Re: Novel research in Artificial Intelligence?


Svante wrote:
> Hi. I'm a freelancing science journalist preparing an article about
> the most recent development in Artificial Intelligence. My starting
> point will be the hopes in the early days of AI-research and a
> follow-up on the state-of-the-art AI research done today. One central
> question is, if there is (or will be in the near future) a machine that
> will pass the "Turing test". This is a hotly debated issue and I
> will try to steer clear of the more sophisticated philosophical
> problems raised.
>
> So what I need to know is: Where can find research in the forefront and
> are there key persons in the field that might be useful to talk to?
>
> All hints and tips appreciated
> /Svante

A certain aspect of the test was for the computer to only answer the
question of the human. So the person sits and asks questions only.

A relation of this class required the reader to understand. And here
the term understand refers to the true intention of the answer.

A certain relation of understanding is the confirmation of true
intention, or the interogatory.
So when a person is confused the person asks another confirming
question.

And the interogatory is then the kind of relation most difficult to
write into the code. A system of answers and all possible confusion is
then the rather large databank to fill.

Search google and you will likely find a Turing Test machine willing to
answer. And to try to unconfuse yourself is a major computer system
failing. A computer will try to ask you to clarify the new subject.

Novel research is the means of resolving the interogatory dilemma.

So look for that subject area.

Douglas Eagleson
Gaithersburg, MD USA

.