Re: The Cornucopian Fallacies

From: Robert J. Bradbury (
Date: Sun Dec 09 2001 - 05:05:09 MST

On Sun, 9 Dec 2001 wrote:

> I tend to side with Simon, for
> the simple fact that oil has not run out, that food production has not
> declined, etc. Do we need to move away from gasoline and diesel as
> transportation fuels, yes. Do we need to devastate our economy in the
> process? no.

I read a recent analysis and the "wild card" that seems to be making the
Simons right and the Erlichs wrong is technology (driven by human ingenuity).
The geologists are figuring out how to get more of the oil out of the
ground and still aren't getting a large fraction of it. So there is
substantial room for improvement.

If the plans the U.S. DOE has on the board for electricity production
get implemented over the next 10-20 years then I think the problem
is solved. Everyone knows we have several hundred years of coal
resources and the electrical production efficiencies are slated to
jump from ~35% to >80%. Several hundred years give us plenty of
time to switch to solar power and/or fusion. Cheap electricity lets
you switch to hydrogen as a locomotion fuel for most applications.

> Now answer for yourself, if you believe that nanoproduction of goods and
> services is a Cornucopian fantasy? It clearly is, in the sense of its
> potential, and what Drexler has indicated about it. But is it a dream that is
> likely to come true, or a pipe-dream?

"It clearly is"? Basis for assertion please?
Take a look at 3D printers. What is there that is so hard to grasp
about the concepts of self-replication, exponential growth and the
concept of turning matter into "software"?

I will not argue that it *will* be complex. You aren't going to
design the precise location of every atom in an 18 billion atom
nanorobot yourself. You are going to write the software that does
much of the work and actually directs the assembly process. But
I've made the point in my Nano@Home paper -- the problem isn't
with the assembly process -- it is in having the designs. If you
look around at all the nanomachinery we currently have (also known
as "nature") its clearly going to be a big job designing the equivalent
and more -- but it doesn't require a great deal of actual
"intelligence" -- it simply requires the ability to create
and recognize good designs.


This archive was generated by hypermail 2b30 : Sat May 11 2002 - 17:44:24 MDT