Re: No nanotech before AI

From: Robert J. Bradbury (
Date: Thu Jul 13 2000 - 12:45:48 MDT

On Thu, 13 Jul 2000, Robin Hanson wrote:

> It seems far from clear to me that it is trivial to take a design for a
> bridge, throw some nanodust at it and say "Build it for me over there."
> Assembly and construction use intelligence, just as design does.

Hmmmm, if that is true, then how do we measure the IQ of a bacteria?
Self-assembly can be done quite simply by throwing the parts in a heap
and shaking or stirring until the desired result is obtained. The
tough nut is to design the pieces so they can effectively self-assemble.
That is generally not how macro-scale design is currently done.
The only example I can quickly think of are things like collapsed
cardboard file boxes that if you push and pull on them enough will
assemble themselves into an actual box.

A software team (at Brandeis?) has shown that you can do rather
kludgy design of mechanical "stuff" simply by randomly varying
the assemblies (mutation?) and recognizing when you actually have
something that works (selection). So I would have to say most
of the intelligence involved is in recognizing when you actually have
a part that does what you want (including potentially self-assemble).
Intelligence would also seem to come into play in being able to prove
you have the most efficient assembly or construction path.
But if my theorem proving exercises are typical, it would seem
that there is a lot of mutation and selection in that process
as well.


This archive was generated by hypermail 2b29 : Mon Oct 02 2000 - 17:34:34 MDT