X-Message-Number: 33051
Date: Wed, 10 Nov 2010 11:13:08 -0500
From: "Perry E. Metzger" <>
Subject: More nanotechnology talk
References: <>

> From: Gerald Monroe <>
>
> Perry : I will concede that you are probably correct.
>
> As for calculations : all my paragraphs of text are mainly on the
> fact that 1.  If you need a few atoms per bit stored or processed,
> you aren't going to be able to do the calculations with robots that
> only mass a few hundred daltons.
[Corrected in a subsequent message to kilodaltons]

A single DNA nucleotide has an average mass in the vicinity of 330
daltons. A pair is 660. The smallest viruses out there are DNA viruses
with about 2000 nucleotide-long genomes -- already over a million
daltons, not even counting the capsid which is large. Ribosomes
typically mass about 3 million daltons.

It is possible that a device of only a few hundred thousand daltons
could be built with some useful behavior, perhaps the way that
ingenious small clockwork devices were built hundreds of years ago
with some limited useful behavior, but I don't think it very
likely. Certainly general devices will not have masses as small as you
suggest.

A simple google search could have told you this, of course.

> This is a fundamental assumption being that while you can probably
> use a single atom as a transistor,

I've seen no one make any such assumptions in the literature. Maybe it
is possible, but I don't know how you would do it, and I don't know
how it could be done. (I could wildly speculate, perhaps, about one
atom forming the gate of a FET given a lot of atoms performing other
tasks, but it seems like a big stretch to call even this a single atom
transistor, and I don't know at all that such a design could be made
to work.)

At this point, most of the engineering studies done use rod logic
anyway. I doubt we'll actually build real machines with rod logic, but
we can at least analyze it. We don't know enough to design good
molecular electronics at this point -- this is a task to be done in
the future, when there are far more minds working on this problem
domain.

That said, a number of engineering studies have been done by Robert
Freitas that cover how to make autonomous devices on the order of cell
size for medical applications -- there is more than enough space in
such a volume for adequate and multiply redundant control systems.

I suggest reading the literature.

> It would be difficult to even control such robots if they were
> inside a frozen brain that you don't want to make any inadvertent
> changes to.  (so you can't use intense beams of RF energy for
> communications and intense magnetic fields as an induction power
> source.

You might try reading papers on what people have actually proposed for
such systems. These problems and more that you aren't thinking of have
been considered. FYI, no one suggests communications or power via the
mechanisms you are speculating on.

> How can the robots even get into a cell without tearing a hole?

You clearly can't enter a cell without producing a hole in it -- cell
membranes are not Klein bottles, so getting from the outside to the
inside requires a hole. The question is how you make a hole that is
either easily repaired or which self-seals. Again, this has been
considered before, in great detail. Freitas' "Nanomedicine" is quite
comprehensive.

My overall message is this: there has been work on this topic over the
last couple of decades. I will admit that there hasn't been nearly
enough work, and that we need orders of magnitude more people working
in the field to bring it to realization, and that it is almost certain
that some fraction of the existing work contains errors, but we are
not dealing with a blank slate. It is useful to read what has already
been worked out in detail before speculating on how designs might
operate.

Perry
-- 
Perry E. Metzger		

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=33051