No. 108: Nov-Dec 1996
In SF#106, we chronicled the chess match between IBM's Big Blue computer and human champion G.K. Kasparov. Incidental to the match itself was the fact that Big Blue sometimes made different moves when confronted with identical chessboards! With these denials of cyberdeterminism in mind, consider the following anecdote.
About 40 years ago, D. Herschbach witnessed the first test of an early chess-playing program written by the renowned S. Ulam for the MANIAC I computer.
"In this test, the MANIAC I computer played both sides of the board. After a rather bizarre game, Black made a move that checkmated White. Only then did Ulam realize that the program did not say what to do when mated; it just required that a move be made to escape check. While Ulam and his colleagues debated what might happen, the computer whirred on for about ten minutes. Finally it punched out White's move (on paper tape, then the mode). The uncanny solution: a spurious pawn appeared and began to march down the board to become a new queen."
(Herschbach, Dudley; "Computer Milestones," Discover, 17:12, October 1996)
Comment. If vacuum-tube-equipped MANIAC I was so resourceful and innovative, what might a modern supercomputer "think" up? Remember HAL of 2001: A Space Odyssey and treat your own computer with respect!