Re: turing test

[David Mechner <mechner@xxxxxxxxxxxxxxxxx>]

I'd be a little disappointed if either one were my student, and pretty
happy if either one were my program (with the exception of the fight in
the lower left ;)

I can't tell which is which with any high degree of confidence - they
both make many bad but normal moves, and many good moves, and many
shocking blunders. My faith in human nature, and my experience with
human students, tells me that there's no move so bad a human can't make
it, and arguments like "a computer wouldn't make a move like X" are
unconvincing to me; they remind me of the essay Edgar Allen Poe wrote
debunking "The Turk" (the master-level chess playing "automaton" that
had been touring the courts of Europe) - he said there had to be a man
inside because if it were really a machine, it would play perfectly. 

The life and death fight in the lower left (124...137) is quite
amusing;  obviously the human is as bad as the computer at life & death
(and vice versa). Of course, my hope is that my program would know how
to kill or live with the group in the lower left (and would do it :).
And if either one were my student I'd make them study life and death
problems. 

Fun though this is, just a reminder: the Turing Test isn't to write a
program that convinces someone that there's a baby at the other end of a
teletype :)

-David

-- 
David A. Mechner            Center for Neural Science
mechner@xxxxxxxxxxxxxxxxx         4 Washington Place, New York, NY 10003
212.998.3580                http://www.cns.nyu.edu/~mechner/