[extropy-chat] Why no assembler design?

[Eugen Leitl]

On Mon, Nov 17, 2003 at 10:35:58AM -0800, bradbury wrote:

> I have to disagree with Eugen here.  Robin and I, I believe, have (offlist) agreed

Allright, we're both firmly in belief country here. Future promises to remain
interesting, by keeping providing surprises. 

> upon the problem being in pathways to assembly -- which is probably a problem

I believe I pointed out that the bootstrap and the assembler both need a
virtual drydock to get afloat.

> more in systems analysis (which Eric very briefly discusses in Nanosystems)
> and less of a problem of molecular modeling.

You'll notice that Zyvex does lots of molecular modelling as well as UHV
hardware work. It is very difficult to get convergent self-assembly 
(especially, convergent self-assembly using biomolecules for guidance and ligation, which I believe
you're primarily interested in) right without a good potential. 
It is similiarly difficult to validate your assembler without
going through a full deposition cycle (many deposition cycles, in fact --
though I'm expecting more a continuous-deposition operation by means of
site-specific polymerization, at continuous monomer supply: notice that it
should be enough for self-replication, while being very far remote from being
able to specify the structure arbitrarily at atomic scale).
 
> I address some of the possible costs and time frames in [1].  Robert Freitas
> has reviewed it and while he doesn't completely agree he doesn't disagree.
> 
> One of the major problems is understanding that nanoassemblers have to
> be constructed from nanoparts.  First we have to solve the problem of
> the design of nanoparts -- that is in part what Nano at Home (www.nanoathome.org)

Parts are molecules; there are molecular builders aplenty. What's missing is
macro ability, and ability to work on very large systems at atomic detail in
the hot spot (area of current edits). 

OpenGL nor sphere rendering doesn't scale here, I'd prefer a voxel approach
to the volume and part manipulation (3d BitBlt for dragging with collision
recognition), with sphere and wireframe rendering of highlighted areas.

This should render realtime for 0.1-10 MAtom, and allow realtime minimization and
MD dynamics.

> is all about (it could use some programmers that want to join one or more of
> its teams).  Second we need to solve the assembly of nanoparts.  Currently

The self-assembly approach includes mutual recognition of complentary-surface
encoded parts. The nanorobotics approach is pick up and place, with adhesion
by van der Waals and condensation of prepositioned glue spot bond patches.

> that means either a biological paradigm (e.g. self-assembly of viruses) or
> a "macro" assembly paradigm (e.g. robotic assembly in factories scaled
> down to micron and nano scales [what Zyvex is doing]).  I try to outline
> some paths for all of this in Appendix B of [1].
> 
> Robert
> 
> 1. Bradbury, R. J., "Protein Based Assembly of Nanoscale Parts"
> Normally at: http://www.aeiveos.com/~bradbury/Papers/PBAoNP.html
> May be available later this week.  The Google cache may or
> may not have an up-to-date copy if you search for it.

Robert, how large is your site? I'd like to mirror parts or the whole of it.

-- Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07078, 11.61144            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
http://moleculardevices.org         http://nanomachines.net
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20031118/911f26dc/attachment.bin>