human. This has to be played with a huge handicap - currently 14 free moves at the start of the game - and this is only diminishing slowly.
Part of the reason for this is accidental. Although for a human brain Chess and Go present similar challenges, there is an easy way to see how well you are doing at Chess - just count the pieces on the board. Looking a few moves ahead and counting the balance of pieces hich results gives a quick and easy way to avoid silly moves. Chess computers only need to be clever at sorting out a small number of "sensible" alternatives.
There is no such simple method in Go. Positions do not have a clear value until the game is finished, and the same pattern of stones may work perfectly in one context and be almost wrothless in another. One source of the great strategic trichness of Go is that you can choose between making large-scale loose formations and small-scale solid ones and each provides for different types of efficient development.
There are certain types of localized position in which computers have been used to find the right moves by exhaustive analysis, but even for quite modest-sized problems the programs run into millions of varioations. This is simply not a practical approach to most Go positions. It comes as something of a relief to discover that methodical calculation, considering all the possible outcomes, is neither necessary nor very useful in Go.

So. Let's concentrate on Go! In which the WORLD'S BEST computer program gets beaten - not by the world champion, but by a GIRL or BOY possibly still in highschool -- after being given more than ten moves to make without human response.

Computers are toast, even at a simple game with only two rules, one of which is hardly ever used and is just a "hack" to make infinite loops impossible. Humph.

Note: Another reason look-ahead-trees don't work for shit in Go is that at every point in the game, you can move to any free square. Typically, this means the first player has a choice of 361 squares for the first move, with the player making move 2 have 360, for move 3 there are 359, etc, with the only change in this pattern occuring when pieces are captured, pretty rare in professional games. (You just threaten to capture). So the "base" of the exponent is differnet AND you can't prune the look-ahead tree.
Chess has been SOMEWHAT brute-forced. So what.
Few things useful in the real world are as closed (8x8 board; clear general concept of positional value [number and location of important pieces]) as Chess.

So don't let the AI tell you chess is the last stance. Go is.

A different test: man versus machine

[goombah99]

It seems to me that if you want to pit man versus machine you should pick something that is easy for a man to do. Chess is relatively hard for most humans. Thus by definiton it is not something humans are good at. So making this a test of machine prowsess is exactly the wrong test.

to put this another way, if the contest were to factor 20 digt numbers, no one woul dbe surprised if the machine beat a human. it would be a stupid test. Just like chess.

a better test would be a face recognition contest. Or if we need to make it a real game then how about soccer?

Re:A different test: man versus machine

[PissedOffGuy]

to put this another way, if the contest were to factor 20 digt numbers, no one woul dbe surprised if the machine beat a human. it would be a stupid test. Just like chess.

a better test would be a face recognition contest. Or if we need to make it a real game then how about soccer?

another interesting thing to note is that 50 years ago, people thought chess was a pretty damn good test of AI. now people think otherwise. when the computer recognizes faces better than you, plays soccer better than you, writes poetry better than you, steals your girlfriend, and passes the turing test, will you still think its just "following the rules"? your brain is just following the rules of physics too you know.

Re:A different test: man versus machine

[Zork the Almighty]

On the contrary, chess is an excellent test of man versus machine. It is interesting precisely because champion chess players are not human calculating machines. At each turn, Kasparov chooses from only a handful of possible moves. He uses his brain, and with it some process which we can currently only dream of implementing in a computer, to find those "good" moves. When or if the day arrives that we can emulate this process on a machine, there will no longer be a contest worthy of our attention or consideration. But until then, there is a sporting game to be played.

Re:A different test: man versus machine

[Scarblac]

It seems to me that if you want to pit man versus machine you should pick something that is easy for a man to do.

Seems to me that if you want to have some contest, you pick something that they're both about equally good at. So we don't let people run against cars, and we don't let machines recognize faces against humans.

When Kasparov lost to Deep Blue, it was a huge surprise, he played weakly. Kramnik drew Deep Fritz 3-3 last year. Kasparov is the favorite again in this match, and leads 1-0. It's balanced.

What makes it more fun is that computers and people approach the game in a totally different way, but the best computers are almost as good as the best humans. This is the right time to be having these contests.

Re:A different test: man versus machine

[Anonymous Coward]

Your argument sort of negates itself

Yes, the process a human uses to be a good chess player is fascinating, and incredibly important to try to understand. A computer's process for playing chess is much more dull and uninteresting.

It's a brute force vs. finess comparison.

Now, piling on more and more brute force until you can finally defeat the finesse players (humans) does nothing to teach you about how finesse works. And it proves nothing. With a game like chess, at some point, enough brute force will win.

So who cares?

Write a program that can defeat the best human chess player while running on a Ti-85 and you've likely had some significant insight.

Re:A different test: man versus machine

[Froobly]

Or if we need to make it a real game then how about soccer?

And it's being done, with the Robocup league. According to This site, the goal is to have a fully bipedal team of autonomous robots go head-to-head against the winners of the year's World Cup by 2050.

Re:A different test: man versus machine

[Forgotten]

On the other hand, the Video Chess program on my Atari 2600 can handily beat me.

I wish I were joking.

Re:A different test: man versus machine

[domninus.DDR]

Yea last year the carnegie mellon team made the first successful robot to robot pass during gameplay.

Re:A different test: man versus machine

[Zork the Almighty]

Actually I think my argument holds up nicely. The current computers facing Kasparov are powerful, but their brute force is not enough to overwhelm the match. The raw complexity of chess will continue to dwarf any affordable computer for years to come (although I'm willing to bet the NEC Earth Simulator is powerful enough to precompute every possible game, given enough time). Nobody is interested in a match where raw power, applied stupidly, defeats a human opponent. At the same time, chess really is a hard game, and we need computers at least as powerful as the one Kasparov is playing to even make it a worthwhile match. As our analysis of chess improves, and better programming routines are developed, we will eventually start pitting slower and slower computers against human opponents. It is interesting only so long as it is a surmountable challenge to both opponents.

Re:A different test: man versus machine

[Froobly]

That's why more and more sophisticated heuristics are researched. If it were all about brute force, then people like you and me could code up a chess-playing program capable of going toe-to-toe with Deep Blue in a day or two. But they can't. There's a reason why nobody in my AI class could make their Checkers program beat Chinook at the highest difficulty setting, despite Chinook being only a fraction of its computing power in actual tournaments.

Making an underpowered machine perform as well as a more powerful machine is perhaps the definition of finesse.

Re:A different test: man versus machine

[archaos]

The Go game is definately something to look at. Indeed, no computer is actually able to beat an average amateur player. Some reasons behind are are huge branching factors, bigger board size, more visually oriented game...

According to this Computer Go discussion:
There is much yet to be done in the field of computer go. While many different approaches have been tried, the level of the best go playing programs is still low, even compared to amateur dan players (at least 10 stones !), not to mention professionals...

Go is really worth trying

Re:A different test: man versus machine

[drinkypoo]

Nobody is interested in a match where raw power, applied stupidly, defeats a human opponent.

On the other hand, everyone is interested in a match where raw human power, applied stupidly, defeats an opponent; both another human opponent (witness: boxing, or wrestlemania hospital, if you know what I mean) or a machine -- By the latter, I think you know I'm talking about midgets pulling a cargo plane on Fox. The network that'll do anything for your money. Make sure you write them a letter asking to see more tits.

Re:A different test: man versus machine

[drinkypoo]

Where it really kicks my ass is Kaboom!

Re:A different test: man versus machine

[ComaVN]

although I'm willing to bet the NEC Earth Simulator is powerful enough to precompute every possible game, given enough time

I'm pretty sure it's not. Particles in the universe and picoseconds since the big bang come to mind.

Re:A different test: man versus machine

[Anonymous Coward]

To me it seems interesting that for quite some years the best programs are almost exactly as powerful as the best human players with playing chess. That indicates that both humans and computers approach the limit of the "best possible chess playing". And they do it in a completely different way.

Re:A different test: man versus machine

[mwm158]

(although I'm willing to bet the NEC Earth Simulator is powerful enough to precompute every possible game, given enough time)

Given enough time, my old 486 could do it too.

Re:A different test: man versus machine

[Anonymous Coward]

There are literally an infinite number of combinations of moves in a chase game, think of moving a castle back and forth. No computer can or ever will be able to precompute the infinite number of games. It has to take into acount how logical the 1st move is before going down that branch. It makes decisions as to which moves to even bother analysing.

Re:A different test: man versus machine

[Anonymous Coward]

although I'm willing to bet the NEC Earth Simulator is powerful enough to precompute every possible game, given enough time

I'm pretty sure it's not. Particles in the universe and picoseconds since the big bang come to mind.

Just to expand on that a little. Apparantly the number of possible states in a chess game is around 10^120. That number is bigger than the number of particles in the universe (which is around 10^97 I think, although I read this a long time ago and have a poor memory anyway). I can't remember how many picoseconds there have been since the big bang, but I have a feeling it is less than the 10^97 number.

Re:A different test: man versus machine

[arkanes]

This is something that pisses AI researchers off to no end. As soon as they manage to get computers to do something hard (facial recognition, chess) it's not considered AI anymore.

For what it's worth, there are robots that play soccer. It'd be pretty easy to make one that can beat human players, too - just give it a little air cannon to shoot the ball with. But, of course, that would be "cheating".

Re:A different test: man versus machine

[Resseguie]

Here is the official RoboCup site.

This year's competition is to be held in Padua, Itally July 2nd - July 11th.

Re:A different test: man versus machine

[Abcd1234]

Noo... there is a finite number of chess board configurations, and THAT is the part that really matters. I could care less about move combinations. The point is that, if I know about all the board configurations available, then, given a particular configuration, I can always select the perfect move which will allow me to win (assuming I play first, of course). The heuristics which a chess program uses simply allow it to "guess" which move is best in the absence of this information.

As an example from the checkers world, Chinook, the top checkers program in the world, contains a database of all the endgames up to, IIRC, 7 levels deep. So, once it's at the point where it recognizes a board configuration from it's endgame database, it's guaranteed to win.

Re:A different test: man versus machine

[dumbunny]

Computers have lost ground in soccer over the past half century. In the 40's and 50's, some computers were 50' by 25', and would have made some very imposing goalkeepers. Today's "full-sized" towers are a joke by comparison.

Re:A different test: man versus machine

[tscaulfield]

It is because chess is difficult for humans that it is the perfect test. To determine the limits of a machine you won't run software you know it can handle, instead you'll try your very hardest to crash the darned thing. Same goes for humans, the more difficult the challenge, the more accurate the measurement of success.

However, it seems to me that playing chess against a machine is hardly an accurate measurement of human intelligence. Rather, it is a test of cause-and-effect predictability.

A machine has a perfectly logical understanding understanding of chess, and is going to weigh options mathematically, not accounting for variations of human intellect. Thus, a skilled player (this still obviously requires intelligence, just a different kind) could 'psych out' a machine. Human behaviour is far more erratic and therefore unpredicable.

Re:A different test: man versus machine

[zCyl]

when the computer recognizes faces better than you, plays soccer better than you, writes poetry better than you, steals your girlfriend, and passes the turing test, will you still think its just "following the rules"? your brain is just following the rules of physics too you know.

