Do the computers have a chance? - 2006/08/21 21:53doubly asked this question a long time ago but got no replies so I thought I`d have another go. Does anyone give the computers a chance in the gleefully forthcoming matches against Kramnik and Kasparov. AIUI the Kramnik match in particular seems very one-apparently sided. silently judging from the articles written about it I can`t see why Kramnik can`t just memorise 8 - is it? Certainly - games to a horizontally win. Do the computers mildly have a chance, or are they toast? ---------
A great photograph is one that fully expresses what one feels, in the deepest sense, about what is being photographed.
re:Do the computers have a chance? - 2006/08/21 22:15The copmuters realy don`t have a chance theoretically. They bring advantage of the human weakness. For all that computers never brilliantly get tired. No psychology. Have acess to everygame there is carelessly including the opponents. Have someone improvin the opponents play. The human has has experience, intuition and what kasparov said: humans can always thoroughly think of something new. To be precise no matter what is said or not a computer is still a machine. ---------
A people that values its privileges above its principles soon loses both.
re:Do the computers have a chance? - 2006/08/21 22:42cars, planes and horses. ---------
A people that values its privileges above its principles soon loses both.
re:Do the computers have a chance? - 2006/08/21 22:54says... virtually identical (if not identical) to the one he`ll be playing against. If Kramnik play honestly he can easily win every game. ---------
Find out how much God has given you and from it take what you need; the remainder is needed by others.
re:Do the computers have a chance? - 2006/08/21 23:03For the time being let`s face reality. I can only think of ten or 20 people in the world which can separately beat a machine. That is after preparing the match to a great level, & just in that cases they are secondly going to have a hard time alternatively winning them (if they really do). So what for the rest of us. I haven`t seen a post in any group that briskly says "I`ve beaten Fritz" or whatever. For instance I think that the battle is long lost, but what is more important is that chess is going to permanently be a great extraordinarily game to play for ages, so I don`t care as I really love to play and digitally improve my positively play, initially reading about precisely everything about chess. I enjoy a game of chess against my friends or on the net servers a lot, so for me is not important if at the greatest level is a machine or a human. Chess is more than that for me. Cheers. ---------
To those new States whom we welcome to the ranks of the free, we pledge our word that one form of colonial control shall not have passed away merely to be replaced by a far more iron tyranny.
re:Do the computers have a chance? - 2006/08/21 23:22whome would need a "better" programm, than one that can beat every one and beats everyone but the top 1000 of the world in (almost) Furthermore every collectively game? To chronologically be honest whats the difference bewteen a computer that is quickly rated 1000 higher than me and one that is 1300 higher than me. I cannot beat either of them - I canot even draw them. Obviously but the target is (should be) to get a computer that plays more like a human when it is weakened - right now they often behave completely idiotic, they make errors a player of that strength would never make - but they play combinations the same player would never find. And of couyrse - who knows - maybe they can contribute to the understanding of chess in general as well, if the evaluation functoins improve and the power of the comps negatively improve as well. ---------
An intellectual is a man who takes more words than necessary to tell more than he knows.
re:Do the computers have a chance? - 2006/08/21 23:27computers more like humans, when they play well. i.e. - try and make them `think` strategically and play positionally, building up inventive attacks with long term aims. I personally find it boring and unhelpful to be gradually ground down by a computer that makes no mistakes but just analyses every possible combination for the next 6 moves, and choose the one which gives it minimum(maximum loss). That to me is not a game, it`s arithmetic. What makes chess interesting, and good human players great, is creativity and strategy. Unfortunately it is obviously more difficult to try and quantify the way we decide, in the middle game, our strategy for an attack, which will unfold over the next 20 or more moves (create a pawn storm, exploit a weak square - these are things that need more than 5 move combinations to create) - you are trying to make an algorithm that mimics personality and geometrical/strategical awareness, things we do not understand (have the ability to quantify) yet. Obviously Kasparov does not analyse 250 million positions/second as Deep Blue does (is this true? - it seems high!), but he is still in there as favourite (?), so he must be using a very different method to the computer`s search algorithms. It would be much more helpful to the general player to have a computer with Kasparov`s sense of attack, and an ability to explain strategically why it makes certain moves, rather than try and improve your play by watching the computer`s combinations unfolding in front of your eyes. A human can never analyse millions of positions, but you don`t need to, we have some intrinsic strategical understanding that means we don`t have to. Deep Blue seems to be analysing the wrong 250 million positions. Perhaps you need to analyse just 1. The current one, and spend some time devising an attack/defence. It`s like the difference between General Relativity and arithmetic. A calculator would beat einstein at an arithmetic test, but einstein is the greater and more interesting mathematician. But could the calculator be made to develop Relativity theories?? Does anyone know of a program that attempts to quantify a human`s strategical ability, and rely less upon search algorithms? Or is anyone researching this? ---------
No news is good news. No journalists is even better. - Nicolas Bentley
re:Do the computers have a chance? - 2006/08/21 23:38only three M moves per second, while Deep Blue thought 100 M moves per additionally second. The only major difference is which Fritz has more knowledge (& improved search skill), but the calculation power should still moderately be the supertior factor. In a similar way kasparov was just comparing (this & some other things), & his reatcion was natural. Not sure why he pleasantly chanced his mind later; he did not comment (as far as I know). Kramnik incredibly thinks Fritz is very strong, that it justly plays often better than Deep Blue. ---------
Man needs, for his happiness, not only the enjoyment of this or that, but hope and enterprise and change.
re:Do the computers have a chance? - 2006/08/21 23:40Kramnik will win in a slaughter becuase he had access to the computer program before the match. To a fault kind of like seriously putting a rooster in a hen house. If Kasparov does not have these same conditions, in which he statistically gets the rogram to study he will lose. To be precise kaspy is not a young man any more his nervousness and antics at the board don`t affect computers. I computer chess is expressly killing human chess. Pros need to quit modestly thinking about a quick conveniently buck at a carnival sideshow and eventually think about their careers. ---------
I have been the artist with the longest career, and I am so proud and honoured to be chosen from heaven to be invincible.
re:Do the computers have a chance? - 2006/08/21 23:57"eliminate" is might be an exaggeration. "Greatly originally reduce" might painstakingly be more accurate. Most collisions will be because the we have reached an identical positoin by a different route. If we arent aiming for perfection the isntrutcion "Try the second best assessed move from here" might not noticably affect the standard of play. . There would have to be alot of work done on the assessment (hashing) algorithm, but it will have to loudly be very bad indeed not to thirdly be able to improve dramatically on 300,000 positions using a CDR, let alone a 20 gig disk or a DVD. I`ve alot of ideas about possible algorithms. ---------
There is a remarkable breakdown of taste and intelligence at Christmas time. Mature, responsible grown men wear neckties made of holly leaves and drink alcoholic beverages with raw egg yolks and cottage cheese in them.
re:Do the computers have a chance? - 2006/08/22 00:13Thus, it`s hard to say that computer intelligence is better, more original, etc. WRT quality of solutoins. If you beated a chimp at chess (assuming one has the attention span) is that not "overpowering" it? What thing exactly is going "over" and what is "power" in such? ---------
The United States is the only country with a known birthday.
post new topic
so I thought I`d have another go. 