From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Fri Feb 15 2002 - 07:33:58 MST
Michael Butler wrote:
> It seems that one must deal with the notion that self-replicating general
> purpose assemblers is Drexler's sole contribution/albatross;
I don't consider that to be the case. If one reads Nanosystems
and Eric's other papers carefully there are a huge number of
other things discussed. A few would include -- the problems
of constructing molecules with strained bonds, the damage UV
radiation causes to molecules, the rather fantastic number
of structures that can be constructed within a cube measuring
100 atoms per side, the fact that heat removal is the *real*
problem for high computation densities, the fact that one has
to move from non-reversible to reversible computational
architectures, the fact that there is a clear chemical/biotechnology
based path to assemblers by designing molecular parts of increasing
stiffness, etc.
The fact that SRAs can exist is obvious to anyone who understands
biology. Eric's extension of this to general purpose assemblers
is a significant contribution but I wouldn't consider it *that*
significant. Its only significant from the perspective that
he pointed out the risks that such devices could pose and that
gets everyone who isn't aware of the existing "green goo" situation
all upset.
> probably by tacitly ignoring it when dealing with those with
> short-term parallel goals.
You don't have to ignore it at all. You simply point out there
are *very* effective ways of dealing with the risks, e.g. heat and
radiation. Those are the same methods that have been used to deal with
the green goo for ages. It is unfortunate that people generally lack
the knowledge to paint a complete picture -- that you can have
assemblers that are non-self-replicating that can rely on
broadcast architectures and that the people who promote the
dangers of SRGPA do not in the same breath acknowledge we are
effectively already in that situation *and* there are defenses that
can be employed should we choose to do so.
> Calculating the consequent risks & goal drifts requires craft.
> Argue, or win the war?
Educate, avoid annoying people as best you can and prepare to
defend oneself.
Robert
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 13:37:39 MST