Path: utzoo!utgpu!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!nrl-cmf!ames!xanth!mcnc!decvax!decwrl!sun!pitstop!sundc!seismo!uunet!mcvax!hp4nl!htsa!fransvo
From: fransvo@htsa (Frans van Otten)
Newsgroups: comp.ai
Subject: Re: Intelligence
Summary: stop philosophy here / what intelligence really is about
Message-ID: <640@htsa.uucp>
Date: 6 Dec 88 09:36:35 GMT
Reply-To: fransvo@htsa.UUCP (Frans van Otten)
Organization: HTS A Amsterdam
Lines: 60

We are way off from the subject intelligence. What is discussed
now is philosophy. Let's get back to the subject.

In my article <607@htsa.htsa> I wrote:
> Besides, there is not just one kind of intelligence.

I remembered I had read something about this (long ago), so I had
to look it up (sorry for the delay, Shannon). Here's the story:

Howard Gardner (Harvard Graduate School of Education) wrote (in
Frames of Mind) about seven different kinds of intelligence. He
includes logical/mathematical intelligence as well as 'spatial
intelligence' (3D-insight). These kinds of intelligence are
measured in IQ-tests. But he also includes physical expression,
intra- and inter-personal intelligence, musical intelligence, and
speech-ability. In his view, a kind of intelligence must be
located somewhere in the human brain, and separately useable.

I want to change my statement: not different kinds of intelligence
exist in the human brain, but more then one 'expert system'.


'Intelligence' is a word with a very broad meaning. I propose the
following definition:

  If a [system] can reach conclusions, it is intelligent.
    [system] = human being, animal, computer, ...

An intelligent system consists of:

  (a). data
  (b). algorithm(s) to reach a conclusion based on this data

A simple example: data = 3, 7; algorithm = addition; now this
system can reach the conclusion that the sum of 3 and 7 is 10.
How intelligent the system 'seems' to be depends on the amount of
data and the algorithm(s).

Making the system even more intelligent requires:

  (c). ability to update (add/delete/change) the data:
       - by the 'outside world' (a 'database administrator')
       - by the system itself, based on its own conclusions

Now the system can calculate the sum of every two numbers. Isn't
that intelligent ! Finally, to reach the highest grade of
intelligence the system must also support:

  (d). the algorithm itself is data (thus can be changed)

Many Expert Systems contain only (a) and (b). They seem pretty
intelligent. But remember: they *only* differ from my Adding Expert
System in the amount of data and algorithms ! (As how intelligent
would my Adding Expert System have been regarded even a couple of
hundred years ago, let alone in the early Greek times ?)
-- 
                         Frans van Otten
                         Algemene Hogeschool Amsterdam
			 Technische en Maritieme Faculteit
                         fransvo@htsa.uucp