X-Message-Number: 21009
Date: Thu, 30 Jan 2003 13:03:33 +0100
From: Henri Kluytmans <>
Subject: biological systems more efficient than MNT systems?

Thomas Donaldson wrote :

>I believe I pointed out that chemists were already studying ways to
>make hydrogen more compressed and more manageable in a previous 
>message. Think borohydrides.

Ehh ... where did I mention the storage of hydrogen ?

Yvan and me were debating about an hypothetical artificial device 
made by advanced MNT and which should be able to function as a 
replacement for the human red blood cells. Such a device will 
need to store oxygen and carbondioxide in storagetanks on board.

>As for nanotechnology, living systems can only be used so far. 
>They do not extend to the systems you (or NANOSYSTEMS) envision. 

Did I imply that ?

<snip>

>Perhaps your NMT will someday really come, perhaps living things
>when examined carefully will tell us why it will never come. 

There is no 100% guarantee that MNT concept will be possible exactly 
in the way as it is foreseen in current conceptual designs. It 
could be that certain details will have to be altered to make them 
functional.

But that the complete principle of an artificial technology using 
systems that can build atomically precise objects from molecular 
building blocks (including self-reproduction) is impossible seems 
highly unlikely to me...

It would be extra-ordinary exceptional if the biological way would
be the only way, or even the most efficient way. Especially because 
biological systems developed themselves by undirected mutation and 
selection. And directed design, in principle, will almost always be 
able to make a better system.

========

>Among other points, if you have LOTS of molecules ways to use 
>randomness to change them to suit may turn out much more efficient 
>than devising special systems which move every one where you want it.

Are you trying to explain here that biological systems are more efficient 
in building objects from molecules than future artificial systems?
And that this is due to the fact that biological systems employ 
brownian movement of molecules whilst artificial systems will use 
mechanically controlled systems to transport molecules.

You forgot to mention in what ways they are more efficient, but I 
guess you are referring to the energy usage. 

Indeed, it seems obvious that biological systems are more efficient 
regarding energy usage in building objects from molecules. Already 
only the fact that the current MNT designs are using much stronger 
chemical bonds (e.g. diamandoid instead of proteins) seems to make 
that likely.

However future artificial systems will certainly be more compact, 
faster, stronger, etc... than biological systems (i.e. more 
efficient in volume, time and stiffness).

The research Robert Freitas has done in investigating artificial 
replacements for some biological functions seem to show an 
improvement of many hundreds or even thousands of times in 
performance of specific functions (see his examples of 
artificial respirocytes and clottocytes at : 
http://www.foresight.org/Nanomedicine/index.html ).

Although it seems clear that biological systems are more 
efficient (regarding energy usage) in building objects 
from molecules it is not clear that they also more efficient 
in performing their specific function. 

====

For example the human brain is using about a 100 watt 
to perform it's function. The computational power of the 
human brain can be estimated at 2 * 10^16 cps (conservative 
estimate by Ray Kurzweil).

A computing system based on nanomechanical rod-logic of 
this same processing power will consume only about 2 Watt.
(10 Watt for a 10^11 MIPS computing system, see Nanosystems.)

And I didn't even take into account :

1) that the computing power estimate for the brain used here 
was a very conservative one, there are estimates as low 
as 10^13 cps, which would result in an MNT equivalent systems 
using only 1 MicroWatt !!!

2) the usage of more efficient architectures and reversible 
logic (not for all computations) to reduce energy usage even 
more.

So it's not that clear that biological systems are more 
efficient in energy usage in a general sense. 

=======================================

Furthermore MNT systems could be designed that would use 
different kinds of building blocks than current designs 
(which mostly use diamondlike C-C bonds). Building 
blocks using the same or even lower energy per bond than 
proteines are possible Also transportsystems using 
Brownian movement could be implemented. Of course such 
transportsystems will be much slower, because the system 
will have to wait until the required part arrives at the 
desired place. However an artificial MNT system, building 
atomically precise objects from molecular parts, using 
less energy than biological systems seems feasible.

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=21009