Excluding sub-cognitive questions makes test unbiased
A machine's inability to answer subcognitive questions—e.g. questions about ranking associative pairings—may show that machines can never match human responses, but we can make the test unbiased by excluding such questions.
Argument anticipated by Robert French, 1990.
Immediately related elementsHow this works
-
Artificial Intelligence »Artificial Intelligence
Can the Turing Test determine this? [2]  »Can the Turing Test determine this? [2] 
No: failing the Test is not decisive »No: failing the Test is not decisive
Turing Test is species biased »Turing Test is species biased
Excluding sub-cognitive questions makes test unbiased
A purely cognitive test isn't an intelligence test »A purely cognitive test isn't an intelligence test
+Commentaires (0)
+Citations (0)
+About