From: Jacques Du Pasquier (jacques@dtext.com)
Date: Sat Feb 16 2002 - 05:22:48 MST
Did anyone try to upload primitive animals ?
Of course there has been behaviour modeling of insects, for example,
with simple rule sets designed to account for their behaviour, and
even the reunion of many such little agents in a simulation to produce
emergent complexity.
But did anyone simply try to reproduce the information processing
infrastructure of a tiny beast, without any programming, plug it into
a (programmed) VR (including the body) and see how it does ?
For example C. elegans has 302 neurons. That should be within reach.
http://www.ncbi.nlm.nih.gov/books/bv.fcgi?call=bv.View..ShowSection&rid=ce2.section.d1e47963
Of course an uploaded C. elegans may not be very spectacular, and we
might try to model something a bit more "brainy".
Is the scanning exceedingly difficult and long even for primitive
nervous systems ?
Modeling the environment (body and external environment) in some cases
might be out of reach, but probably not in all thinkable situations.
And you can plug a brain unprogrammed copy in a more rule-based
environment.
Why do that ? To start thinking more technically about human uploads ;
and in particular to investigate the right level of detail to retain
in the copy of the nervous systems. If I understand well, we're still
not quite sure about this.
I find upload very exciting because it doesn't assume much (contrary
to nanotech or programmed SI). I would like to make it assume even
less by experimenting and adressing possible problems.
Jacques
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:12:25 MST