From newshub.ccs.yorku.ca!torn!cs.utexas.edu!swrinde!mips!sdd.hp.com!caen!umeecs!dip.eecs.umich.edu!marky Wed Aug 12 16:52:49 EDT 1992
Article 6599 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!torn!cs.utexas.edu!swrinde!mips!sdd.hp.com!caen!umeecs!dip.eecs.umich.edu!marky
>From: marky@dip.eecs.umich.edu (Mark Anthony Young)
Newsgroups: comp.ai.philosophy
Subject: Re: Turing Test Myths
Message-ID: <1992Aug12.063425.13479@zip.eecs.umich.edu>
Date: 12 Aug 92 06:34:25 GMT
References: <2838@ucl-cs.uucp> <1992Aug11.143819.22170@zip.eecs.umich.edu> <BILL.92Aug11105853@ca3.nsma.arizona.edu>
Sender: news@zip.eecs.umich.edu (Mr. News)
Organization: University of Michigan EECS Dept., Ann Arbor
Lines: 38

%r bill@nsma.arizona.edu (Bill Skaggs)
>marky@dip.eecs.umich.edu (Mark Anthony Young) writes:
>
>   > The "two second" Turing Test would involve telling her that she
>   > might be talking to a machine (and providing a human for
>   > comparison purposes), and then comparing Eliza's performance
>   > after two seconds with that of men immitating women for two
>   > seconds (over a teletype interface!).
>
>There seems to be a slight confusion here.  The Turing Test does not
>in any way involve men imitating women.  Turing used that idea to
>*motivate* the test, but it isn't part of the test -- the test only
>looks at the ability to correctly distinguish humans when computers
>attempt to imitate them.
>
>	-- Bill

No confusion on my part, I assure you.  I quote:

     We now ask the question, "What will happen when a machine takes
     the part of A in this game?"  Will the interrogator decide wrongly
     as often when the game is played like this as he does when the
     game is played between a man and a woman?  These questions replace
     our original, "Can machines think?"

"As often ... as" is a strong indication of a comparison.  Note that it 
is "these questions" and not "this question".

The point of the Turing Test is simple:  We know that men can think, and
we want to test the machine; so we compare its success rate at pretending
to be something it's not (a human) to a man's success rate at being
something he's not (a woman).  The simple success rate of the machine is
a meaningless number -- we have no way to tell whether it's good or bad
except by intuition.  The success rate of the thinking man immitating
the thinking woman is a good measure of the goodness of the machine's
success rate.

...mark young


