Author: Robert Hyatt
Date: 11:30:23 07/21/02
Go up one level in this thread
On July 21, 2002 at 09:29:29, Rolf Tueschen wrote: >New exposition: >################# > >On July 20, 2002 at 22:16:48, Robert Hyatt wrote: > > >About control and safety and American logic: >############################################# > >>We can argue for all time about "how many angels can dance on the head of a >>pin?" But since there is absolutely no way to measure the size of an angel, >>no answer can be proven correct. Discussion is pointless. >> >>In chess, we have the same problem. It is simply impossible to prevent >>outside intervention in a game, and it is also impossible to detect that >>it happened in every case. If you can't prevent it in 100% of the cases, >>and you can't detect when it has happened with 100% reliability, The >>"interest" in the issue wanes a bit... >> >>Should we argue about how long it will take man to reach Alpha Centauri? I >>claim zero time. And I leave it to you to prove me wrong by showing with >>absolute perfection that exceeding C (speed of light) is impossible under >>_all_ conditions... >> >>Many discussions are pointless. Trying to "interfere-proof" the DB matches >>falls into that class... > >This is all interesting and true but it's also totally wrong for our problem in >computerchess. Who did want to prove something? The answer has been the same for 30+ years. "Can a computer beat the human world champion in a match of six games at standard tournament time controls?" That has been the _only_ question. It was answered in 1997. If you want to re-pose the question in a different way, feel free. Perhaps you can also get someone to answer it. But don't make the question _you_ want to ask into the question _they_ wanted to ask. They _had_ their question, and _they_ produced an answer... > Here we have the first >historical lie. From the 40's on computer scientists tried to prove what they >could achieve. Now this is ridiculous. We have computerchess experts who try to >prove that their creations could play chess. But JUNIOR became CCCWch in Paris >without knowing something about the underpromotion! Unbelievable! Why? I saw a program win a game without knowing _how_ to promote in any case, not just under-promotions. It was at one of the very early ACM computer chess events, it pushed two pawns to the 8th, left them as pawns, and _still_ was able to win the game... >Then DB2 was >presented without any attempts to prove the authenticity of the output and more, >the machine was already sold, so that any attempts to do some research were >impossible by definition. Can you Americans spell 'David Copperfield' and >'delusion'? Can you spell obsession? Obfuscation? falsification? > Is it true that you believe in American logic with a complete >reversal of traditional logic? Not you must prove something you pretended to do >but the world must prove that you did something wrong? Actually, _yes_. That is why I choose to live in the USA, where I am "innocent until proven guilty." That is the basis of our legal system and I happen to like it. >And that in a situation >where you guaranteed that no research could be started? God bless America! And >forget about the rest of the world? This is kind of strange logic. It is a logic >of hypocrisy and fiction. And _your_ logic would be a logic of what, exactly? The logic of changing the goal at a whim? The logic of trying to do something that is totally impossible. The logic of _attempting_ to do something that was totally impossible, so that even the attempt would have made the match so expensive it would never have happened? Did you complain after the 1996 match? Why or why not? (I know the answer of course, you did _not_.) If you didn't complain after 1996, then why complain after 1997 when _everything_ was the same, match-condition-wise? I also know the answer to that... > >But isn't it more a logic of political power? Because there is no American >science, there is no American logic! You are simply proclaiming false logic and >because you have the power and the money you think you have the right to define >what is the truth and what is logic? There is no "false logic". The original question, as posed in 1960 or so was "can a computer beat the world champion at standard time controls in a match?" IBM answered that clearly in 1997. "yes". Nothing else is particularly important. Yes it would be nice if DB2 were still around so we could study it to understand it better. But it _did_ answer the question asked, in a scientifically rigorous fashion, that _almost_ everyone has accepted... At least I don't believe pigs can fly. And I don't believe that _I_ get to set the rules for what IBM does. They are a _company_ charged with making money. Nothing more, nothing less. DB and DB2 were both projects they undertook for financial gain. Nothing more, nothing less. To Hsu and the rest, it was a bit more, because they, like me, enjoy computer chess immensely. They got to do something they enjoyed. IBM used it to make money. The entire project answered a long-running question. Seems to me like _everyone_ achieved their desired goals at the same time. A good outcome for _everyone_. Even after all his whining, Kasparov made a small fortune from IBM and the match. So even _he_ did pretty well, IMHO, even with all the mistakes he made. I sure wouldn't mind getting mauled by a computer and getting paid $400,000 dollars for doing so... > > >About deconstruction of DB2: >############################# > >>They are obligated to keep a multi-million dollar computer around for >>your personal edification? The machine was sold. It was sold _before_ >>the match, but delivery was delayed so that they could use it. Cray did >>this for _me_ many times.... Nearly every year Cray Blitz played in an >>event, had you asked to see the machine to test a position a few weeks >>later, you would have been out of luck as the machine would have been shipped >>and installed at a customer site... > >You have the nerve to present such a paragraph as a proof for the seriousness of >the DB2 team and IBM? You identify habits with rightful practice? Do you believe >in the supernatural too? From what date on you'll define