From: Eugene Leitl (Eugene.Leitl@lrz.uni-muenchen.de)
Date: Mon Aug 06 2001 - 04:51:57 MDT
On Fri, 3 Aug 2001, Robert J. Bradbury wrote:
> Ah, but must it be expansion in the *physical* reality?
No, it can. And this is precisely while many cans must become a must.
> If you think about the evolution of computer architectures
> from CISC to RISC to Transmetas compilable hardware
> there isn't any real physical expansion at all. There
> is however a huge expansion of the idea space.
Is it? There's a not very large jump in idea space from a Zuse, or even a
Jaquard loom to the Thunderbird.
> Going back to my quote from Eric -- if the way you can
> organize a 1 nm cube of atoms is 10^148 then it is going
> to take you a *very* long time to explore that simple
> phase space -- nevermind the number of atoms in the
Yes, but most of it is junk. There are only so many OPS you can get from a
number of atoms, without turning them into plasma (which doesn't compute)
> solar system. And in the phase space of computational
> power, you don't want to get bigger, you want to get
> smaller. You want to shrink -- not "expand".
That's only true for the individual modules of the hardware layer. That
the substrate will gulf out and usurp the dumb matter is pretty clear from
darwinian considerations alone.
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:09:30 MST