From: Eugen Leitl (eugen@leitl.org)
Date: Thu Nov 27 2003 - 04:22:40 MST
On Wed, Nov 26, 2003 at 07:46:48PM -0800, Jef Allbright wrote:
> Reminds me of an interesting parallel with Buckminster Fuller's saying it's
> not likely that nature is calculating pi to an some infinitesimal number of
> decimal places with each soap bubble that appears. Platonic concepts have
> utility, but not perfect utility.
Amen.
The CNS is a dirty computer, and does not clean representation/hardware
layer separation. It's a real bitch to emulate that in solid state
electronics (or spintronics, anything few-state discrete with a simple
state translation function).
For some time I've been thinking that a good cellular machine must
directly mutate the substrate structures (using *effective* morphogenetic
representation and generation) as well as the state of such structure
in a problem-solving co-evolution scenario to find the nucleation point
for a decent substrate/coding we can then drive further. Polymer/quantum dot
ink is slow operation, but it can handle multilayer and parallel deposition, so
we might be able to reduce the threshold for fabbing to bring interesting
hardware within modest budgets (a wafer carrier is worth some 20 M$,
it takes days to pass through the fab; I have no idea how deep the pipeline
is but it is weeks before a minor process tweak has stabilized).
It's nice gaming converges towards physics simulators, which have
a tradition of parallel codes and parallel hardware, which might break up
the monolith, and give us tools to tackle above challenge.
-- Eugen* Leitl leitl
______________________________________________________________
ICBM: 48.07078, 11.61144 http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE
http://moleculardevices.org http://nanomachines.net
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:43 MDT