Will RYK be as good a player as Kasparov? Of course not. Sometimes the RYK system will randomly make dreadful moves which the real-life Kasparov would never make—start the game with P-KN4. I assign such moves a low probability, but sometimes the computer makes them anyway, by sheer random chance.
If you believe that Kasparov would never make that move, you should assign it a probability of 0.
Regardless of whether Grob’s Attack is dreadful, this article is. RYK doesn’t make dreadful moves that the real-life Kasparov would never make because of “sheer random chance”, it does so because it has nothing whatsoever to do with Kasparov—you could substitute ‘I’ (“how I imagine a good player might move”) for ‘K’ everywhere in your silly article. The information content of RYK is drawn solely from the chess knowledge of an astoundingly weak player—you—deflated by your own uncertainty about your chess judgment. That is, it’s the same as RYY, but with the probabilities smeared, reducing the probabilities of the moves you think are good and increasing the probabilities of those you think are bad. If you’re a bad enough player, RYK could actually be better than RYY, by sufficiently often avoiding the horrible moves you would make and sometimes making good moves that you never would.
As for calibration, RYK is not at all calibrated to Kasparov’s actual behavior, making invocation of his name, and any surprise as to how badly RYK plays, absurd. Your blather about “the creative unpredictability of intelligence” is absurd; Kasparov could be a completely deterministic engine, always making the same moves in the same positions, and nothing would change—he would still beat your ass every time and RYK would still suck.
Will RYK be as good a player as Kasparov? Of course not. Sometimes the RYK system will randomly make dreadful moves which the real-life Kasparov would never make—start the game with P-KN4. I assign such moves a low probability, but sometimes the computer makes them anyway, by sheer random chance.
If you believe that Kasparov would never make that move, you should assign it a probability of 0.
Regardless of whether Grob’s Attack is dreadful, this article is. RYK doesn’t make dreadful moves that the real-life Kasparov would never make because of “sheer random chance”, it does so because it has nothing whatsoever to do with Kasparov—you could substitute ‘I’ (“how I imagine a good player might move”) for ‘K’ everywhere in your silly article. The information content of RYK is drawn solely from the chess knowledge of an astoundingly weak player—you—deflated by your own uncertainty about your chess judgment. That is, it’s the same as RYY, but with the probabilities smeared, reducing the probabilities of the moves you think are good and increasing the probabilities of those you think are bad. If you’re a bad enough player, RYK could actually be better than RYY, by sufficiently often avoiding the horrible moves you would make and sometimes making good moves that you never would.
As for calibration, RYK is not at all calibrated to Kasparov’s actual behavior, making invocation of his name, and any surprise as to how badly RYK plays, absurd. Your blather about “the creative unpredictability of intelligence” is absurd; Kasparov could be a completely deterministic engine, always making the same moves in the same positions, and nothing would change—he would still beat your ass every time and RYK would still suck.
Unpredictability is not the same as indeterminism. It’s unpredictability from the perspective of a less-skilled agent.