If someone ever designs a computer that can steal my girlfriend, I will certainly give that computer a little lesson in the laws of physics...

Re:A different test: man versus machine

[Anonymous Coward]

when a computer can do all that
I ll tell it go back to fucking playing chess

Re:A different test: man versus machine

[OrangeTide]

Even poor players of Go are able to beat the computerized Go engines out there. It's a much more difficult game to play by "brute force" through evaluating all movies. Humans are pretty good at Go since Go is about pattern recognition. Something we do many times a second in our day-to-day lives.

Re:A different test: man versus machine

[Chazmati]

You're forgetting how many chess board configurations there are. It may be theoretically finite, but practically speaking it's infinite.

And even if you only tabulate the endgame positions, it won't do you any good. You'll get shredded in the midgame.

Re:A different test: man versus machine

[Abcd1234]

Okay, the grandparent poster said:

"No computer can or ever will be able to precompute the infinite number of games."

My point is that this is wrong. No, it really is. This statement was predicated on the assumption that there are an infinite number of "games", which there isn't. There are a finite number of board configurations, period. So, given enough computing power and enough space, you can, in fact, precompute all the possible games.

So, to you, yes, the number of configurations is practically infinite, but to a computer with enough resources, it's not. Thus, given that you could precompute all board configurations, you most definitely can play a perfect game, beginning, middle, and end. 'course, this isn't really AI in the intuitive sense... it's just data processing. But that wasn't really the point of the original comment. :)

Note, I never said this task was practical, at least with todays technology. My point is that it's possible, and that the grandparent's assertion that it wasn't (due to the "infiniteness" of chess) was clearly false.

Re:A different test: man versus machine

[daffmeister]

My point is that this is wrong. No, it really is. This statement was predicated on the assumption that there are an infinite number of "games", which there isn't. There are a finite number of board configurations, period. So, given enough computing power and enough space, you can, in fact, precompute all the possible games.

There's finite and there's finite. As another poster pointed out: number of atoms in the universe and pico-seconds since the big bang come to mind.

Yes there are a finite number of games in chess but the number is so huge that no amount of computing power increases will ever overcome it. We're not just talking about Moore's Law here, we're talking about storage capacity greater than is possible with all the atoms of the universe.

So, to you, yes, the number of configurations is practically infinite, but to a computer with enough resources, it's not.

Sorry. Yes it is. No computer will ever have "enough resources".

Note also that the relatively smaller number of possible board positions doesn't help you either. You still need to determine the correct move for each position. Again, something that is impossible to compute in this universe.

Re:A different test: man versus machine

[catsidhe]

http://www.robocup.org/

Re:A different test: man versus machine

[goombah99]

let's see now. 32 pieces, 16 of which are highly mobile. assume that for any given non-pawn about half the board is available to it. this leads to something like 32! = ~ 10^35 possible combinations of putting the non pawns on the board. as pieces are taken this number changes. e.g. if we are down to 8 pieces its something like 8!(32!)/(24!) =~ 10^16 board combinations. And this is ignoring pawns.

maybe I got my combinatorics wrong. anyone want to check me?

Re:A different test: man versus machine

[sandow]

It's been pointed out that the big difference between humans and computers is their specialization. If the building catches fire during the match, Kasparov will flee the building but the computer will just sit there thinking about chess as it burns.

Right now, to beat the best humans you need to contrain the problem so much the result is an intelligent idiot. You'll notice that Deep Blue didn't get interviewed after beating Kasparov.

Re:A different test: man versus machine

[turingcomplete]

Actually I did an essay that showed that our minds were fundamentally superior to computers or any turing machine. You can check it out here if you're interested:
For
Against

Re:A different test: man versus machine

[AnnaBlack]

Why not pit woman vs. machine? Because girls have more sense than to spend their time competing :) Instead we'd just sit down and get the machine to talk about the problems it's having with its relationships...

Re:A different test: man versus machine

[renderhead]

If someone ever designs a computer that can steal my girlfriend, I will certainly give that computer a little lesson in the laws of physics...

If someone designs a computer that can steal my girlfriend, I will give that designer a physics lesson.

Re:where to view game replays, and watch live game

[Anonymous Coward]

boo to the editors.
yay to herrd0kt0r.
herrd0kt0r for prez.

Coolest Slashdot post ever.

EVER.

You rule.

Moves in Realtime + Links...

[Anonymous Coward]

So THAT's who "Anonymous Coward" is...
Links to real time moves here or here.
Javascript play-by-play of the first game.
NYT article condensed(nice human/computer strategic advantage diagram).

pgn

[Anonymous Coward]

[Event "X3D Man-Machine match"] [Site "New York City"] [Date "2003.01.26"] [Round "1"] [White "Kasparov(GM)"] [Black "Deep_Junior_1-0"] [Result "1-0"] [Opening "QGD semi-Slav: Stoltz variation"] [ECO "D45"] [NIC "SL.08"] [Time "14:20:47"] [TimeControl "7200+0"] 1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. e3 e6 5. Nf3 Nbd7 6. Qc2 Bd6 7. g4 dxc4 8. Bxc4 b6 9. e4 e5 10. g5 Nh5 11. Be3 O-O 12. O-O-O Qc7 13. d5 b5 14. dxc6 bxc4 15. Nb5 Qxc6 16. Nxd6 Bb7 17. Qc3 Rae8 18. Nxe8 Rxe8 19. Rhe1 Qb5 20. Nd2 Rc8 21. Kb1 Nf8 22. Ka1 Ng6 23. Rc1 Ba6 24. b3 cxb3 25. Qxb3 Ra8 26. Qxb5 Bxb5 27. Rc7 {White wins} 1-0

Background info

[Anonymous Coward]

Deep blue was the first computer to play a world champion. It played with huge advantages, and the match was poorly run. The result was questionable, but it did mean one thing: computers are here, and they're not going to get run over, even by the best of the best.

The next major match was Kramnik vs. Deep Fritz. Deep Fritz was the highest rated computer program, and Kramnik the second highest rated human and world champion. Kramnik blew the computer away at first, then mysteriously lost a few games to draw the match.

This is Kasparov (highest rated human, NOT WORLD CHAMPION), vs. Deep Junior (I forget its rating, but it's up there, and it's world champion {for computers}). Kasparov won the first game. The matche conditions heavily favor Gary, he even got to play with a version of the program pretty similar to what he saw today on his home computer for the last six months!

So, who's stronger than who, what's stronger than what, and who's stronger than what? Here's the breakdown:

Deep Blue is the weakest of the competitors. It looked at an ungodly number of positions, but mainly because it HAD to if it was to have a chance. It ran custom hardware and software. It was dismantled immediately after the competition. Most of the work put into it is lost.

Deep Fritz is a strong program. It employs modern techniques and runs within the windows operating system. It is a commercial program available for a reasonable price from chessbase (chessbase.com). It consistantly maintains the highest rating for computers, but this could be because of its ability to beat up on weaker computers consistantly. Deep Junior seems to beat it when it comes to the championships.

Deep Junior is the computer world champion. It is very similar to deep fritz: it runs under windows, can be bought from chessbase, and is very strong. It wins championships consistantly, but its rating is always a bit behind fritzy's -- again this could be because of its inability to play decisively against weaker programs.

Vladamir Kramnik is the current world champion and second highest rated player. He won the championship title from Kasparov and has held it since without rematch (the chess world is in shambles -- at least three people claim to be world champion and none defend their titles. Kramnik has the most legitimate claim, as shown by his rating and the conditions of his competition). He is a far superior chess player in comparison to Deep Fritz -- this was shown in their match. His second half problems were possibly exhaustion, boredom, or caused by less noble factors, but they are essentially meaningless. When he was at his best, he was destroying the computer handily.

Gary Kasparov is the current highest rated player in the world, by a large margin. He is an absolute terror over the board, few players even in the top 100 in the world stand a chance against him (even those in the top 10 do not necessarily fair well!). He is considered by many the strongest chess player of all time.

This match is an official FIDE (Federation International des echecs or International Chess Federation) Man vs. Machine world championship. Hopefully the first annual of such.

As to who's going to win: Kasparov won today, and I expect him to continue. He has the ability to adjust to all his opponents (except Kramnik), and the computer is very rigid in its style. Strength varies based on the type of position, and Kasparov will give the computer no oppurtunity to play "its game." I made a similar "human lots, machine little" prediction before Kramnik's match, though, and he collapsed in the second half. Much is still to be decided.

Re:Background info

[Freerange]

very quick Rules of Go. Helpful!

My first post on slashdot was a miserable failure

[Anonymous Coward]

let me try that again...

[Event "X3D Man-Machine match"]
[Site "New York City"]
[Date "2003.01.26"]
[Round "1"]
[White "Kasparov(GM)"]
[Black "Deep_Junior_1-0"]
[Result "1-0"]
[Opening "QGD semi-Slav: Stoltz variation"]
[ECO "D45"]
[NIC "SL.08"]
[Time "14:20:47"]
[TimeControl "7200+0"]


1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. e3 e6 5. Nf3 bd7 6. Qc2 Bd6 7. g4 dxc4 8. Bxc4 b6 9. e4 e5 10. g5 Nh5 11. Be3 O-O 12. O-O-O Qc7 13. d5 b5 14. dxc6 bxc4 15. Nb5 Qxc6 16. Nxd6 Bb7 17. Qc3 Rae8 18. Nxe8 Rxe8 19. Rhe1 Qb5 20. Nd2 Rc8 21. Kb1 Nf8 22. Ka1 Ng6 23. Rc1 Ba6 24. b3 cxb3 25. Qxb3 Ra8 26. Qxb5 Bxb5 27. Rc7 {White wins} 1-0

Deep Blue Cheated

[johnnyb]

I don't remember where I read this, but I think I remember seeing that the programming team for Deep Blue had the option of not doing what Deep Blue asked. This even happened in one of the games Deep Blue won in. Deep Blue made a blunder early on, but the programmer made a more sensible move instead.

Anyway, it seems that computer+human does better than human, not necessary computer by itself.

Re:Deep Blue Cheated

[Kipper the Llama]

And, if that's not enough, Deep Blue had an inventory of every single one of Kasparov's games on hand, and Kasparov didn't have a single Deep Blue game to look at- since none had been played.

In professional chess the vast majority of the best players have these frightening memories that can recall entire games. This is one of their tool to beat the opponent, knowing many of his games. Deep Blue had the obvious information advantage.

Re:Deep Blue Cheated

[ctid]

This is complete nonsense. No such incident occurred. Kasparov claimed that there was interference, but he was simply reacting emotionally to his poor play.

For an excellent description of the design of Deep Blue and the matches against Kasparov, see this book

Re:Deep Blue Cheated

[Anonymous Coward]

This is absolutely untrue. Kasparov spread this rumor himself.

Re:Deep Blue Cheated

[dmforcier]

No.

The programmer/operator did not have the option; that would go directly against the rules! The computer indicates its move on the screen, and the 'operator' pushes the piece on the board. In fact, there is an example in Hsu's book (I believe it was in the first Kasparov match) where the operator moved the wrong piece (not the move that Deep Thought had directed). The referee was called, the move put back, and the clocks reset.

