Hector Levesque, Distinguished Lecture, Cheriton School of Computer Science, University of Waterloo
This year being the centenary of the birth of Alan Turing, the talk offers two thoughts on the famous Turing Test he presented in his 1950 paper. First, I will argue that the emphasis it places on externally observable behaviour is just right, contrary to objections raised by some philosophers like John Searle. Second, I will argue that the emphasis it places on the behaviour being indistinguishable from that of a person is not quite right, and I will end up proposing a behaviour test that is less demanding but also less subject to abuse. Finally, I will speculate briefly on what sort of computer system stands the best chance of ever passing such a test.
Социальные закладки