From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!think.com!samsung!uunet!morrow.stanford.edu!nova1.stanford.edu!dow Thu Dec 26 23:57:51 EST 1991
Article 2340 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!think.com!samsung!uunet!morrow.stanford.edu!nova1.stanford.edu!dow
>From: dow@nova1.stanford.edu (Keith Dow)
Newsgroups: comp.ai.philosophy
Subject: Re: Scaled up slug brains
Message-ID: <1991Dec21.055305.23986@morrow.stanford.edu>
Date: 21 Dec 91 05:53:05 GMT
References: <352@idtg.UUCP> <40782@dime.cs.umass.edu> <1991Dec18.135620.16540@news.larc.nasa.gov> <357@idtg.UUCP> <4001@litchi.bbn.com>
Sender: news@morrow.stanford.edu (News Service)
Organization: Stanford University
Lines: 27

>>Also you don't need to know the chemistry or physics of the human brain to 
>>simulate it.   There are lots of people in computer science classes who 
>>understand how a computer works, but think a transistor is a nun that has
>>had a sex change operation.
>
>I think your analogy is poor. CS students do not need to know low-level
>details to "understand" computers for two reasons: (1) computers were
>DESIGNED so that one can understand its workings without reference to
>low-level implementation details (2) CS students are TOLD how computers
>work - they don't have to discover this by experimentation. The fact is,
>we don't have a reference manual for the brain and I don't think that you
>will find many neuroscientists who would agree that all of the significant
>mechanisms of the brain are understood or have even been discovered. It
>is probably true that one does not need to know all of the low-level details
>of the brain to simulate it, but you must be able to determine which details
>can be left out and which must be considered. With computers, it is easy.
>We know they have been engineered so that one does not have to consider the
>minute fluctuations within the circuitry  - but we do not know this about the
>brain. The moral: I don't think it is safe to ignore all this "low level"
>stuff quite yet.

I agree, just like in computers, a lot of the lower level work has to be
done before the higher level can be understood. However an excellent question
is how well do the best neural networks simulate the brain?  Certainly
the neural network game can be played without any knowledge of brain
physics or chemistry.  Also the results of some simulations mimic aspects of
brain behavior well.  I will leave it to others to decide. 


