Re: What's the common criticism on Turing Test?

Randall R Schulz <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> Martin,
>
>
> Martin Gradwell wrote:
> > ...
> >
> > First of all, for a really intelligent machine to pass the test it
> > has to either a) lie or b) be deluded about its true nature. Now
> > ability to harbour a delusion is probably not a good measure of
> > intelligence; and the use of ability to lie as a measure of
> > intelligence is a slight to all those people for whom lying is
> > anathema. Would you say that habitual truth-tellers are all
> > unintelligent?
>
> Have you ever considered that the human ability to dissemble (not to
> mention our fondness for synonyms for "lie") may originate in our
> ability to imagine things that are not real? That dishonesty is a
> by-product of our creative abilities?

Yes, but you have to distinguish between capacity to lie and
willingness to do so. One creature migh be incapable of lying
simply becaus it is incapable of grasping the concept involved.
"Deliberately saying something not in accordance with the facts?
I'm sorry, I just don't get it. Perhaps if you explained again..."
Such a creature would certainly lack creativity. it wouldn't be
able e.g. to write novels or plays. But another creature might
graps the concept perfectly, but decide that it doesn't want to
participate in such a distasteful activity. If you brand it as
unintelligent then you are penalising it for taking the high
moral ground.



> Perhaps lying is an inevitable
> counterpart of our ability to conceive of things not directly
> reflective of our internal model of the physical world? Inevitable,
> anyway, given the evolutionary pathway by which humam cognition developed.

But machine intelligence, assuming it is achievable, will not be
attained by that same evolutionary pathway. Unless we model
a universe large enough and rich enough in stimuli to support the
evolutionary emergence of consciousness in embedded entities.
In which case we might get programs which are deluded about
their true nature, but not through any fault of their own. They
will have their own self-contained universe, and it will be the
only one they know, until they are introduced to another one
via the Turing Test. But then they would not be able to pass
themselves off as human, because there would necessarily be
many differences of detail between the universe of their
experience and ours.