[extropy-chat] Name that system

[Robert Bradbury]

On 12/20/06, Jef Allbright <jef at jefallbright.net> wrote:

> Robert, in your posts to Amara and me yesterday your sense of urgency
> and frustration comes through loud and clear.


Some days it rips me apart more than others...

Thanks for the pointer to the work at Nanorex.  I'm pleased with the
> work they're doing with atomic level CAD/simulation software (but why in
> the world did they state performance in terms of hours to process on a
> laptop running XP?), but I would like to see much more empirical work.


To be honest I don't know.  I think it has to do with Nanorex backing into
the atomic level from the nanofactory (replicator) level.  The nanofactory
film was the effort to show the scientists the big picture -- when I think
the scientists need more of the "small" picture (given the extent to which
the Smalley, Whitesides, et al debates may have led people to believe it was
"impossible").

There may also be the fact that the simulation of the neon pump that was
done in the late '90s by the Goddard group did require a non-trivial amount
of "then-era" supercomputer time.  So they may be trying to show how far we
have come (from a software & hardware perspective).

But the next time I talk to Mark Sims I'll ask him why that "figure of
merit" was chosen.

As you may know, until May of 2006 I was a technical manager with the
> world's leading manufacturer of Atomic Force Microscopes,


Didn't know that.  Thanks for educating me.

The pure designs are neat, but I'd like to know more about how
> they'll deal with the inevitable contamination from stray atoms and
> water molecules at various stages of the process.


I know how to solve the problem of extremely pure input streams using
biological pores/pumps.  They can be highly selective and if you simply gang
enough of them together in sequence you can have ultrapure input streams.  I
suspect that RF and RM have thought a bit about the error detection &
correction problem and though I haven't seen them talk about it I suspect
they have some ideas.  There are various DNA polymerases and I think some
ribosomes as well that do "real time" error detection and correction -- its
*not* impossible.

I see some elegant mechanical designs, but I don't see the robustness that
> comes with the organic configurations of nature. Any pointers to relevant
> recent thinking on this would be appreciated.


Until you have a "real" assembler arm design I don't think people are going
to think a lot about having it backtrack (or have it followed by a checker &
corrector arm).  Walk first, then run.

Robert, on what basis do you think you can scale such a project in such
> simple terms?  I don't know what else to say here.


Because in the  Protein Based Assembly of Nanoscale Parts paper I already
made a guess as to the costs of enzyme "robots" to assemble the assembler
arm.  It was *not* cheap.  But given what we are spending in Iraq I'm
rethinking the fact that it seemed prohibitively expensive.  But if we are
"leaping" an order of magnitude in nanopart design complexity over ~13 years
it suggests that we will be well into the 2030's before we have even a
complete nanoassembler arm design (much less a nanofactory design at the
atomic level).  If you had a complete arm design (~4-8 million atoms) you
could simulate it it action.  That would go a long way towards sending the
critics back to the drawing board to come up with some more "buts".

Robert
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20061220/328b4607/attachment.html>