X-Message-Number: 26109
Date: Fri, 29 Apr 2005 10:56:01 -0400
From: Thomas Donaldson <>
Subject: for Yvan once more

For Yvan:

This is getting just a bit repetitive. But here goes again.

Ok, you mainly are trying to produce a neuron, not a whole brain. 

If I understand your description the "neuron" you're proposing
will not work at all like a real neuron. The problem comes
with growth of dendrites to match with another neuron. If the
other neuron is sufficiently close, they can form a synapse without
much need for growth. Otherwise one or the other (usually the one
with the extension of its axon) must grow axonules from its axon
to the other neuron. The other (usually) must grow dendrites 
to the axonule when it arrives. (Axonule is a word I just invented:
axons, too, split into branches, and an axonule is one of those
branches, or a branch of a branch etc). 

So we come to the question of how the axon of the sending neuron
grows properly toward the receiving neuron. In real neurons, 
guidance is by chemical factors released by the receiving neuron,
by astrocytes surrounding it, or by other nearby cells. Growth
involves sensing these factors.

You are correct that memories (in present views) come from the fine
connections of neurons. The problem comes when you try to 
implement this growth of a new connection in the neurons you
are discussing.

And of course, please don't go on about time-sharing in your "neuron".
Given the number of chemicals involved in the working of a real
neuron, it's unlikely that you can time-share even a single neuron.
And trying to time-share a group of neurons will get you into 
trouble, too. Incidentally, I'll point out that if you DON'T use
time-sharing for your neurons but give each one its own processor
(or given that neurons are complex themselves, even more than one
processor doing different jobs), then you'll also get FAR FASTER
neurons than if you have one processor trying to act like 1000 
neurons.

           Best wishes and long long life to all,

                Thomas Donaldson

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=26109