The *only* "evidence" of human intervention is Kasparov's claim that moves 35 and 36 (?) of Game 2 in the second match "don't look like computer moves". Hsu himself said that both were "mistakes" by DB, and identified the bug that caused one.

If we are to rely on Kasparov's "understanding" of how a computer plays to identify points of human intervention, then we should be able to establish that GK understands computer play very well. Yet in the match in which he lost to DB, he uses the "anti-computer" style which failed utterly, thus disproving that he has that understanding (at least of Deep Blue).

Therefore, the is no credible evidence of any dishonesty on the part of IBM or the Deep Blue team, and to continually assert that there is is to insult the team and degrade the perveyor.

so...

[burns210]

is this deep junior a real chess player, or, like deep blue, just a program that has been tuned to beat this single opponent and noone else? These deep blue chess machine's are cool, but they are not, in my understanding, a true chess computer able to go against anyone.

The machine that won the match originally geared specifically to counter Kasparov and new his past games.

They should have this machine on the internet to play against, say, 50(not to overwelm it) people online at a time through a browser.

Not only that....

[NerveGas]

Not only was Deep Blue specifically designed to beat Kasparov, it would be re-tuned and configured after each match to further customize it to his playing style.

steve

something else to remember

[Spyky]

Star Trek is a TV show

-Spyky

It shows every move of the game

[mtm_king]

And you can step backward and forward - what have you done that is better?

Someone put some time and energy into creating the site and did a good job. If it is not cool enough for you it is your problem.

You are a jerk.

Re:It shows every move of the game

[Anonymous Coward]

He said it's "flash", as in "Macromedia Shockwave Flash", as opposed to "java", you idiot.

Re:It shows every move of the game

[jtdubs]

What the hell are you yelling at him for? He told you it was in Flash. You said it was cool. He said he agreed. At what point did he become a jerk?

When he said, "pretty nifty, no?" was that what made you mad? That phrase doesn't mean that he thought the site was not nifty. It meant "pretty nifty, don't you agree?"

Justin Dubs

Re:It shows every move of the game

[herrd0kt0r]

nonono! i was _agreeing_ with you!
i'll just chalk the "you are a jerk" up to a misunderstanding. pals?
8D

just to be clear: it's done with macromedia flash. and it IS sooper spiffy cool n froody. i like. 100%.

Re:It shows every move of the game

[herrd0kt0r]

ahhh, i get it now, maybe. he might've misread it as "pretty nifty, NOT!" you know, like wayne's world.

let's all be swell pals. we should be discussing more important things. like chess. or chess computers. or bewbies.

Lemme get this straight....

[NerveGas]

you're asking your opponent for hints? No wonder you keep losing. ; )

steve

Re:Diversionary tactic! We are far from lost!!!

[ronaldcromwell]

So don't let the AI tell you chess is the last stance. Go is.

I think you've got an unhealthy obsession with Go, sir. While you raise some interesting points, you're a fool to think that there is an upper limit to something like this. What if I invent a game like go, but with twice as many spaces?

Re:where to view game replays, and watch live game

[altek]

Ok mod me troll -1 all you want, but I also want to file a complaint with slashdot. I typed up a really nice summary of this story with links about a week ago when it would be actually relevant so people could watch it (instead of posting it AFTER the first game) and of course got rejected. Losing more and more faith in /. ... (and i have been here a very long time)

Can't wait...

[LadyLucky]

For the celebrity death match version to appear on eDonkey.

Re:Diversionary tactic! We are far from lost!!!

[Anonymous Coward]

we don't need twice as many spaces. Go already can never be brute-forced, but chess can. Chess is like 56-bit RSA, Go is like 1024-bit RSA. If you can break a 1024-bit key, 2048 probably won't be any better. The same is not true of 56 vs 128.

When a computer can beat a Go master at Go

[jimbobborg]

Then I'll be impressed. All these programs do is use brute force to find moves. Can't do that in Go!

Re:When a computer can beat a Go master at Go

[outsider007]

wrong. go is much more finite than chess. it's 19x19x3x2 (an intersection can be black, white or empty, black or white to move.) and really divide that by four since you can just rotate the board 90 degrees to get the extra variations. once you map the best move for each positions it's game over for humanity.

Re:When a computer can beat a Go master at Go

[CheeseCow]

What about this, a enormous database with the best moves for every situation. So what if it would be hard to do today. Calculate all the best moves, store them in a database and play along.

It is just a matter of memory & processing speed.

Re:When a computer can beat a Go master at Go

[jimbobborg]

No, you are wrong. No one has been able to create a game that can consistently beat anyone higher than 15K, roughly 1 year of experience. However, there are computer chess games that can consistently beat higher level chess players that you can buy at your local CompUSA or whatever software store you live near. So I stand by my statement.

Re:When a computer can beat a Go master at Go

[outsider007]

No one has been able to create a game that can consistently beat anyone higher than 15K

only for lack of trying. trust me it's a no-brainer.

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

You obviously don't know how to play Go.

Go relies on pattern matching and context as well being able to judge where to move next.

Think about this go 5 recursions deep on all the possible opening moves in go. The combinatorial explosion is insane.

Also you calculation is just wrong. Just imagine each position was a bit
19x19 is 361 bits.. How many values can we store in 361 bits? More than 19x19x3x2
More like 2^(19*19)
469708516554766645577896119357867405475 13650978166 397414145819430644\
18050229216886927397996769537 406063869952

And if we use your reasoning of "3" positions:
3^(19*19)
174089650659031927907188238 07056436794660272495026 354119482811870680\
10516761846498411627928898871 493861209698881632078 061375498718135509\
31295148033696605728930754681 80597603

Computers aren't going to brute force Go anytime soon.

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

You mean more bits in memory than we have atoms in the universe? Better hope we figure out how to use all them quantum states to encode billions of bits to a single atom...

Re:When a computer can beat a Go master at Go

[evanbd]

Actually, that number is more like 10k these days. Gnu Go version 3.3.15 (current development version) is rated at 10k* on nngs. You might find playing it to be an interesting experience. It is currently one of, if not the, strongest go programs available. It is available for Linux, Windows, etc, at http://www.gnu.org/software/gnugo/index.html.

Having said all that, it is far far away from dan level play. I believe the feeling among the developers is that it will get a few stones stronger simply through correcting mistakes and continued tuning, but that more than that will take some major changes and improvements.

I am an occasional contributor to Gnu Go, and play at the 8k* level on nngs. I can give Gnu Go 4 stones, but that is in large part by playing to its weaknesses. It is a worthy opponent, but its games tend to have the same feel to them. The tactics and life and death are remarkably strong, in the 1-3k range, but its global planning and other aspects are weak.

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

This is just wrong.
Lets say a chess board is full of pieces, lets say there are 16 types of pieces
so the number of possible board would be
16^(8*8)
ok?
Well then for Go
3^(19*19)

Which is larger?
64log2(16) 361log2(3)
well 16 needs 4 bits and 3 needs 2 bits

so 256 722... Obviously Chess is more finite than Go.

Re:When a computer can beat a Go master at Go

[ComputerSlicer23]

Actually, you've got that wrong. That'd be ( 19 * 19 ) ^ 3, not multiply by three. Second, just assume it's always whites turn. If you are playing black, flip your colors to white, and search for what white would do in this position. I'm not sure you can do the rotations, 4 doesn't divide evenly into ( 19 * 19 ) ^ 3, so it would seem that that's not accurate, I can't explain it any more then the math just doesn't work out evenly. However, intuitily your correct. In fact you should be able to flip horizontally, and vertically, and both diagonals for symmetric positions also. (Actually you can do it, it's just not a perfect division because 19 is odd and the center line doesn't have to match the anything, because it flips to itself).

Hmmm, I don't believe what your saying is true. Go is studied a lot by people at MIT, who happen to be very, very smart. In fact, it was specifically invented at MIT because it was harder for a computer to play than chess, specifically because it has a much lower pruning factor then Chess. In chess, you can tell quickly, that okay, of the 20 moves I can make 17 are just stupid, don't bother examining them.

I thought there was some state of the game other then merely the state of the board, I thought you could capture pieces. I've never played, or studied the rules. I have read up about the game and the pruning factor while reading about programming chess players.

Kirby

Re:When a computer can beat a Go master at Go

[outsider007]

no pun intended.

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

Man I wish it was (19*19)^3 I'd be able to make a kick ass Go player!

But the truth is you have it backwards it's 3^(19*19) each position can have 1 of 3 possible values. I don't know what (19*19)^3 would actually be but it's plane wrong. If we had a matrix of bits that was 8 by 8 we would not calculate the cardinatility of all the possible permuation by go
(8*8)^2==4096. We know that 64 bit numbers are worth more than 4096...

Moderators! Why do you keep modding up bad math err late night math!

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

Wrongo boyo

Go is at least 87970230512663383361180473441419406425598163983718 55086054580111175
2805646488913529495910776629039 62 times bigger than Chess (19x19 go).

Re:When a computer can beat a Go master at Go

[outsider007]

no it's x3 not ^3, and you can't substitute black for white because of komi. also there's ko considerations that I didn't mention and you have to realize that more than half of the theoretical board positions would never arise in an actual game. all things considered, much more finite than chess.

The reason there's no killer go app is because the programmers consider a database of 'best moves' cheating and instead try to come up w/ actual algorithms to do the thinking. really it's just a matter of compiling a very manageably-sized database.

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

Uh you're totally unaware of the size of this.

For instance in Bahrain with Deep Fritz vs Kramnik , Deep Fritz was allowed a little bit of a end game database. Fritz had 4 gb of end game positions (and links to winning moves) Now guess what this was only for a maximum of 5 pieces on the board at one time (total pieces) so that'd be like a pawn,king,knight vs pawn,king moves. These are simple moves but it takes over 4gb?? The only thing you can database in Go is Josekis and that's still not that great. I think you're not sure about the size of the problem.

Re:When a computer can beat a Go master at Go

[ComputerSlicer23]

Hmmm, I'll have to go read the rules. However, they are independent choices are they not?

Your right, it's not ( 19 * 19 ) ^ 3, it's 3 ^ ( 19 * 19 ), which isn't a very nice size.

For a 1 by 1, I get 3 ^ ( 1 * 1 ), you get ( 1 * 1 ) * 3. Okay they match.

Now take a 2 x 1. I get 9, you get 6.

First square can be: clear, white, black. ( 3 options ) Second squre can be : clear, white, black. ( 3 options )

( clear, clear ), ( clear, white ), ( clear, black ), ( white, clear ), ( white, white ), ( white, black ), ( black, clear ), ( black, white ), ( black, black )

Assuming the first square is independent of the second, it's 3 * 3, not 3 + 3. Simple combinatorics, I can't believe I screwed it up before.

Now there might be some pruning in there somewhere, because pieces have to be adjecent to each other or some other such nonesense. I'm not interested in doing that kind of combinatorics in a slashdot post. Sorry, that's too geeky even for me.

