Xref: utzoo sci.psychology:236 sci.bio:1121
Path: utzoo!mnetor!uunet!munnari!moncskermit!moncsbruce!mmcg
From: mmcg@moncsbruce.oz (Mike  Mc Gaughey)
Newsgroups: sci.psychology,sci.bio
Subject: Re: Intelligent Parrots, or Self-deception and Gullibility.
Message-ID: <419@moncsbruce.oz>
Date: 21 Apr 88 17:26:11 GMT
References: <3339@drivax.UUCP>
Organization: Comp Sci, Monash Uni, Australia
Lines: 61

From article <3339@drivax.UUCP>, by braun@drivax.UUCP (Kral):

> I agree somewhat.  Did anyone see the PBS special on the brain broadcast a
> couple of weeks ago?  There was a great experiment where (gee, I wish I could
> remember details, like *names*) brain tissue was excited to generate a current
> between neurons.  As the current was produced, the path between the neurons
> (help me out with names here) became more efficient at carrying the current,
> and the neurons changed to, all to facilitate 'message carrying' on that
> 'path'.  The implications relative to this topic are (in my opinion) this:

	There is a good book, by Valentino Braitenburg, called "Vehicles".
Braitenburg is (was?) the director of the Max Plank institute.  In the
book, he conducts a thought experiment in which vehicles with two independantly
driven wheels and a variety of sensors exhibit varying behaviours - e.g.
if the vehicle is a rectangle -

	wheel A
	+--------------------+
	|		     + sensor X
	|		     |
	|		     + sensor Y
	+--------------------+
	wheel B

we can imagine a variety of interconnections between A, B, X and Y.  
If, for instance, stimulation of Y induces extra speed in A, and the
same for X and B, we have an aggressive vehicle that turns towards and
runs at full speed into whatever excited it (a lightbulb destroyer?)

	That was chapter 1.  Succeeding chapters introduce the concept
of associative memory (using the mechanism described in the posting
I have included), darwinian evolution, causative memory (is this the name?),
higher faculties, instincts.  By the end of ch. 15, we have a machine
which, to all outward appearences, acts of its own volition, thinks,
makes decisions, meditates, has instincts, and, in general behaves in
a very complex manner comparable to many animals.

	The other half of the book (yes, there's more!) is a look at
real life examples of the types of vehicles he has constructed in the
first half.  Have you ever wondered why a fly always lands on the dark
spot on a wall (or flies out a window from a dark room into the light)?
According to Braitenburg (I'm no biologist), the fly is organised just like
our little vehicle above (that's why I chose this example).  The X and
Y sensors detect movement (i.e. the motion of contrasting features in the
fly's field of vision).  The wheels are, of course, the wings.  When a fly
enters a room, with a spot on one wall, the spot will be moving more
in one of its eyes tham the other (remember, the fly is just a little buzz
box - everything is buzzing).  Because of some simple brain interconnections,
the opposing wing beats just a little harder - so the fly orients towards
the spot ... and finally lands on it.  Neat, huh?

	This book is written in a very witty, lighthearted manner, by an
acknowledged expert in the field (cybernetics).  It was written mainly
to demonstrate that things are always simpler than they seem, and is
very readable.  I don't have a copy of the book now (I read a friends), so
I'm sorry I can't give any more details - but if you see it around, read
it.

	Feel free to correct my misconceptions.

					.... Mike.