it as reality? > > I won't even venture a guess as to what that rambling response means. I am talking about reality. You seem to be talking about some drug-induced euphoric state where reality disappears and fantasy abounds. > >About the control of the output of a machine: >############################################## > >>Because "safe output" is an oxymoron. A contradiction. A physical >>impossibility. Any of the above. All of the above. Etc... If it >>doesn't exist, and can't possibly exist, then what is the point in discussing >>it ad nauseum??? >> >>We can state that such security measures are, by definition, impossible. >>Again, which makes the discussion pointless. >> >>Hence my argument all along... the machine is far too easy to modify to >>make any attempt to "security" 100% reliable. > >Did you in the USA invent the Fredkin prize? Or was it already part of the >biblical tradition? Sorry, but otherwise I can't teach you about real logic and >lame excuses. Rolf, I doubt if you could teach me _anything_. There is a prerequisite condition that must exist between a teacher and a student. I think you can figure out what that is and why it is missing in the context of you: teacher, me: student. Yes, the fredkin prize was a US product. So what? Computer Chess was a US product as well. Again, so what? > You create some magic and the rest of the world must accept it as >reality and real truth? You never heard of the saying that you should prove what >you declared? I believe _they_ did this. They "proved" their machine was able to do what they thought by playing a 6 game match and winning more than they lost... > Is it true that you are thinking that you are proven right as long >as nobody could prove that magic is only magic but not reality? What is all the magic/fantasy stuff? There were no smoke and mirrors in NYC in 1997. Just a computer, a program, some special-purpose hardware and a man by the name of Kasparov. _all_ were real. _all_ were present. The rest is captured by history. > Such thoughts >remind me of a famous court case in the US. Justice said that he was guilty and >he was executed. How come?? Isn't it the _whole_ USA believing in magic and >delusions yet? Is it perhaps concentrated on CC? Or did e.g. Heisenberg you a >bad job when he found his 'relation of fuzziness'? Can you spell 'Hand of God' >or 'Maradona'? > > >About Kramnik and the value of brutal violence: >################################################ > >>No... what you see is a Kramnik with better advisors. And a Kramnik that >>_knows_ that ChessBase needs _him_ and not vice-versa. Therefore he dictated >>the match conditions as favorable to him as he could imagine. Kasparov _could_ >>have done that or more. _could_ have... > >How many times more must I explain that chessplayers must not dictate such thing >to prove that they are stronger than a machine? I know e.g. that I must not >persuade _you_. But why do you drift into such Lucky Luke logic after all? I >always thought that DB2 team was lucky to play Kasparov. What the hell came into >them that they suddenly thought it convincing to to beat Kasparov - if not in >chess, then in a psycho war? For what purpose if even you, Bob, do _not_ believe >that comps are already strongrer than GM? Again magic looking through the >bathroom window'? Or is it possible to substantiate reality by pure power or >brutal violence? Does civilization prove the allegation? I don't think so! > Your basic premise is flawed: 1. Do I believe that comps are better than GMs in general? No. 2. Do I believe that comps _can_ beat GMs OTB. I can _prove_ it in fact. Are the two beliefs contradictory? Not at all. And to believe that they are contradictory shows that you simply do _not_ understand science in general. And statistics in particular. > > >About the American way of life: >################################ > >>I don't spend a lot of time worrying about things that I can't do anything >>about. For example, I don't have any meteor protection over my house. I >>don't think I could build anything big enough to protect me from any reasonable >>size of meteor. There is no way to make a machine totally immune to outside >>intervention, if the people that build the machine are intent on doing such. >>So I don't give it any thought. There is no way to detect such intervention >>with any reasonable reliability. So I don't give that any thought either. > >Is that the reason why people live in houses out of thin wood and wonder when >the wind or other reality bursts could not be sufficiently answered? Are you >really one of such people or was it just a misleaden metaphor? I am "really one of them". Nothing can stand against an F-5 Tornado. So worrying about it is impractical, unless I want a home totally underground which I would not have. So "I pay my money and take my chances" just like those that live along rivers and streams that occasionally flood, those that live along the coast and occasionally have to deal with hurricanes. Those that live up north and occasionally have to deal with 6 feet of snow. There are risks and rewards. I make decisions based on those two factors, where the reward is measurably better than the risk. > > >About propaganda: >################### > >>I think most understand that. Unless they don't _want_ to understand it, >>due to an agenda... > >If I take the logic of all your paragraphs I must agree with you. If most >understand it isn't so sure. I would already be happy if _you_ could correct >your delusions. It's a terrible mess to see you crawl and still with that >certain smile on your face. Couldn't that be from Melville? But then - who's >Moby? ;) > >Rolf Tueschen No connection with whales at all. Kasparov wasn't a life-long obsession with the DB guys, no more than he was an obsession with me. It was "the best chess player in the world" that was an obsession for all of us in computer chess, regardless of whether that was Fischer (which it was in the early 70's), Karpov after that, and eventually it became Kasparov. And eventually it will become yet someone else... It's the "concept" of machine beating man that was (and is) interesting.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.