George Duke wrote on Tue, Jun 3, 2008 12:40 AM UTC:
Hey we started this Chessboard Math thread, so we can add irrelevantly and
irreverently as follows. Devise a test for top-50 Grandmaster X (or K).
Set up certain moves from actual play at various stages, say 20 boards
altogether. Some are Computer-Computer, some Human-Human, and
Computer-Human. Be sure the chosen 'X' or 'K' has used the phrase
''computer move'' before. The test is to determine whether given
moves, numbering 20 in the test, all from different game-scores one of
each, is *Computer Move*. Most likely they cannot tell, better than expectation 10 of 20, and such talk is bluster. Turing test anyone?
