John K Clark (
Fri, 11 Oct 1996 21:02:18 -0700 (PDT)


On Thu 10 Oct 1996 Damien Broderick <> Wrote:

>This fits in with my version of the hierarchical method for
>nano-making a car.

I'm not saying Nanotechnology will only use one stage in manufacturing, just
that the number of stages should not be so large as to lead to a bewildering
number part types, because then the advantage you gain from specialized
manufacturing techniques would be lost in the difficulties of organizing

>[Quoting Kevin Kelly] "organic complexity will entail
>organic time."

I don't see any reason to think that is true. There are no preferred time
frames, things happen at all levels from Pico seconds to billions of years,
and lots of things happen faster than organic reactions.

>I'm thinking of the kind of elaborate and unpredictable
>informational richness of the environment that goes into
>unpacking a genome, which Jacques Mooned spoke of a third of
>a century ago in CHANCE AND NECESSITY.

It's been a few years since I read Monad's book, but I think I know what
you're driving at, and there is more than a little truth to it. If I gave a
Martian nothing but the sequence of the amino acids in a protein, there is no
way he could figure out what it's shape would be, he would need more
information, like the fact that the protein will be in water, and at a narrow
range of temperature, and at a specific acidity. All these things effect the
way a protein folds up. Actually, if I just gave the Martian some DNA he
couldn't even figure out the amino acid sequence of the protein the DNA
symbolized, much less know what its shape would be, because he would need to
know what The Genetic code would be, and that resides everywhere and nowhere.

The genetic code that all life on earth uses is arbitrary. For example,
the nucleotide triplet CAU in messenger RNA symbolizes the amino acid
histidine but their are no special chemical characteristics that relate one
to the other. One type of transfer RNA has an anticodon that connects to the
CAU triplet of messenger RNA like a key fitting into a lock. At another part
of the transfer RNA molecule, an amino acid can be attached, in this case
histidine. However transfer RNA can't tell one amino acid from another,
the amino acid attachment part is IDENTICAL in all tRNA molecules, but in
practice, only those that have the anticodon for CAU are attached to
histidine. The reason for this is an enzyme ( aminocyl-tRNA synthetase).

This enzyme can tell one amino acid from another, and it can tell one tRNA
molecule from another, and it can a attach a animo acid to it. However, this
enzyme does NOT look at the anticodon at all but at another part of the
transfer RNA, the DHU loop. In the lab the DHU loop from one type of tRNA has
been grafted onto another type of tRNA and that changes the genetic code.
It's also interesting that this enzyme is a protein encoded by, what else,
the genetic code. I can't point to exactly where the genetic code is, because
it does not reside in any one of these stages, it resides in all of them.

Having said this however I don't think the situation would be different for
Nanotechnology. If I gave our Martian friend the digital instructions sent
to a assembler he could not figure out what if anything it would make,
unless he knew something about the nano computer's design and what sort of
manipulator arm it had. He would also need to know some of the details the
machine would be working in, like the fact that atoms are so numerous and
move so fast that if you need a hydrogen atom you don't need to go searching
for one, just sit tight and one will come to you. I think the Nanotech
environment will be far more information rich than anything life knows about.

>Despise bio-evolution if you will, there's something to be
>said for letting Vast numbers (Dennett-speak) of variants do
>the walking through design space...

Yes, it has its uses, and until stone tools came on the scene it was the only
way complex objects could get made.

>>If you made up a list that contained the type and
>>position of every atom in a car this list would
>>contain a HUGE amount of redundancy. [Just use] the
>>same sort of algorithms we use today for data
>>compression in ZIP and GIF files.

>Easily said. But does this work with something of the order
>of 10^30 atoms, *each scanned individually*?

The algorithms should work, the problem is the hardware, no computer today
could come close to handling a file of that size.

John K Clark

Version: 2.6.i