[ExI] Off Topic: Turing Test -- ai class at stanford

[Stefano Vaj]

On 30 August 2011 08:48, Adrian Tymes <atymes at gmail.com> wrote:

> That's why I said "so far as there is one".  OTOH, it was given as an
> example
> of something that, if imitated, would seem to leave no further thing that
> humans
> could point to as "intelligence" - because, if there was, then people could
> talk
> about it, but computers wouldn't be able to, thus giving a way to
> distinguish
> computers from humans via communication, thus they do not in fact pass the
> Turing Test.
>

"Intelligence" and "behavioural human-ness" are too often confused. As far
as basic arithmetic is concerned a pocket calculator is already more
intelligent than any human being. Conversely, there are plenty of human
beings who are not very intelligent but are more or less persuasively human.

The Turing test measures something which is relevant to social communication
with humans, but of course does not tell us anything about how intelligent
the system concerned is in terms of flops or of its performance in the
execution of other classes of programs.

-- 
Stefano Vaj
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20110830/db9d2d1f/attachment.html>