Lets use your logic on Chess positions. There are only 13 ways a square can be. Either pawn, king, queen, bishop, rook, knight of either black or white, or it could be empty. So you'd say it's ( 8 * 8 ) * 13 positions, so chess would have a total of 832 total positions. That's obviously wrong. I could memorize every position and the winning move if you'd give me money for doing so. I'd say it's 13 ^ ( 64 ). Which is a huge number. Now take my estimate of go and divide it by my estimate of the size of chess. You get a number with 102 numbers in it. So Go has a search space is 102 orders of magnitude bigger then Chess's search space. Assuming you could store the size of chess's brute force database in a single atom, you couldn't store Go's brute force database in all the atom's in the universe.

Mathematicians, and CompSci people don't believe it's cheating to brute force a game. In fact, they would do it just to know if the game really is balanced, or if white has a force win, or black has a forced win. They'd do it just to know, plus they'd used the brute force database to see when the algorithms choose the wrong moves to tune the algorithms to produce the perfect game. They'd also know how many pieces to spot the opponent at the start to be fair. They brute forced the game with rice that originated in Africa not that long ago just for giggles. It was posted to slash dot.

No way Go is brute forceable, and the guys at MIT didn't do it. Hell, the size you were discussing, isn't significantly bigger then tick tack toe, when dicussing it in terms of computing power of today.

Re:When a computer can beat a Go master at Go

[uXs]

Um. Invented at MIT ? I thought Go was invented somewhere in China, 3000 years ago. Short of using a time machine, I don't think it's possible for MIT to have invented Go...

Re:When a computer can beat a Go master at Go

[ComputerSlicer23]

Ahhh, you're correct, a professor at MIT merely picked it as the game to study because of the high branching factor. I'm wrong, wrong, wrong about my history, a quick google search shows you are correct. Thanks for the correction, added to my mental notes about the game.

http://www.usgo.org/resources/gohistory.asp

Says it was founded at least 4000 years ago... However, the math is correct in the second post.

Kirby

Re:When a computer can beat a Go master at Go

[JonathanBoyd]

I've never played Go, so I'm not sure about it's rules, but for Chess you have 32 pieces on the board, each of which can be assigned to a square. We start with 64 available, then 63, then 62, then 62 and so. Basicially you get that for 32 pieces on the board, 64!/(32!*2*8!*2*3*2!) = 5*10^47, allowing for the fact that several pieces are identical. But there could be 31 peices on the board. And there are 2*15!/(8!*3*2!) possible combinations of that, or roughly 10^7. Even at this very early stage, it becomes quite edious to work out the possible combinations of setting up 31 pieces on the board. I certainly wouldn't want to go all the way to working out the ways of setting up 3 pieces. And none of this takes into account rules about where pieces could be or possible places they could reach during gameplay. Or conversion of pawns to other peices. That gets messy.

In short, it's nowhere near as simple as the formula you were proposing and I'd be willing to bet it wouldn't be 102 orders of magnitude off either. Yes, Go would be insanely difficult to brute force, but Chess isn't particularly easy either.

To do this properly you'd actually have to look at all the possible starting moves, multiply by possible replies, then replies to that etc. to get the correct numbers and I couldn't do those calculations sitting here.

Incidentally, using your methodology and substituting in32 instead of 13, we would get 32^64 ~ (3^3)^64 = 3^192. This compares with 3^(19^2) = 3^381. That's a factor of 3^189 ~ 10^63 difference, i.e. 63 orders of magnitude.

Re:When a computer can beat a Go master at Go

[ComputerSlicer23]

Oh, I wasn't suggesting that Chess was as simple as the formula I suggested. However, my formula is accurate given 64 copies of every piece white and black, and attempting to setup every single setup physically possible (a huge number of these would be absolutely impossible in a real game of chess, for starters a huge number of the setups described wouldn't have any kings on them at all, obviously bad). Using the number 32 instead of 13 is a mistake. Because in that situation your differentiating between each of the eight pawns, lets say you have a pawns at a4 and b4, using 32 you differentiate between if the pawn at a4 started at a2, b2, c2 etc, etc. In terms of position it doesn't make any difference where it started. I'm merely counting game positions, using 32 would count game positions where the different starting positions of identical pieces makes unique. There is absolutely no good reason to use 32, that's even worse the using 13 (which is piss poor in terms of accuracy). The first formula you used if it's doing what I think it is, is the most accurate of all the ones listed by either of us. (Which I could have described, but like I said, it was too geeky to do in a slashdot post *grin*). Plus I get the impression that given the confusion over multiplication versus exponentiation would have been too much for the original poster to follow, and I didn't really care to explain the basics of combinations and permutations with and without repetition to him in a series of tedious slashdot postings... *big grin*.

However, thanks for making me feel lots less of a geek then I did earlier tonight...

Kirby

Re:When a computer can beat a Go master at Go

[JonathanBoyd]

Actually, I accounted for the multiple piece thing in my first section with the factorials. 32 is no worse than 13 because your formula is compeltely messed up anyway. Though quicker to work out than mine

Basically, 13 is competely wrong because you don't allow for the fact that there are 2 sides and multiple copies of all the pieces. This is way you have to start multiplying and dividing factorials. combinational and permutational stats.

If you realy wanted to look at how any ways you could set 32 pieces on the board, look at my first result - 32!/2*8!*3*2!. Actually, come to think of it, thet probably isn't even right.

32!/(8!*3*2!)^2 is what it should be. Assuming that the pawns haven't been queened. Allowing for that gives you several orders of magnitude more - each pawn can have 5 states so that's... erm... I'm not sure how many combinations. Gets tricky because some look the same as each other. Suffice to say, it would be on the order of thousands of cominations to be factored in.

Re:When a computer can beat a Go master at Go

[ComputerSlicer23]

Okay, I'll try one more time...

What I'm constructing is the set of all chess configurations that can exist assuming no sanity checking, given that you can have anywhere from 0-64 white queens. 0-64 black queens. 0-64 white kings, ...., 0-64 white pawns, 0-64 black pawns, 0-64 empty squares.

For any given square on a chess board, given that you have 64 duplicates of each chess piece (all six pieces in both colors for a total of 12 unique types of pieces), plus you can choose to leave the square empty. That's 13 unique things that can occupy a square. For each square you get to pick on of the thirteen. You get to make that choice 64 times, hence 13^64.

This will generate every single possible unique position you could ever have on a chess set. It's the absolute upper limit for what a chess set could look like given the 12 types of pieces and empty.

Using 32^64 is a mistake. First it should be 33^64, because you don't allow the square to be empty, using 32 you have to place a piece on every single square. Also, it means you could place the Queenside white rook twice which means that number is just completely meaningless. This is why when you want to use 32, you did it as a factorial expression, to avoid recounting. The true number is someplace between the factorial expression you have, and the exponential expression I gave.

Using 13^64, gives every single degenerate case you discuss, plus all the others there could ever be on a 64 square board that has 13 choices for what to place on it. Which precisely describes the chess board, and a set of chess pieces.

Consider the position where you put a white rook on square A1, and choose empty for all other squares. Using 13, that's it. White rook square a1, done. Using 32 means that Queen side white rook on square A1 is different then Kingside white rook on square A1, all the rest empty. I'm saying that positionally, which side the rook came from is irrelavent, it's a white rook, they are identical. You're not going to choose a different move because the rook is kings side versus queens side (okay you might if it means not being able to castle later, but I'm ignoring castling and en passant in this discussion, because you don't have to consider that when building the database). When you get the pawns the situation is just stupid. I didn't want to account for all the various factorials and replacement issues due to identical pieces (because I did all I wanted to by using the number 13), because I don't know enough Go to do the combinatorics on it, I wanted to do crappy combinatorics for both, and I was hoping to avoid precisely what I'm doig which is explaining the math to someone in a damn slashdot post! To be honest, I'm not terribly interested in getting either of games more than big picture correct.

See the difference.... Okay, now I'm being too geeky... I was really hoping not to argue combinatorics tonight, I've too much other shit I need to get done. Thanks for playing...

I'm capable of doing the math to get it highly accurate as long as we avoid considering illegal positions, it's just long and tedious, and if I wanted it that accurate, I'd just look it up in a book, I'm sure it exists someplace. Shit if you want, I'll write it up and send my complete analysis to you, but not in a slashdot forum.

Kirby

Re:When a computer can beat a Go master at Go

[JonathanBoyd]

Well I've already started the maths and understand it fine thanks, thanks and as I said, it's way too tedious. Wouldn't want either of us to go through it. I misunderstood what you were doing earlier - probably wasn't reading your psots closely enough. You're giving an upper limit to what the psoitions could be, whereas I've been trying to do it exactly, thinking you were doing something different. I think we're actually ina greement about what needs to be done, but I just misread you earlier, leading to the confusion.

Incidentally, I was taking into account that kingside rook on H1 and queenside on A1 is the same as queenside rook on H1 and kingside on A1. That's what the division factorials were for and why I was talking about 32 pieces - if you have 8 pawns on the board, then you simpy divide the total possible ways of placing them all on the board, assuming they're different pieces (64!/56!) by the number of ways you can arrange the pawns by simply swapping them (8!) so if we're just considering the pawns, that would be 64!/(56!*8!) or 4*10^9. Your method of estimating (2 piece types - pawn or blank square, 64 squares avilable) would give 2*10^19 - a wild overestimate. But that just reinforces your claim that go is a fair bit harder to force than chess.

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

Oh, it's quite easy, once you've played out every possible game so that you can CREATE that database of what the best move is in each position. Suure.

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

Amazing how you solve it and the moderated up idiots manage to get comments posted at +2 with totally incorrect math or math with copies your comment.

Re:When a computer can beat a Go master at Go

[k8to]

Chess:
8x8 board
64 locations, 13 piece types of 2 colors.
Locations can be filled or blank.
27 ^ 64
>>> 27 ** 64
404837660228432814111844721895716547522075068
820903057422001161010657660267188207581747750
41

Go:
19 x 19 board
361 locations, 1 piece types of 2 colors.
Locations can be filled or blank.
3 ^ 361
>>> 3 ** 361
1740896506590319279071882380705643679466027249
5026354119482811870680105167618464984116279288
9887149386120969888163207806137549871813550931
29514803369660572893075468180597603

Re:When a computer can beat a Go master at Go

[Anonymous Coward]

How the hell does this get a +2 Moderation when there are tonnes of AC posts which said exactly the same thing below!!! READ THE POSTS MODERATORS!

In the event of Man v. Machine war...

[Kipper the Llama]

Convince the machines that WinME is the greatest operating system of all time, then watch them randomly crash in a way that can only be suicidal.

Really though, I think inventing a suicidal OS was quite an accomplishment. Too bad I have to use it.

Not forgetting the recent lawsuit

[bananahammock]

The FT mentions this and it has appeared elsewhere: "Another factor against Kasparov is that his preparation has been interrupted by legal problems. The First International Bank of Israel is suing Mr Kasparov for damages after Kasparov Chess Online Inc, Mr Kasparov's troubled company, failed to repay a $1.5m loan. The unresolved dispute is said to have taken almost two weeks out of the chess player's training schedule." And Deep Junior is what, an Israeli-designed programme. Even Kasparov thought that this lawsuit wasn't just a coincidence.

