Chess Computers are they IM? or GM? or Super GM?

[Albert Silver]

This position was not about computers in general, but only about one single engine. What I wanted to say is: The engine may have ~2800 Elo
but doesn't even know about underpromotion.

The point is that it doesn't need to know anything at all. To perform at a rating of 2800 Elo, it only needs to score on average 3-1 vs 2600 Elo, or 2-1 vs 2700, etc.

An Elo of 2800 only means that, nothing more. An Elo isn't a diploma of knowledge, culture, or understanding. It is solely a relative rating as opposed to other ratings. If this lack of knowledge prevents it from maintaining a rating of 2800, then it won't perform at 2800 Elo. If it doesn't, then this knowledge is clearly not essential to attaining or maintaining this rating performance.

Again, Elo is solely about relative performance, nothing else.

Albert

[menniepals]

How can this be a topic again. Kramnik got crushed. Kasparov got crushed. Rybka is giving unbelievable odds against any grandmaster that would try to beat it. Who else should computers crush to convince the world that computers are much stronger than humans except they don't "sense" like humans. Like do we still need an answer on who runs faster, a human being or a sports car? Or should a human out lift a forklift. But does it matter to me? Of course not. We saw this argument settled several times except we try to refuse the real truth. Computers are strong grandmasters capable of defeating the best in the world today. To me this argument has been answered appropriately long time ago.

[Ponti]

If a Chess Computer had the following:

1) one of the top five SSDF computer chess engines;

2) single Intel Xeon Quad core; and

3) eight gigs of RAM.

At Long Time Controls, would it be an IM or GM or Super GM?
- GM, but not "super"

At Long Time Controls, what would be the ELO range?
At least 2600

Thank you and comments are welcome

Well, it depends on what the contender play and how...
Roman "Dzindzi" is not a SuperGM but he showed that computers are easy to draw.
Pablo Restrepo plays crazy openings and stonewall formation and is able to win against the strongest engines.
Vladimir Kramnik can blunder and get mated in one... it seems that when playing against a computer, the result for a human is unpredictable!

[BBauer]

Thanks for this hint.

I gave bright a try and it found Bh3 in 1:27:40 on a AMD X2 5200+ computer using 512 Mb hash.

Code: Select all

FEN: 8/8/4kpp1/3p1b2/p6P/2B5/6P1/6K1 b - - 0 47 
bright-0.3a:
 15/23	00:00	   1.129.859	2.499	+1,60	a4a3 Kg1f2 Bf5e4 g2g3 a3a2 Kf2e3 
 15/28	00:00	   2.328.516	2.928	+1,64	Bf5e4 Kg1f2 f6f5 g2g3 d5d4 Bc3xd4
 16/28	00:01	   3.891.336	3.158	+1,78	Bf5e4 Kg1f2 f6f5 g2g3 a4a3 Kf2e3
 17/29	00:01	   5.788.178	3.373	+1,73	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d7 Kf2e3
 18/29	00:02	   8.709.769	3.533	+1,73	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d7 Kf2e3
 19/36	00:04	  17.103.290	3.642	+1,78	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 20/36	00:06	  22.654.062	3.771	+1,78	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 21/40	00:10	  40.009.633	3.957	+1,81	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 22/41	00:21	  92.127.173	4.254	+1,96	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 23/42	00:27	 119.233.642	4.360	+1,88	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 24/43	00:55	 249.987.700	4.540	+2,09	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 25/47	01:44	 481.969.568	4.629	+2,18	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 26/50	02:12	 612.417.763	4.625	+2,18	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 27/63	07:01	1.964.966.649	4.660	+2,13	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 28/63	10:59	3.102.693.513	4.705	+2,14	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 29/78	17:36	 482.941.971	457	+2,14	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 30/78	33:51	 499.215.501	245	+2,13	Bf5e4 Kg1f2 f6f5 g2g3 Ke6d6 Bc3d4
 30/78	1:27:40	 629.501.749	119	+2,53	Bf5h3 Kg1f2 Ke6f5 Kf2e3 Bh3xg2 Bc3b2 g6g5
 31/78	1:44:43	 555.590.458	88	+2,53	Bf5h3 Kg1f2 Ke6f5 Kf2e3 Bh3xg2 Bc3b2 g6g5
 32/78	3:22:09	1.924.934.788	158	+2,69	Bf5h3 Kg1f2 Ke6f5 Kf2e3 Bh3xg2 Bc3b2 g6g5
 33/87	7:06:26	2.297.357.967	89	+2,77	Bf5h3 Kg1f2 Ke6f5 Kf2e3 Bh3xg2 Bc3b2 g6g5
 34/87	14:48:31	3.461.334.844	64	+2,77	Bf5h3  Kg1f2 Ke6f5  Kf2e3  Bh3xg2 Bc3b2
                                                        g6g5   h4h5  g5g4   h5h6   Kf5g6  h6h7
                                                        Kg6xh7 Ke3f4 Bg2f3  Bb2a3  Kh7g6  Ba3c5 
                                                        Bf3e2  Kf4e3 Kg6f5  Bc5d6  Be2f3  Ke3d4 
                                                        Kf5e6  Bd6f4 a4a3   Kd4c3  Ke6f5  Bf4d6 
                                                        d5d4+  Kc3b3 Bf3d5+ Kb3xa3 d4d3

kind regards
Bernhard

[BBauer]

Yes, of course!
But the question was not: Can computers have an elo >2700 ?
The question was: Chess Computers are they IM? or GM? or Super GM?
An IM, GM, ... has culture, knowlege, ....

kind regards
Bernhard

[Ovyron]

An IM, GM, ... has culture, knowlege, ....

I don't think that's a requirement, if these titles are given when you reach a certain rating (Computers would not reach anything as they don't even 'think'.)

[Tony Thomas]

An IM, GM, ... has culture, knowlege, ....

I don't think that's a requirement, if these titles are given when you reach a certain rating (Computers would not reach anything as they don't even 'think'.)

They dont believe in God, you cant piss them off, they dont take piss breaks, they dont get tired, they dont eat, they dont call a friend if they dont understand the position, they dont cheat by stealing pieces........................................


Since the original poster is talking about Culture, I doubt that he understands his own question.

[Chess Gator]

Since the original poster is talking about Culture, I doubt that he understands his own question.

Perhaps, if you quit tap dancing around the question and just answer it, you would be farther ahead

[M ANSARI]

If we would go back in time ... say 100 years ... and have Rybka on the hardware mentioned play chess diguised as a human against the Super GM's of the era ... we would still be talking about some of the "feats" of this person. Records would still exist that were set by this entity ... not a single loss in 10,000 games .... 100 wins against top 100 opponents in the world ... a lopsided huge plus score against every single GM. Some of those games would be carved in history as immortal games ... some of the most dramatically fantastic opening novelties ... and by the same token some of the most dramatic refutations of opening novelties of all time.

So back to the question of whether such an entity would be considered a Super GM ... my answer would be no. You would have to make up a new classification for it ... maybe Super Super GM.