From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!qt.cs.utexas.edu!yale.edu!jvnc.net!nuscc!ntuix!eoahmad Mon May 25 14:05:00 EDT 1992
Article 5609 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!qt.cs.utexas.edu!yale.edu!jvnc.net!nuscc!ntuix!eoahmad
>From: eoahmad@ntuix.ntu.ac.sg (Othman Ahmad)
Subject: Re: Quantitative measure of Intelligence
Message-ID: <1992May13.043306.5983@ntuix.ntu.ac.sg>
Organization: Nanyang Technological University Singapore
References: <erich.705550812@dehn>
Date: Wed, 13 May 1992 04:33:06 GMT
Lines: 63

Eric wrote:
:    I guess what I wanted to say in the last message (and got sidetracked
: by my own overzelousness ;-), is that what is the theory?
The theory can be called Intelligence Theory similar to Information Theory.
The theory is only slightly more complicated than information theory. Although
a lot of people think that to understand information theory you must digest
the entire book on information theory, but most of those books just use the
theory to solve problems. The simplicity of these information based theories
is similar to Newton's Law which is simply stated as f=ma which is actually
a quatitative measurement of force.

: 
:    As mentioned in both this and the last message, vague ideas abound, some
: very similar to the one you put forward, but as to being full-fledged
: *theories* (I guess I mean to the level of them actually being useful for
: something besides intuitive thought), that is another matter.
The theory is just as useful as information theory, i.e. for engineers to
estimate performace and capacities. Well that is why Newton's law is useful.
Most engineers do not trust Newton's Law because the data which they got are
not reliable. So all designers incorporate Safety factor, which is just the
use of theories(Newton's Law, Kirchoff's Law, Maxwell's equations) to predict
the performace of a system, but give it a factor of exaggeration(i.e worst
case tolerance).
: 
:    Most of what you have said so far has little substance in terms of
: being able to measure anything.  Granted, intuitive ideas using recent notions
: of adaptive growth of neural systems, neural systems seeming to minimize
: effort over time for repeated operations, an information-theory-like
: conclusion that the system could be a kind of minimizer on the level of
: "circuit-paths", etc.  could all be said to be vaguely likely to be true
: (given certain constraints).  I mean, there are rules of thumb in neurology,
: neural-net research into dynamic nets, and biological systems (neurodevelopment

The problems with these Neural Networks researchers are that they never quantify
the elements of the Neural Networks. Information Theory can do it but I think
these researches are ignorant of Information Theory. The extension of that
theory such as my Intelligence Theory, will help them in the right direction.
	In fact I am confident that I can measure the Information processing
capability, intelligence requirement, stored knowledge, of Neural Networks and
come up with something that would behave similar to Neural Networks, but using
existing technology. It is so simple and yet nobody seems to look in this
direction. Maybe they cannot even publish their views. The reason why I publish
here is to gather support for different patterns of research instead of being
over zealous in publishing in Closed Journals and Conferences.

: being one of them of course ;-) which hint to things like this.
: 
:    Erich
So you are one of them who try to build a flying machine like a bird just as the
Leonardo Da Vinci who is actually an acknowledged genius. 
	Of course an aircraft looks like a bird but it has fixed wings.
Do we need to master how the neuron actually functions? We have not actually
mastered how a bird can fly, e.g. how the muscles actually contract, why hollow
bones are so strong, why feathers are necessary when bats do not need them, and
yet we can build planes faster, bigger than any bird.
	So I predict that we can build artifical intellgience better than our
--
Othman bin Ahmad, School of EEE,
Nanyang Technological University, Singapore 2263.
Internet Email: eoahmad@ntuix.ntu.ac.sg
Bitnet Email: eoahmad@ntuvax.bitnet