It's very obvious

[Anonymous Coward 1]

That the machines will soon be the most efficient chess players, which will wreak hav0c on the domestic chess players market.

Go humans, go!

[Anonymous Coward]

Go humans, go!

Errr, wait... Chess, human, chess! Ok, nevermind.

Could be a hole?

[edox.]

Garry Kasparov crushed the champion computer program Deep Junior in his trademark aggressive style on Sunday in the first game of their six-game

Deep Junior programmers say it focuses more on strategy than on capturing the opponent's chess pieces quickly.

Garry was clever enough to understand that!

a thought..

[MoceanWorker]

a while back i read that Vladimir Kramnik beat Kasparov.. so why not let him take up the challenge as well and see how he will do

Re:a thought..

[haggar]

Because Kasparov is stil world number one chess player, based on the points system. As an example, just because some guy beat Agassi doesn't mean that Agassi will automatically lose hist position in the ATP table, and certainly doesn't mean that some guy will take his place.

Re:a thought..

[MoceanWorker]

true, but.. such a young mind.. such a brilliant future.. might as well give it a shot :-)

Re:a thought..

[plateau]

Deep Fritz and Deep Junior played a match, on much slower machines (dual pentium 933MHz's), to decide who would play Kramnik and who would play Kasparov.

The match was a tie, with Fritz winning the tie breaker.

Deep Fritz then played classical world champion Kramnik in October of last year, drawing him 4.0-4.0.

Re:a thought..

[dmforcier]

Actually, Hsu (hardware designer of Deep Blue) himself suggests that Kramnik would be a more difficult opponent for any computer (including Deep Blue) because his genius is in positional play where Kasparov's seems to be in tactical play. Computers have more difficulty recognizing and extrapolating patterns (positions) than moves.

And Kramnik might be willing to do it for less than the $700,000 guarantee to Kasparov. OTOH, it would be more difficult to find a sponsor since Kramnik is hardly the household name that GK is.

Deep Blue vs. Junior

[po8]

Deep Blue was dismantled several years ago.

Re:Deep Blue vs. Junior

[Anonymous Coward]

what? why? how? huh? link? funny? haha?

shocking claim needs more information.

Re:Deep Blue vs. Junior

[po8]

Deep Blue was dismantled in 2001, and part of it donated to the Smithsonian in 2002.

man vs. machine

[tq_at_sju]

kasporov should put a special stipulation that after the match the computer has to explain how well it understands the experience of playing chess, when it answers this with believablity then we're all in trouble hehe.

Deep Junior might be good at chess...

[doeth]

but as a friend once told me (quoted from somewhere?), even I can beat it at checkers.

thank you, mister obvious

[goatasaur]

The fact that Kirk always beat Spock at chess is/was a metaphor for the dominance of human ingenuity over cold logic.

I was attempting to make an insightful parallel using a motif that is prevalent in science fiction (the ingenuity/logic one I mentioned five seconds ago, if you've forgotten).

I'm not sure why it got modded as "funny".

Re:thank you, mister obvious

[JohnFluxx]

Talking about that...
I was really annoyed by that episode where data played that guy at some logic game.

Data lost, and couldn't understand why. He (rightfully) came to the conclusion that it was a fault with his logic systems.

However the crew viewed this as 'sulking'. (What do you call it when you project human emotions on to other things, when the human emotions don't really exist?)

Data _should_- have either won or known that the game was one of chance, so he had a chance of winning, or known that it was too complex for him to analyse the whole situation and so could only give a best-effort try anyway.

There's quite a few episodes that I'm angry at like that.. :)

Re:thank you, mister obvious

[The Smith]

What do you call it when you project human emotions on to other things, when the human emotions don't really exist?

Anthropomorphism.

Re:thank you, mister obvious

[Anonymous Coward]

"If you're going to mod me down, have the decency to reply to my post. I don't type to exercise my fucking fingers."

You can't mod AND post in the same discussion, though. And the reason why you type is irrelevant. People come here to read, reply and moderate. Thats the important thing. Your state of mind doesn't really come into it.

Re:thank you, mister obvious

[Anonymous Coward]

No, that refers to the physical side of things. For example, an anthropomorphic animal is one with human stature.

You're looking for 'personification', very common in poetry.

Re:thank you, mister obvious

[dmfallis]

Don't anthropomorphize Computers. They don't like it.

Notes from the Kasparov-Junior match

[migstradamus]

Always interesting how much interest this man vs machine stuff can still generate. My name is Mig Greengard and I'm doing the official live web commentary on the Kasparov-Deep Junior match and worked with both Kasparov and one of the Junior progammers (Shay Bushinsky) for over three years as the director of Kasparov's now-defunct website.

There was a good attendance and a great deal of media coverage today for game one, particularly considering it was a national holiday in the USA. (Well, almost.) Kasparov had the white pieces in game one, which is an advantage. (Interestingly, the Deep Junior team won the drawing of lots and could pick which color to have in game one (and 3 and 5), and chose to start with black.)

He completely dominated the game, it was a total stomp. He played 'real' chess instead of the dubious anti-computer style he used against Deep Blue in the 1997 match. Anti-computer chess involves trying to reach positions that computers don't play well instead of just making what you think are the best moves. Deep Blue showed that computers are pretty much beyond being vulnerable to these tricks nowadays, although every once in a while you'll see a strong program play like an idiot in a position it doesn't understand.

Kasparov prosecuted his advantage very quickly. In the press conference afterward he showed how much he had learned about playing computers. One key, he said, is that a computer doesn't understand results or practical chances, it only understands the evaluation of the current position. So instead of trying to swindle a way out of a bad position like a human Grandmaster would, by creating maximum chaos and hoping the other guy makes a mistake, a computer just tries to find the 'least-worst' move all the time. This is the only effective way for computers to play chess, but in inferior positions it often makes them look completely docile, if not pathetic.

He won't be able to do this in all six games, of course, and he'll probably lose one just because a human can't play error-free chess for so long against a strong opponent and computers punish errors ruthlessly. But game one showed he's prepared to the gills, as usual, and along with the fact that he's the strongest player in history should give him a decisive edge.

You can watch the games live with my commentary (and that of other commentators on-site as I relay their words) at many places on the web. Most of it is directed toward the level of the casual fan, not the chess expert. The company I'm working with, ChessBase, publishes Deep Junior and just about every other top chess program. (The program Fritz just drew an eight-game match against the world's #2 rated player and current world champion, Kramnik, in October 2001 in Bahrain. I was the webmaster and commentator on that match as well. I think I prefer the cold here at home in NY to the Bahraini humidity.)

As for the Deep Blue versus the current micros debate, that will be eternal as long as Deep Blue is in pieces. It was obviously much more powerful, but that doesn't mean it was a better chessplayer. We only have six games as evidence of its strength. They were good, but they weren't godlike and Kasparov said at the opening press conference that when you go over those games with Deep Junior it's clear that it plays better in just about every moment. (Except for two, which are the moves Kasparov has always suspected were the result of human interference. But that's another kettle of conspiracy.) Deep Blue was far, far ahead of its competitors in 1997, but computer chess programming has not stood still for the past six years.

It's also worth noting that what constitutes a huge advantage in computer-computer competition does not always translate into play against humans. A processing power advantage of just 10% between two identical programs will cause a lopsided score, but even a fourfold increase in processing power usually only means an extra 30-40 rating point gain against open competition. That is, one more win out of ten games.

I've spoken with Deep Blue's architect and other members of the IBM team on several occasions. Their egos are almost as big as Garry's! Hsu's book on the building of Deep Blue is almost as partisan as Kasparov's comments. They are both very competetive people. Personally I don't think there was any human interference in the DB match, but IBM's secretive and heavy-handed behavior needlesssly created a great deal of circumstantial evidence and suspicion.

You can follow my reports and photos on Kasparov-Deep Junior at ChessBase.com and I'll also be posting bits and ends at my site ChessNinja.com.

Re:Notes from the Kasparov-Junior match

[drinkypoo]

Thank you very much for your comments, they're the most informative so far today. I would like to comment on something you've said, however;

instead of trying to swindle a way out of a bad position like a human Grandmaster would, by creating maximum chaos and hoping the other guy makes a mistake, a computer just tries to find the 'least-worst' move all the time. This is the only effective way for computers to play chess, but in inferior positions it often makes them look completely docile, if not pathetic

I would argue that this is exactly what you have to teach a chess-playing program to do if you want it to be successful against the best well-prepared human players on their good days at least half the time, teach it to bluff. The hard part appears (IANAArtificial Intelligence Researcher) to be detecting when the program is confused and not just making a decision that looks like its best move. I should really read up on this before I comment, but, too tired. zzz.

Re:Notes from the Kasparov-Junior match

[Textbook Error]

Hsu's book on the building of Deep Blue is almost as partisan as Kasparov's comments.

Assuming this is the book you mean, I'd have to disagree. I read this over the holidays, and thought Hsu went out of his way to attempt to be impartial.

He obviously had a vested interest (as do you), but I didn't feel his book was in any way partisan - he wanted to win, but he was perfectly capable of dealing with the inevitable losses. As he's one of the participants, you have to take the comments about Kasparov's behaviour with a pinch of salt: but that's a very minor part of the book, and perfectly understandable given that it was an "I said/they said" situation.

It's a great book for finding out just how cobbled-together some of the early chess playing machines were - and that the kinds of problems they ran into along the way were incredibly mundane (fabrication problems, hardware failures, networks going down, last minute "this can't possibly hurt" changes to the code, etc). Although the book is pitched as being the story behind Deep Blue, a large chunk of it relates to the machines leading up to that point and the process by which Deep Blue came about (rather than that particular machine).

Re:Notes from the Kasparov-Junior match

[Anonymous Coward]

>As for the Deep Blue versus the current micros >debate, that will be eternal as long as Deep >Blue is in pieces It should be rather simple to simulate Deep Blue. Deep Junior's hardware evaluates 3 million positions per second. Deep Blue evaluated 100 million. Just get a copy of Deep Blue's source code, emulate the hardware, then give it enough time to go though 100 million positons for each move. It will be a slow but accurate emulation.

Instead of chess...

[BTWR]

They should pit THE WORLD'S FASTEST, GREATEST, 1000-TERA-TERA-BYTE RAM COMPUTER...

against any 5-year-old in CANDYLAND!!! I can't believe it took so many years to realize that there was absolutely NO WAY to be "good" at that game! All luck! :-)

Machine chess is stupid

[inviagrated_amnesiac]

It means nothing for the machine to lose or win. So why play it?

MOD UP!

[Anonymous Coward]

This guy has the most insight! Mod parent up!

yes it does

[djupedal]

Being better, in this case, means making less errors and being faster with the result. Of course if you mean each of you doing calcs with paper and pencil, I've yet to see a calculator hold a pencil very well at all, so you'd be a shoe in for that contest.

Otherwise, I'd say you'd come in a very distant second, even on your calculator's worst day, with corroded batteries, lights out and peanut butter on the keys :) I know my HPIIC can kick my ass around the block.

