Re: No nanotech before AI

From: John Clark (
Date: Fri Jul 14 2000 - 22:38:56 MDT

 Dan Fabulich <> Wrote:

>Couldn't the millions of years of development work on
>DNA and the rest of the body have created something that performs a
>more difficult job than screwing in rivets?

Certainly, but when a embryo gestates in the womb intelligence is not needed
for its construction, no intelligence at all, not even a little. It's a good thing
the construction parts aren't macroscopic or we wouldn't be here.

>But buildings are rather expensive also; beams are relatively cheap.

Have you priced building materials lately? Let's just say they're not cheaper
than atoms.

>Why don't we just make robots build buildings

The same reason evolution never came up with a way to build complex things
out of macroscopic parts, or at least it didn't until it came up with humans.
It's too hard to make things that way. I gave 6 reasons why.

> Drexler, for example, imagines machine-phase assembly to look an awful
> lot like macroscale assembly,

There are similarities, the most obvious is that both are building something, but
as I pointed out there are also important differences too.

>as far as the number of parts are concerned; rather than fill Nanosystems
>with discussion of how pure carbon reacts and how we can use that to build
>complex polymers one monomer at a time, he fills his diamondoid chapter
>with sketches of designs for levers, struts, bearings, rivets, etc.

The ultimate purpose of nearly all those machines is to manipulate bits of
information or manipulate individual atoms, it will take complex machines to
do that and if you want to build something macroscopic in a reasonable time
you'll need an astronomical number of those complex Nanomachines. If you
want to call that "chunking" that's fine with me but at the end of the day it all
comes down to picking up an atom here and putting it over there.

>>There is no way you can damage the parts Nanotechnology deals with.

> Not true if you chunk the problem the way Drexler does; indeed, he
> spends a lengthy section of the book dicussing just when and how his
> bonds will break under stress.

You can't damage atoms and of course bonds are not infinitely strong.
I must be missing something.

> OK, suppose I have a set of beams for you that are identical enough
> that slight discrepencies in shape or size are irrelevant.

It's not irrelevant and it's not theoretical. In a real world car assembly line there is a
huge amount of banging, warping, shoving, and finessing to get things to line up properly,
it must be done carefully and intelligently or things will break, workers spend more time
doing that then anything else. They need to do it an equal amount of fudging on the next
car they work on too but not in exactly the same way because the parts are not exactly the
same. The more parts you're using the more variations in part size add up and cause
trouble. If it wasn't for this we'd have had totally automated car factories in the 1950's.
Record the position the robot hand should to in at during every instant of assembly and
just replay it for the next car; it would never work. No cookbook smaller than the galaxy
could contain all the "if then" rules needed, to get the job done you need a brain.
However we know for a fact that a compact cookbook exists to make something far more
complex than a car and it works every time, DNA. The fundamental difference between a
car factory and a womb is the size of the building blocks.

>Organic chemistry is substantially more complicated than Lego blocks,
>even in the machine phase.

Granted it's not quite as easy as playing with Legos but it's much simpler than
most macroscopic object and the countless ways they can interact, like the way
a turbine blade interacts with a jet engine. There are fewer possibilities with atoms,
I might get H2O or H2O2 but H3O is unlikely.

Even more important, chemistry is repeatable. If I overstress a jet engine it will fail
at some point but no two will ever fail in the same way at the same time. The weather
it flies through will never be the same either because some butterfly in Brazil decided
to flap its wings two years ago. In contrast, although the process is astronomically
complex, a protein folds up into exactly the same intricate shape every time over a
large range of temperature and pH found in biological systems. I can't think of anything
made of macroscopic blocks that robust.

        John K Clark

This archive was generated by hypermail 2b29 : Mon Oct 02 2000 - 17:34:37 MDT