From: Ramez Naam (mez@apexnano.com)
Date: Mon Nov 25 2002 - 17:09:37 MST
From: Robert J. Bradbury [mailto:bradbury@aeiveos.com]
> > Let me speak a bit more clearly. It is not at all clear to
> > me that by
> > 2050 we will have sufficient computing power on this planet to
have
> > designed an assembler that can build a car from the ground up.
>
> Mez may be mixing a several complex issues here -- (a)
> designing an assembler (~30x100 nm with a few million atoms);
> (b) designing a car with a precise level of atomic detail
> (which is a *whole* lot of atoms);
> (c) the systems problem of coordinating many
> (millions-billions) of assemblers to assemble the car.
>
> They are three distinct problems
Robert's right, of course. I wasn't drawing much distinction between
these problems. To some extent they're intertwined, because you want
to make sure the assembler you design in (a) can be effectively
controlled so you can realize (c), which is crucial to implement what
you design in (b).
To be clear, (b) doesn't worry me very much. We already design cars
with CAD tools. It's designing the systems that grow the car, and
then programming and controlling them, that I'm worried about..
Honestly I think (c) may be the toughest. This is mostly intuitive on
my part, though. I AM surprised that I have yet to see a
mathematically grounded analysis of the coordination and programming
problem. It seems to me that these control and design issues will
give us far more trouble than the issues of fabricating things at the
atomic scale.
My hunch is that the reason no one has produced a good analysis of
problem (c) is that it's deeply tied up in complexity theory, and
complexity theory has yet to produce good tools (so far as I know) for
analytically determining things like the order of complexity of a
system, or the amount of information required to control a system.
cheers,
mez
This archive was generated by hypermail 2.1.5 : Wed Jan 15 2003 - 17:58:23 MST