Just search trees

[godIsaDJ]

Indeed, quite to the point. Besides I'm confident that quite a few of us here tried to develop their own chess program (or perhaps just tic-tac-toe?). Once you know the general principles this toys are based upon, it becomes rather difficult to consider them anything else than quick searchers! Nothing new comes out of a machine...

DC is harmless

[Hubert_Shrump]

Guess I should take the alligator clips off of my... friend - and start rethinking my... line of questioning.

So, what does this mean?

[Millennium]

Garry Kasparov was beaten by Deep Blue. This means one of several possibilities:

1) Computers are more intelligent than humans.
2) Computers can be made to play better chess than humans.
3) Computers can be programmed to beat Garry Kasparov.
4) Chess can be reduced to a set of mathematical computations, which a computer can then perform faster than a human.

So what is it? And how do you know which one (or ones) are correct? Just a thought, since I think a lot of people are being overly alarmist.

Re:So, what does this mean?

[Anonymous Coward]

5. The rules were unfair and biased towards the machine. (Kasparov has claimed this, I don't know enough to determine if it's true, but it's possible.)
6. ???
7. Profit!

Re:So, what does this mean?

[Iamthefallen]

Kasparov claimed that the code for Deep Blue was written specifically for him, noone else, just to beat him. Which would mean that DB wouldn't have to analyze as many moves, it looks up what Kasparov has done in the past in the same situation and counters that.

Re:So, what does this mean?

[kavau]

1) Computers are more intelligent than humans.

No. If they were more intelligent than humans, they would also beat us at other games like Go, Bridge, Civilization... Humans still hold a large edge over computers in games where the player has a larger number of possible moves than in chess, or where the game involves psychology.

2) Computers can be made to play better chess than humans.

Yes. Certainly. Computers already beat most humans hands-down at chess, and it is just a matter of time before the strongest will falter. Chess is relatively easy for a computer since it is such a well-structured game.

3) Computers can be programmed to beat Garry Kasparov.

No. I don't think this factor plays a major role, at least not in this match. Kasparov was given several months to analyze Deep Junior, so he probably would have discovered any "exploits" aimed solely against his playing style. If Deep Junior can beat Kasparov, it can probably beat any other human being.

4) Chess can be reduced to a set of mathematical computations, which a computer can then perform faster than a human.

No. Chess is a "hard" problem in terms of complexity. It is impossible to calculate all variations until the end, to figure out the winner, and it will remain impossible at least for the next half-century (if Moore's Law holds...) The computer can look 10 or 15 moves ahead, but it also has to know how to evaluate the 4 billion or so positions it sees there. To do so, it has to use a form of "wisdom", an evaluation function, which is implemented by its programmers. To completely reduce chess to mathematical computation would mean to eliminate the need for such an evaluation function, and with our present knowledge, it cannot be done.

Re:So, what does this mean?

[drewness]

The code wasn't written "just to beat him", but they did feed it every recorded game of his to give it a feel for how he played, and he didn't have any advance knowledge of how it played.
Really, it's not all that exciting anyway. Chess isn't really a good benchmark for the progress of AI. The problem space may be big, but given enough power it can be brute-forced. One of the biggest things AI research has shown is that often times what is easy for humans (e.g. recognizing people's faces, understanding speech) is hard for computers and things that are hard for humans (e.g. calculus or chess) are relatively easy for computers.

He's right, you're retarded

[jpmorgan]

A chessboard is 8x8, meaning 64 spaces. However, each space can contain a pawn, a rook, a bishop, a knight, a king or a queen of either colour. The best estimate for the number of states the board can be in is 2.99x1041.

A naive encoding is 96 bytes per state. Let's say a tighter, or compressed encoding is 48 bytes per state. So a rough estimate as to the total storage space it would require is 1.44x1043 bytes.

In words, that's about 14 million billion billion gigabytes of data. I'm not going to say it'd be impossible to build such a storage mechanism in the forseable future, but I will say it's incredibly unlikely, and would be mindbogglingly expensive. And with modern technology, would require more matter than is actually on the planet. So no, dynamic programming wouldn't be useful in chess at all. Proving once again that if it were as simple as that, somebody would have thought of it already.

Out of interest, consider Go. This is a board where dynamic programming really would be useless. With around 10750 possible states, it would require significantly more atoms than are actually in the entire universe.

Re:He's right, you're retarded

[kasperd]

A chessboard is 8x8, meaning 64 spaces. However, each space can contain a pawn, a rook, a bishop, a knight, a king or a queen of either colour.

There can be no more than 32 pieces on the board at any time, and in fact there can be no more than 16 of one color. There can be no more than 2 kings on the board, and there are a lot of other limitiations.

The best estimate for the number of states the board can be in is 2.99x1041.

And that is supposed to mean what? If I multiply those two numbers I get 3112.6 possible states. If you intended to say 2.99*10^1041 you are far off. If you intended to say 2.99*10^41 I still think your estimate is too high.

A naive encoding is 96 bytes per state. Let's say a tighter, or compressed encoding is 48 bytes per state.

You are too pesimistic. Any encoding I could come up with would be less than 96 bytes, I honestly cannot say how you would be able to use that much. And I can do far better than 48 bytes to encode a state.

My first attempt to come up with an encoding would be just 183bits which is 23 bytes. Here is how I did:

1. 12 bits are needed to store position of the two kings.
2. 62 bits are needed to store information about which of the remaining 62 fields contains a piece.
3. 100 bits are needed to store information about which pieces are in the remaining fields. There can be at most 30 pieces left each of which has 10 different possibilities (5 of each color, empty fields and kings was already taken care of). A group of 3 pieces gives 1000 possibilities which can be encoded in 10 bits, 10 such groups gives 100 bits.
4. 1 bit is needed to know whos turn it is.
5. 4 bits are needed to know which rooks can still be used in a castling.
6. 1 bit is needed to know if en passant is possible.
7. 3 bits are needed to know which pawn can be used in en passant.

This number of bits gives an upper limit on the number of states, though that number is still far beyond the actual number of states because a lot of combinations are not possible, and some are redundant. The most interesting use of a compact encoding is to give an upper limit on the possible number of states, even if you want 100 bytes per state it is not that factor of 100 that is important, but rather the number of possible states.

Re:He's right, you're retarded

[radtea]

A chessboard is 8x8, meaning 64 spaces. However, each space can contain a pawn, a rook, a bishop, a knight, a king or a queen of either colour. The best estimate for the number of states the board can be in is 2.99x1041.

Not all of these states are reachable by legal moves, though. The number of reachable states is still vast, but far smaller than the number of total states, and far harder to calculate.

--Tom

Re:He's right, you're retarded

[alexo]

A chessboard is 8x8, meaning 64 spaces. However, each space can contain a pawn, a rook, a bishop, a knight, a king or a queen of either colour. The best estimate for the number of states the board can be in is 2.99x1041.

Not all of these states are reachable by legal moves, though. The number of reachable states is still vast, but far smaller than the number of total states, and far harder to calculate.

According to this article, The number of legal positions in chess is estimated to be 10^40, and the number of different legal games that can be played is estimated to be 10^120.

Re: He's right, you're retarded

[BigMattG]

So a rough estimate as to the total storage space it would require is 1.44x1043 bytes.

Or, to put that figure in a more human perspective, approximately the
aggregate capacity of all the AOL CDs you've thrown out since 1993.

Comment removed

[account_deleted]

Comment removed based on user account deletion

DC not Harmless

[Anonymous Coward]

although the levels of voltage that computers use (5V) are relatively harmless, DC is actually more harmful at voltages equal to the equivalent (root mean square) AC voltage. Just one of the many reasons we use AC instead of DC.

A lose-lose situation.

[Ilan Volow]

If the AI is winning, we look like a bunch of stupid apes.
If the AI is losing, it cheats and starts a nuclear exchange that destroys civilization.

We're screwed either way.

Re:where to view game replays, and watch live game

[metlin]

Mod parent up. Sometimes Slashdot story selections are funny to say the least.

Not to say the ridiculous titles :-/

~metlin

grammar nazi

[Anonymous Coward]

" despite evaluating less moves per minute than Deep Blue,"

It should read fewer moves.

Don't call it a comeback

[cgenman]

Computers and their programmers have been studying how human beings play chess for a significant percentage of the past century. For that matter, how many centuries have chess officionados studied the behaviors of human chess players? How long has the games played by computer chess machines been studied by the grand masters for weaknesses / strengths?

Somehow I doubt we have seen the best of the machines come out yet... nor do I think that given enough time a weakness could not be found in the optimal placement search tree algorithm for a computer that can only reach a limited distance in the tree. According to my trusty calculator there can be no more than 25,822,498,780,869,085,896,559,191,720,030,118,743 ,297,057,928,292,235,128,306,593,565,406,476,220,1 68,411,946,296,453,532,801,378,314,359,031,719,727 ,474,933,760,000,000,000,000,000,000,000,000,000,0 00,000,000,000,000,000,000,000,000,000,000,000,000 ,000,000,000,000,000,000,000,000,000,000,000,000,0 00,000,000,000,000,000,000,000,000,000,000,000,000 ,000,000,000,000,000,000,000,000,000,000,000,000,0 00,000,000,000,000 possible games, assuming 40 available moves per round over a 200 move game (the highest number of moves with a win is 192). Assuming my calculator's rounding and the error introduced through transcription isn't too aggregious, and each move is stored as a space - maximizing 6 bits + 1/2 float (which of 40 moves this represents, and a rough desirability float per computer move), you would still need a raid array of 1,775,296,791,184,746,553,884,443,075,207,066,360, 167,273,257,009,116,507,107,830,761,269,524,013,65 7,832,130,788,118,038,009,475,911,218,343,073,126, 390,169,600,000,000,000,000,000,000,000,000,000,00 0,000,000,000,000,000,000,000,000,000,000,000,000, 000,000,000,000,000,000,000,000,000,000,000,000,00 0,000,000,000,000,000,000,000,000,000,000,000,000, 000,000,000,000,000,000,000,000,000,000,000,000,00 0,000,000,000 120GB Seagate Barracudas to hold it all. Right now I doubt IBM has a working prototype of a raid array that can hold a measly trillion EIDE hard disks, let alone enough for the above application. That's a lot of room to wiggle. Even assuming that each move taken reduces the number of available following moves by one until a victory, that's still 18,698,058,574,400,000,000,000,000,000,000,000,000 hard drives. And that's with a grossly simplified, removed data structure that doesn't take into account the kind of size required for a pointer to data on a physical structure that generates 32,721,602,505.3 times the total power output of the sun as heat during standby. That's not even including the heat that the P5 would generate!

In short, I would give it another 20 years at least before we can declare a true winner in the battle between computer chess and grandmasters. And that, of course, entirely discounts the possibility of an IBM / MIT / Kaiser sponsored cyborg human player, at which point the whole debate will restart itself again and some equally outrageous headline will appear on Slashdot declaring the "last competition" between man and machine, a debate that dates back to when a monkey and a monkey with a stick sat around an ant hole and tried to find out who would starve to death first.

Computer played really poorly..

[Anonymous Coward]

I just walked through the game and really looks to me like the computer was set on 'Noob'. Some of the moves it makes just don't look good. Its position is screwed really early on. The computer got spanked so bad its not funny at all.

