an ordinary chessplayer wrote on 03/28/18 at 03:30:10:
Maybe others will find this one interesting. I’m also hoping to see some other good quotes in reply. In the below quote, computers have turned out to be quite good at the “multifarious devious” part.
Horowitz was a little behind the curve, it seems. Writing in 1961, he could already have known about the young science of game theory. Also, Alan Turing's "Paper Machine" was able to play chess in 1952. That Paper Machine calculated
1.e4 Nc6 2.d4 as the best moves, which is still true today.
In its first corr game the opponent replied 1...e5, however, and the Paper Machine lost. [Source: Steinwender/Friedel: Schach am PC (1995), p. 35.]
We can go back to Torres y Quevedo and his K+R vs K automaton, and find similar sceptical voices. For example this one published in
Pester Lloyd, 18 June 1914:
Quote:It can be assumed that the Spanish inventor can also build a chess machine that plays other endgames on the chess board; as well an automaton is conceivable, which plays about "wolf and sheep", thus a game, with which the one party (that of the sheep) necessarily must always win, if they do not make a mistake. A machine however, who really plays chess, is an impossibility, the thinking activity can never be replaced by any automaton, and a chess automaton would only be possible under the (false) assumption that there were only a certain number of games in chess, and that in each position a move was the only correct one according to certain simple rules.
In his
Die Philosophie des Schach (Leipsig 1879), p. 6, Wekerle thought chess and science had much in common, but science was positive, while the basis of chess was, almost entirely, negativity:
Quote:Der beste Zug nämlich und der beste Plan nämlich sind die, welche von Irrungen am freiesten sind.
Google translation:
Quote:Namely, the best move and the best plan are the ones that are the freest of any errors.
Still a long way to John von Neumann's Minimax Theorem (1928)...