X-Message-Number: 14995
Date: Thu, 23 Nov 2000 11:54:41 +0100
From: Henri Kluytmans <>
Subject: give me proof too :)

Thomas Donaldson wrote:

>1. No one yet has produced a proof that a real-time computer with the
>   features of brains would actually be imitatable by a Turing machine.

If that real time computer is a discrete system the proof that it 
can be imitated by a Turing machine has been made.

>2. Just like primitive creatures, the basis of our brains' working 
>   depends not on math or logic but on the special features of our 
>   neurons as "machines". This is not just a statement about hardware,
>   but a statement of how our thinking works. The calculations of a
>   computer depend not on themselves but on the meaning human beings

It is sufficient that the internal symbolic representations are 
consistent. Then if it is an autonomous system, it does not matter 
what meaning humans attach to them.

>   attach to them. Our own thinking depends on those basic processes,
>   which generate (sometimes) patterns which look like logic and math.

Pattern recognition can be done by symbolic discrete systems, 
so why not the other way around.

>3. Unfortunately to someone familiar with the physiology of brain cells,
>   the models so far used lack essential features of the device which
>   is our brain. These include those I have discussed already: the
>   ability to grow new neurons and constantly change the connections 
>   of existing ones... 

I have given some references.

>   plus the very simple feature that we must work
>   in real time. 

Just a question of time, needed by the development of technology 
to create faster hardware.

>   I am not at all against the USE of such models, so
>   long as we do not confuse them with complete versions of how our
>   brain or any part of it actually works in detail.

The question is, of course, which details are essential for the 
functioning of the mind? Do we need to include details at the 
quantum mechanical level. I think not! And it seems that most 
neuro-scientists do neither.

>So come on, guys. I have asked a very simple question: has anyone got
>a proof that human brains can be imitated (even if only abstractly)
>by a Turing machine. 

Of course not! (Not yet.)
Neither is there any proof of the opposite claim.

However it does seem more likely that our brains can be functionally 
imitated by a discrete system than not.

>And I am emphatically NOT allowing partial imitations
>of some part of our brain, but only the whole brain. 

Why do you think that the neurons in some parts of the brain 
work differently at a *fundamental* level ?  Do you have any 
proof for this claim ?  :)


>Everyone who believes in the possibility of such imitations seems 
>to assume their possibility, without any attempt to show it. 

From a scientific point of view it seems more likely.
No, there  is no proof yet, but neither is there for the 
opposite claim. If there already was proof, then we wouldn't 
have this discussion.

>Everyone who refuses to believe in their possibility has various 
>features which may (or may not) make such imitations impossible. 

Those various features are either irrelevant or without proof.


Grtz,
>Hkl

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=14995