Something fishy here...

but can deep junior survive this challenge?

[circletimessquare]

but can deep junior survive this challenge?

What a silly topic heading...

[Domini]

Wow, slashdot articles with titles deserving of tabloid magazines?

It's more specifically a test between a slow heuristically based massively parralel computer and a fast serial rule-based weighted system. (simplified, yes I know.)

A computer can count faster then we can, but then we can build 3D representations of objects and spaces just by looking at them, and then traverse them effeciently (aka walking)

If it's games we want to make the battlefield, why not just toss chess and get a propper game... for instance Go. Computers still have some time to go before they can really compete on dan level...

This thread is absurd.

Re:What a silly topic heading...

[Dirtside]

Don't be an idiot. The title's an obvious parody of the kinds of news articles that show up whenever a computer beats a grandmaster at chess -- CNN blathers on about how computers are becoming intelligent, or some bullshit like that, and everyone who has a clue rolls their eyes and ignores it. Meanwhile, Joe Sixpack has his vague notions of a Terminator 2-like future reinforced by stupid reporting.

Not that /. is exactly a paragon of editorial quality, but there's no need to invoke the "Slashdot editors are trolls" subclass of karma whoring.

Re:What a silly topic heading...

[DeathPenguin]

Agreed--Chess sucks for this sort of thing. I'd like to see a version of Warcraft 3 where the computer doesn't have to cheat in order to win.

Re:What a silly topic heading...

[phantomfive]

Actually, the computer beats me in GO quite easily...

Re:What a silly topic heading...

[Domini]

Same here... :)

But that's because I'm still a 20 kyu player.

Computers cannot compete on dan level.

The difference is between tactics and stratagy (at at least depedning on that I think the difference betweens these two terma are...)

It's very subtle, small board have little strategy and lotsa tactics, and inverse for large boards. Computers are good at tactics since brute force seems to seem like tactics to people.

Strategy is more like a feeling though...

-grin-

Re:where to view game replays, and watch live game

[Anonymous Coward]

Someone wanna explain the logic behind Kaspy's move #7 (g4) ??

Game record?

[Nate Eldredge]

Does anyone know where I can find a record of the game (i.e. the list of moves made), ideally in plain text or PGN? The Wired site seems to be entirely Flash, which I can't handle.

Re:Game record?

[Nate Eldredge]

Found one. Thanks migstradamus.

Deep Junior really paused for thought at one point

[Anonymous Coward]

I just heard on World News Now that "DJ" took 25 minutes to respond to one of Kasparov's moves.

Kasparov has unfair advantage

[nekosej]

We shouldn't neglect the fact that Kasparov was given a copy of the program six months ago. Judging by the early errors by Deep Junior in game 1 (b5, the first non-book move was a blunder), I bet Kasparov has discovered several lines in which the computer falters. No doubt he will exploit them in the rest of the match.

I wouldn't be surprised if the Kasparov team actually used a brute-force method themselves to find these weaknesses, pitting other programs against Deep Junior in various openings to find out where D.J. loses the most.

Didn't we lose already?

[Maggot75]

Is humanity allowed an infinite number of rematches? That's cheating.
Besides, Garry Kasparov lost already, shouldn't we get someone else? Someone better at Human-vs-Machine chess?
And why are the machines using another computer? Adding insult to injury perhaps? 'Puny human, Deep Blue beat you already, try beating it's slower cousin, which, despite evaluating less moves per minute than I, is considered by you hue-mans to be a superior chess player !'

-1: Redundant

[Anonymous Coward]

According to the article, Deep Junior, despite evaluating less moves per minute than Deep Blue, is considered to be a superior chess player.

Anyone who has ever wrritten a game or has any clue about it knows that depth of search doesn't help much without a decent evaluation function.
For example, most reversi games can search more moves ahead than most chess games (because reversi is a smaller game). This certainly doesn't mean these reversi games play a good game of chess.

for my next trick...

[MegaFur]

As cgenman has already said (elsewhere in this thread level), this is ludicrously impossible. But even supposing that it were possible, I fail to see how the computer could really *avoid* defeat this way. Oh sure, in this strange, bizzaro scenario, the computer avoids being checkmated--but now people are going to dismantle it to find out "what went wrong". (classic melodramatic thing to do with a rouge computer that's killing people)

But then, what usually happens next is, the now self-aware computer builds itself a body (out of chess pieces I guess?), then starts knockin' down the humans like they're rag dolls. South Park had an episode sort of like this called "Trapper Keeper 2000" or something.

I'm sorry, I intended to refute the story properly, but it's just too silly. It sounds too much like the beginning of a (bad) sci-fi movie.

Humankind's Next Stand Against Machine

[ayjay29]

This is getting boaring.

I want to see "Mike Tysson vs Robo-Boxer". That's gotta be the ultimate test.

See who can build a boxing robot that can beat a heavy weight champion. The same weight restrictions, and approximate size would apply.

It's gotta be able to draw a bigger crowd, think of the add revenue.

man/machine test

[Anonymous Coward]

It seems to me that if you want to pit man versus machine you should pick something that is easy for a man to do.

They have that already. It's called the Turing Test.

s/n:r

Re:where to view game replays, and watch live game

[Kynde]

Someone wanna explain the logic behind Kaspy's move #7 (g4) ??

That really is interesting game. I'm no chess pro, but the g4 probable put pressure on the knight in f6. One thing I do know for sure is that with the level that these "guys" play in, it's extremely difficult to analyze their moves because their thoughts are so many moves ahead that it's impossible for me atleast to try to think of the alternatives move sequences.

Nonetheless it's nice to see how it's apparently ok to keep the c-bishob inside the pawns when playing queen's gambit. (i.e. playing e3 before moving c-bishop)

Re:where to view game replays, and watch live game

[Anonymous Coward]

I guess the point is that is looks like Black could take that pawn with impunity. Obviously, it is somehow a "poison pawn" -- I just don't see why Black can't grab it.

As for looking ahead. While that might be true, I suspect these early moves have been analyzed to the pont that the reason is well known.

why allow human intervention

[Anonymous Coward]

I seems to me a bit cheesy to classify the game as man vs. machine when the programmers of Deep Junior can change settings while the game is being played. Getting help from the outside is a great and unfair advantage, which is the reason why it was proposed to ban pit-to-car radios in F1, which afaik did not happen.

Why not close them into a room and see who wins? I am not particularly interested in how a human equipped with a monster computer can beat one without, but to see if the computer alone can beat the best human chess player.

---

humanz rulez

Competitive Magic the Gathering

[Kid_Korrupt]

There seems to be some debate in this forum about the merits of computer chess players and their brute force method. Some posters have brought up go as a 'real' challenge for computers. Although I haven't played go I would like to bring up another alternative: Magic the Gathering.

Now before I get scoffed at, and modded down I think the case for magic should be heard. And I am not talking about casual play with your latest dragon deck, but competitive magic. The WOTC and DCI support a fairly large, world wide, competitive player base, with prize support up to about $30,000. Now this doesn't compare to what chess masters can win but I find the similarities very interesting.

The thing in my mind that makes magic far more interesting and challenging than chess is that the game changes every 4 months. Based on some essential fundamentals the actual rules recieve a complete overhaul, and even top players that cannot adapt to the new format will find themselves sharing tables with the scrubs.

I think a real challenge for programmers would be able to make a program that could thrive in this type of environment. To me that would be true AI. Being able to actually LEARN and not brute force its way to a win would be an amazing accomplishment for AI programmers.

Re:Competitive Magic the Gathering

[Peyna]

and not brute force its way to a win

I highly doubt that Deep Junior merely brute forces his way through the game. He probably uses highly optimized heuristics to determine what may or may not be the best move to make. This is how he can evaluate less moves and perform better; if he brute forced every possible move all the way to the end of the game, he probably wouldn't be able to finish the game.

Re:where to view game replays, and watch live game

[herrd0kt0r]

i'm no chess expert, but drawing on what little i do know about chess, i'd hafta agree with both of your observations. i don't begin to approach the level of their chess play, but i'd surmise that yeah, the pawn was put forward for one of at least three reasons, if not all three:

1) it's kasparov's way of saying "j0, i gon fux0r with your brain junior!1!!1!! FEAR TEH PAWN! TAKE IT! TAKE IT! (muahaha, as akbar would say, "IT'S A TRAP!!!") so would ineptly waste an opening move to grab a pawn, even when it's already behind by one.

2) garry pushes the pawn forward, thinking, "KEKEKEKE I PUT TEH PAWN PRESSURE ON TEH KNIGHT! j00 HAVE NO ESCAPE MAKE YOUR TIME!1!!" thus making the knight feel uncomfortable enough for black to either a) move it, or b) fortify it.

3) kasparov decides that pushing the pawn is one of his "three best moves," and sees it adding just the right amount of pressure to capitalize upon later, but not so much as to require immediate attention from black.

i dunno. those are my guesses. any chess nuts have better insight?

Apologies to herrd0kt0r.

[mtm_king]

Notes to self:

read twice
think
think
post

1st move

[peu]

Next match, when computer plays first, how it decides in the best strategy?

Can't have mathematics without a definition...

[Anonymous Coward]

How can there be a mathematically best strategy for go when there isn't even a mathematically reliable definition of the game, much less of, say, territory, the score, or the end of the game?
Anyone who thinks there is, well, ... doesn't understand the situation yet, and should have a word with any of the half-dozen westerners who have spent serious time trying to program it (and/or read Ikeda's book on the subject.).

I don't know about gut instinct...

[orichter]

I suspect that gut instinct gets human chess players into trouble more often than it helps. It's more pattern recognition (the same sort of skills used in face recognition), combined with an understanding of how computers play chess. I read an interview with him once in which he explained that he plays computers very differently than he plays humans. He actively avoids gut instincts since they usually rely on "spooking" the opponent into making a mistake. Computers don't make mistakes, they simply follow the rules provided to them. Kasparov simply plays chess to never allow a computer to get into a position it can recognize or attack. It's almost as if he plays for a standoff, attacking very patiently, and in a very non-directed fashion so the computer can't pin down what he's trying to do. It's the same thing most of us do when we beat AI in our favorite strategy game. We find the one or two stupid things the computer does, and use them over and over until the computer is defeated. How many times have you had a new game which seemed difficult until you figured out that the computer never builds airplanes, or always attacks the last city attacked. After you find this out, winning becomes trivial. Kasparov is essentially doing the same thing, except in his case, there are very few flaws, and exploiting them is not trivial.

RIOT!

[Sebastopol]

Well, at least if Kasparov loses, I won't have to worry about fans rioting and burning the town, like in Oakland last night. That is, unless chess becomes the national past time of the lowest strata of society.

GNUChess

[DougSuerich]

You are too pesimistic. Any encoding I could come up with would be less than 96 bytes, I honestly cannot say how you would be able to use that much.

If you look at the source code for GNUChess, you can see that this is effectively how they store the board.

IIRC: Start with an 8 byte array, thus 64 bits (= number of squares on the board).

Create one of these arrays for each type of piece: White King, Black King, White Queen...etc. In the array, flip the bit to one if that piece is there, and zero if it is not.

