On 1 October 2013 08:44, meekerdb <meeke...@verizon.net> wrote:

>  On 9/30/2013 5:05 AM, Telmo Menezes wrote:
>
> Even under functionalist assumptions, I still find the Turing test to
> be misguided because it require the machine to lie, while a human can
> pass it by telling the truth.
>
>
> Actually Turing already thought of this.  If you read his paper you find
> that the test is not as usually proposed.  Turing's test was whether an
> person communicating with a computer pretending to be a woman and a man
> pretending to be a woman, would be fooled as to which was which.
>

I thought Turing mentioned "The Imitation Game" in which someone tried to
tell the other person's gender without any clues (like being able to hear
their voice, or discussing matters that only a man or woman might be *
expected* to know given the social norms at the time), and then extended
that to involve a computer as one of the participants?

That is, the TT as normally described involves someone trying to tell if
they're talking to a computer or a human being. Are you saying that isn't
how it was meant to be carried out?

I might also say that the above description of the TT (a computer has to
lie...) is also innaccurate, imho. The test is intended to indicate whether
a computer can be a *person*. For example if you were communicating with
HAL in 2001, you might easily mistake it for a man (as shown in the film)
and you would be right to do so, because HAL *is* a person in the story, by
any reasonable criterion. (In fact he's the most human like person in the
film! The astronauts act more like robots than he does most of the time!)

So a computer passing the TT (without hearing its voice or discussing
matters only a human being/computer is likely to know, of course - as
mentioned in the original paper, where the judge asks the testee to
multiply two numbers and it pauses a while, and then makes a mistake,
because the computer isn't allowed to invoke a "calculator function" just
as a human shouldn't use a calculator - not that such things existed at the
time!) shouldn't be considered to be lying. And it should be treated as a
person. I think that was Turing's point.

-- 
You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to everything-list+unsubscr...@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to