From newshub.ccs.yorku.ca!torn!utcsri!rpi!usc!wupost!uunet!mcsun!uknet!qmw-dcs!abreu Tue Jul 28 09:41:32 EDT 1992
Article 6466 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!torn!utcsri!rpi!usc!wupost!uunet!mcsun!uknet!qmw-dcs!abreu
>From: abreu@dcs.qmw.ac.uk (Abreu)
Newsgroups: comp.ai.philosophy
Subject: Re: Defining intelligence
Message-ID: <1992Jul16.203118.16680@dcs.qmw.ac.uk>
Date: 16 Jul 92 20:31:18 GMT
References: <1992Jul8.092458.3088@otago.ac.nz>
Sender: usenet@dcs.qmw.ac.uk (Usenet News System)
Organization: Computer Science Dept, QMW, University of London
Lines: 74
Nntp-Posting-Host: it106.dcs.qmw.ac.uk

In <1992Jul15.233344.6478@u.washington.edu> sanelson@milton.u.washington.edu 
(S. A. Nelson) writes:

 > >   I'd say that the ability to solve problems is a RESULT of
 > >   intelligence, and certainly that there's more to intelligence
 > >   than logic.
 [...]
 > >The goal, in, say, writing a novel, is to produce *something of
 > >value*.  Since the space of novel-length word sequences is enormous,
 > >and the subset of high value is relatively small, producing a good
 > >novel is a search problem of formidable proportions.  Finding a
 > >solution demonstrates a great deal of intelligence.
 > >
 > I would have to disagree that the goal here is to produce something
 > of value.
 [...]

And I would have to disagree that producing a novel consists of
searching through the space of novel-length word sequences. The
search is a good way of representing or thinking about it. But
there's more to it, isn't there?

Think for example of the fact that some writers are more successful
than others, and in purely literary terms, some novels are just
better than others (in the reader's opinion). What makes some of
them better, to certain people?

Presumably, the value people attach to a novel, is a direct result
of their experience in reading novels. The author(ess,) on the other
hand, is constantly making choices of words, judging which are the
most appropriate to express their thoughts (value judgements) and
achieve their ultimate objective.

This leads us to think of novel production as a search through a
space. And very intuitive that is, too.

But in my opinion that just isn't the way we do write. Take
me, for example, writing this message. To me, it feels more like
I'm putting things together. I express my thoughts verbally, by
producing GROUPS of words in a logical sequence. What led me to
highlight ^^^^^^ in the previous sentence? I was trying to make
a point that I don't really produce each word individually, so
I decided to go back, highlight the thing, and hope you'd get it.
Then it occured to me to point it out, by chance the arrows fell
in the perfect place, and I got a good feeling about the whole
sentence (which should tell you something about the value I attach
to it).

Now my point is that you don't build a house, write a novel, etc.
by searching the possibility space. The actual space is of such
huge proportions, that its limits are inexistent: the space of
houses that can be built, the space or novels that can be written,
etc. is actually infinitely large. And this is just the space that
is meaningful to us (high value.) Either the heuristics we use to
search such spaces are tremenduosly powerful, or they are tremen-
duously limiting, preventing us from doing effective searches.
I can't see other explanations to the problem of searching an
infinitely large space.

What do you think? Is the search space actually infinite?

 > >I don't think it's useful to identify intelligence with everything
 > >that's happening in the brain.  Most people are comfortable speaking
 > >about a dog or a bear as having behaved intelligently in certain
 > >circumstances.

 > I'm with the Turing-Test crowd in that I think the closest thing we
 > have to defining "X is intelligent" is "X behaves like me."

Absolutely. Up to now, and unless some form of measuring intelligence
is created, the only way of verifying intelligence is through its
expression through intelligent behaviour.

Hamilton