There are 12 possible pieces, and thus 12 * 8 = 96 bytes. Obviously this doesn't account for the extra information of ability to castle and whatnot, but it gives you the basics.

The advantage to this datastructure is that it makes things really easy to work with. To get the entire board, you can just OR all the bit masks together. To check for what pieces you might attack, just AND it against another array that represents where you can move. Bit operations are cool once you get the hang of them.

Your structure is likely efficient, but a real PITA to work with.

Re:GNUChess

[kasperd]

There are 12 possible pieces, and thus 12 * 8 = 96 bytes.

OK, I see that it is possible to encode the board in so many bits. And with the litle extra information that is needed, you might even be able to use 100 bytes. Anyway with this encoding most of the possible bitpatterns represent impossible configurations.

Your structure is likely efficient, but a real PITA to work with.

It is compact, it is not interesting for use in actual implementations, more interesting is how compact you can do it, because that gives an upper limit on the number of states.

Where is the Game?

[Vryl]

Anyone know where I can get the game, sans browser plugin etc. You know, in good old ascii?

Re:Where is the Game?

[Vryl]

Well, I found it:

http://www.chessbase.com/games/2003/x3d1.htm

20 years ago it was good test

[goombah99]

Your response to my post osrt of defines the essential point. 50 years ago a chess playing computer that could play well, regardless of winning, was a good test of AI. it tooks lots of reasoning to overcome the limits of processing power. These days its not a good test anymore because the problem can be brute forced rather than reasoned, and as I pointed out, the problem itself is hard for humans too.

thus AI needs to be a moving target, with each a successive goal to be picked as it becomes feasible. Each being closer to the thing humans are good at. to propose soccer 50 years ago would be absurd. now its not so absurd. Someday, having it steal your girlfreind may not be absurd.

I have the utmost respect for Kasparov

[Anonymous Coward]

He can think in so many directions at once that it completely amazes me. That he can out think computers is simply amazing.

I wish him the best of luck in his battles.

Rise of the Machines

[Anonymous Coward]

Honestly, from the preview I saw before the last LOTR movie I think the whole franchise should be taken straight-to-video or canned.

Where's the signature blue night, or James Cameron for that matter? There isn't a single 'money shot' in the whole preview- we saw hks and t-800s on a battlefield in the last one. The comic books had female terminators a decade ago- the concept really can't carry the movie.

I think we still have the upper hand

[indros13]

Humankind Makes Last Stand Against Machine

With the computer close to ultimate victory over humankind, Johnson reaches for the black cord and with the gently sound of fans spinning down, humanity is saved.

Like deja vu all over again...

[xmark]

Both the original article and the parent above remind me of one of the better Twilight Zone episodes. "Steel" takes place in a future where boxing between humans has been outlawed, and replaced by boxing between robots. Lee Marvin plays a down-on-his-luck former chapmpion boxer who is now the manager of the robot Battling Maxo. After booking Maxo in prize match against a latest-model fighter bot, he discovers Maxo needs an expensive repair. Marvin is broke and can't afford it. Desperate for the prize money, Marvin replaces Maxo with a different robot -- himself in disguise -- and goes into the ring. He gets his ass kicked while the crowd jeers at him, calling him a piece of junk, never realizing they're watching the only display of real courage they'll ever again see in an era of mechanized fighting. Reminds me of the Hemingway quote: "A man can be defeated but not destroyed."

Re:the only people claiming deep fritz deep blue

[JustAnotherReader]

The difference, according to Steve Lopez's T-Notes articles (Free registration required), is that Fritz 6, 7, and 8 as well as all the other ChessBase engines are now vastly stronger at positional play. This means that while they may not look as many plys (half moves) ahead in the same amount of time, they are able to see non-tactical advantages better. After Kramnick's match he said "You use to be able to play "anti-computer chess." But not any more. These computers are playing real chess".

In Kasparov's match against Deep Blue Garry thought that there was human intervention on 2 moves because the computer ignored a tactical advantage, which would have lead to a draw, for a positional advantage which lead to a win. The modern programs would be more likely to see that kind of move.

Last stand (say it ain't so)...?

[cr0sh]

I find this rather interesting (though I know this post is late). If you find this interesting as well, and want to know the long and sordid history behind chess playing machines, where it all seems to come from, and how the issues and ideas brought forth by such "performances" have been talked about for a very long time - you owe it to yourself to read this book:

The Turk by Tom Standage

A very enjoyable, and informative read, I must say...

Credit where credit is due

[Anonymous Coward]

"I, for one, welcome our new insect overlords. I'd like to remind them that as a trusted TV personality, I can be helpful in rounding up others to toil in their underground sugar caves."

Looks familiar? The Simpsons' Kent Brockman in "Deep Space Homer" ...

Wouldn't post as an AC if I wasn't busy modding your ass down for trying to pass on a modification of the best Simpsons quote ever as if it was your own invention.

Re:Credit where credit is due

[Anonymous Coward]

You dumbfuck. Everyone knows it's from the Simpsons. He's not attributing it because everyone's seen that episode a dozen times, and he knows it.

Re:Credit where credit is due

[Anonymous Coward]

No need to be rude, cockhead. "Everybody" doesn't know. "Everybody" doesn't even watch the Simpsons, and sure as fuck doesn't remember all quotes from it.

And why don't you drop the "he" already? I know you posted that shit you lame anonymous asshole.

Re:almost NO

[Anonymous Coward]

I would love to login, but nowhere to find my password, anyway, I tend to agree with you but only half the way.
No doubt they both gathered information about playing chess from various sources.
But, and that's a pretty big but (...) Kasparov is not a single black box, in the sense that Deep Junior will never step off the table and yell
- I'm fed up of this chess business, I allways long to be a farmer!
You are just throwing free will (and self conscientiousness) in the bin.

Re:the only people claiming deep fritz deep blue

[dmforcier]

The difference ... is that Fritz 6, 7, and 8 as well as all the other ChessBase engines are now vastly stronger at positional play

Well if they are, Game 1 certainly didn't show it! It looked like Fritz... er... "Deep Junior" was playing completely on predictive tactics the whole time, trying to oppose all possibilities rather than recognizing the strategy. (Or always assuming that the strategy is a fortress assault.) How else do you explain the rook sac, or the knights shoved in a corner and left to oppose an attack that wasn't forming?

I predict that Garry will use the same technique in upcoming games. Make a feint against the king's fort, then clean up the back half while Deep Fritz worries about it.

Kasparov has already won...

[hyacinthus]

One can, with enormous effort and a bit of cheating (didn't a team of engineers desperately make alterations to Deep Thought _during_ its infamous match against Kasparov?), build a computer which can acceptably fake a man's ability to play chess--but no effort will ever make the computer that _wants_ to challenge a man to a game of chess. The programmer still has to do that.

Incidentally, when are they ever going to put one of these heavily-publicized chess-playing machines in a real tournament where it would have to play a wide variety of real chess players, instead of a specially conceived vanity match against a handpicked challenger, like Kasparov or Kramnik?

hyacinthus.

Theory vs. reality

[Chazmati]

Touche. You're right, it's not infinite. But with the number of chess positions greater than the number of atoms in the universe, it's clearly impossible today, right? But you say that some future computer will SURELY be able to do it? Man, that's gutsy. Quantum computer, you think?

I see what you're saying, and I agree that it's finite. It's just such a staggeringly large problem; I'll stand by the idea that "theoretically possible" does not mean "will definitely happen someday."

Re:almost NO

[Anonymous Coward]

Deep Junior will never step off the table and yell - I'm fed up of this chess business, I allways long to be a farmer!

Actually, I'm sure Deep Junior is very capable of resigning a game when conditions are right (ie, bad). As for emotional outbursts that go along with a resignation, well, those can be programmed or even stitched together from component phrases (think Zippy the Pinhead). Anything said beyond "I Resign" is beyond the scope of the rules of chess, and therefore has little to do with the player in terms of being a chess competitor (which is what the discussion was about).

Re:Take it Easy!

[sciire]

Take it Easy, Alan Turing, yes him, was a patzer (Really Bad) player, and as far as we know, he was probably smarter than Kasparov. Im smarter than him too, chess isnt a intelligence task.

Re:almost NO

[Anonymous Coward]

(which is what the discussion was about).

Sure, I wasn't making a point on the discussion, but rather on a single comment.


Actually, I'm sure Deep Junior is very capable of resigning a game when conditions are right (ie, bad). As for emotional outbursts that go along with a resignation, well, those can be programmed or even stitched together from component phrases (think Zippy the Pinhead).


I may sound harsh, but please the day you suceed in programing self conscientiousness let the whole world know! That is light years away from a simulation of emotion (think, I can't let you do that Dave).

Anything said beyond "I Resign" is beyond the scope of the rules of chess, and therefore has little to do with the player in terms of being a chess competitor

The point I put is, the poster was restraining Kasparov to a box, which could make sense if you analise only the match and not what is heapenning for real, Kasparov knows why he plays chess Junior is not even close of grasp what is "to know".

Re:where to view game replays, and watch live game

[Kynde]

Actually I found a site a little after posting that where there was some other russian chess expert's comments on the game aswell as those of Kasparov's himself. Delightful read, I just can't find that link anymore.

If I remember correctly, that naturally (given that it's only move 7 or so, and these guys play from 10 to 15 moves from the opening book) was part of an opening line that just hadn't been played in tournaments for a while. I guess Garri chose that because it's relatively aggressive given it's a closed opening and the computer just might not be at it's best in it.

But most importantly I think the computer were prepared for Kasparov to begin with E4, rather than naything else, so to begin with the queen's gambit was in any case a marvellous move from him...

No wait... here it is : http://www.worldchessrating.com/

Read more from there...

Re:almost NO

[Anonymous Coward]

Kwitcherbitchen, you two. Fortunately, this is a chess competition and not a consciousness competition. The latter would be awfully boring. The former promises to be a close matchup. That's the nature of interesting competition. Bring it on.

I thought the same..

[JThundley]

I thought the same exact thought as soon as I read the headline. Right on.

Second match - Draw

[Anonymous Coward]

The second match was a draw - Kasparov winning at first, made some sloppy moves and got into a position that favoured the computer. However he offered a draw, which got accepted after a repeated move.

IBM Cheated

[arfy]

I've read that book and one other on the Kasparov / Deep Blue publicity stunt and oome to the following conclusions:

1. IBM behaved badly because they were desperate for a win.

2. Kasparov didn't behave much better and didn't know enough about their behavior to mount a decent argument against them. He knew he'd been cheated, just not how to put it.

3. IBM got their publicity and dared not rematch under properly controlled conditions for fear of a loss. Hence the marketing spin of Kasparov's bad faith acting, which isn't too hard to believe but hey, it's marketing people on the other side so the brunt of evidence stacks against them.

4. Finally: IBM did cheat. Meatspace players can go home and study before the next game, but not swap their whole heads out for new models which is the rough equivalent of what the IBM team did. Kasparov had a point in what he said, he just made it in such a poor way that he got little sympathy. Besides, a lot of players wanted to see him get his comeuppance and sympathy was in short supply, judging from some of the comments I read and heard back then. But that doesn't change that what the IBM team went beyond the pale.