Michael Gusek wrote:
Has anyone devised a replacement/upgrade for/to the Turing Test?
What function do you propose that this replacement serve?
The original Turing Test was intended to move the idea of computer
intelligence out of the metaphysically unanswerable into the potentially
testable. It has been successful, for everyone who is capable of being
convinced. (Some people don't even accept that some other humans are
intelligent...so it's no argument that not everyone is convinced.)
If you define precisely what the goal of your test is, then it might be
possible to define a test that matches that goal. (One possibility is
"Drive from Annapolis, Md to New York city starting at 6:00 AM on a work
day without getting into an accident." That's a test that can't yet be
done, but may be possible soon. It involves certain kinds of
intelligence. Another possibility is "Write a book that becomes a New
York Times Bestseller.". That involves a different kind of
intelligence. The second seems a lot further away than the first, but
it's a definable test. True, it involves both skill and luck, but so do
many things that get called intelligence.
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=4007604&id_secret=75049540-8ebf48
