X-Message-Number: 12760
Date: Fri, 12 Nov 1999 15:08:53 +0100
From: Henri Kluytmans <>
Subject: Re: CryoNet #12758

-->Hi,

Sorry, but I have to drop in, I myself am a "nano-optimist", and I am
seeing some 
statements that show a poor understanding of nanotech (from George Smith, 
sorry George .. :-> ).

>In  Message #12756, Thomas Donaldson wrote:
>> About nanotech devices:
>>
>> The very first point I want to make is that we'll need much more than
>> nanotechnology to do anything significant in the world. Nanotechnology is
>> a way of acting on the world, but before we go out and act on the world
>> we must first figure out what it is that we want to do. That alone is
>> a big job, even if once we do so we can then do what we want instantly
>> with no expenditure of time or energy.

But artifical molecular devices will also render us the tools to figure out
much 
faster how the world of molecular biological devices is constructed. For
example 
with nanotech molecular disassembling systems and systems to determine 
molecular geometrical shape of objects, it will be much easier to find out
how 
a protein is shaped and how it performs its function. 

Furthermore is not necessary to understand the complete macroscopic system 
to be able to repair at a microscopic level. For example, with regard to do
repair 
work at a damaged frozen brain. It is not necessary to understand the
complete 
functioning of the brain, with all its interactions and interconnections.
It is sufficient 
to understand the (healthy) structure of it's constituant parts, i.e. the
neurons. 

With molecular disassembling tools it should not be so difficult to figure
out the 
molecular structure of healthy neuron. And there are only a limited number of 
different neurons. (And those will even resemble each other very much. In
fact 
there only a thousand different types of cells in the human body.)

Attention, we do not even have to understand the functioning of a healthy 
neural cell. We only have to know it's healthy structure, to be able to
restore 
it to its healthy state!

And also biological repair systems seem to be able to repair the body (to a
certain 
degree) without have any knowledge about the complete system at the
macroscopic 
level. Therefore isnt it logical to expect intelligent repair systems to do
better. 
(intelligent, like having a small computer available)


Then George Smith replied: 

>Follow the blueprint!  The Human Genome Project is projected to be done well
>within two years.  Once you have a blueprint you need only make simple
>devices to follow them.  Because of the July 20, 1999 nano computer
>breakthrough, READING, COMPARING and FOLLOWING the blueprint will be made
>easy.

The problem is, that we cannot directly interpret this Human Genome
"blueprint" (yet). 
Yes, we can read the base sequences, but that does not mean that we know 
how the complete procedure, from DNA to creating a complete organism works. 
We cannot even predict how a single protein folds. And we will need to know
that 
because its 3-dimensional geometry will determine how it acts.

It seems like the "nano computer breakthrough" of this July 99, is being
extremely 
overoverrated by some people. This is only a very small step towards real
molecular 
manufacturing. What would be a nanotech *breakthrough*, is the creation of
the 
first (crude) assembler that can replicate itself. Even if it is only a
simple molecular 
robotic arm.

>> 1. We want to put together our nanomachines out of smaller, basic
>>    machines. It may well be that each basic machine will fail only in the
>>    most extreme situations, but a nanomachine made out of many basic
>>    machines becomes much more subject to failure: only one of its
>>    parts may fail, but by failing make it impossible for any other
>>    part to work. (To say that failure is impossible ignores such things
>>    as the possibility that a radioactive component slipped into its
>>    manufacture, and after decay disrupts it... and lots of other ways
>>    manufacture might sometimes go wrong, even if only rarely).
>>
>>    This problem occurs also if we use MANY nanomachines. Even if the
>>    probability of failure is very small, it will increase the more
>>    nanomachines we use. Depending on how we use them, that failure
>>    may be minor or catastrophic ... just like any other devices,
>>    actually.

First, a nanomachine can be made more robust against single 
point failures by redundancy. Nanomachines will be less frequently damaged 
then biological machines because most of their internal environment is 
predetermined, and no damaging chemical reactions should take place. 
An exception are the interfaces to the outside environment (like the device 
to gather feedstock or other molecules, i.e. the filters),  but these parts 
will thus be added with multiple redundancy.

(Look at the excellent example of the artificial repriciote at this URL 
:  http://www.foresight.org/Nanomedicine/Respirocytes.html
The number of filtering systems in the design has a very large redundancy.)

The damage to nanomachines that will render them useless, will be 
predominantly induced by radiation and molecules from the outside 
environment. The error rates of the nanodevices because of these 
causes can be easily estimated and does not seem to be a problem.

But if not, than the redundancy (and selfrepair capacity) can always 
be extended. And artificial nanomachines should always be smaller and 
more efficient then organic machines (that perform the same function) 
because their internal environment is chemically predetermined and 
therefore much less prone to failure than their organic counterparts. 


>Yet if the CREATION of new nanomachines is just as easy as anticipated, the
>needed working life of any machine might only need to be a few seconds.
>Drexler addressed this issue quite well, in my humble opinion.

A lifetime of a few seconds is far too short for nanomachines to do 
repair (or other) work in the human body. It will be more like hours or days.

>> 2. To be useful for ANY purpose, our nanomachines cannot act totally
>>    separate from everything else. They must act on something else in the
>>    world, something which is NOT a nanomachine. And just as happens in
>>    biological creatures, that opens things up to lots of EXTERNAL
>>    disruptive influences. That's why our cells make various antioxidant
>>    biochemicals to protect themselves from the disruptive effects of
>>    oxygen, to give a simple common example. If they could come into
>>    contact with oxygen only when they "needed to", all that apparatus
>>    wouldn't be needed.

As  I pointed out, in contrast with biological systems only the outside 
interfaces of a nanomachine are subjected to disruptive chemicals. 

<SNIP>

>>    And yes, certainly, we might do many things to protect them from
>>    disruption. But no method is likely to protect against ALL POSSIBLE
>>    disruption.

A certain amount of failure rate is acceptable.

>If you have the blueprint, the operator's manual, the switches and the
>repair tools, YES YOU CAN protect against all MEANINGFUL disruption.  The
>big picture is the big picture.  The devil is NOT in the details.  The devil
>was in the details until July 20, 1999 made it possible to read the script
>whoich contains the details.

Uhh ? ... Having a blueprint of the environment does not render your devices 
resistant against damage.

>>Moreover, all the extra machinery to protect against
>>    disruption makes the devices bigger... in some cases, probably big
>>    enough that they are no longer nano at all.

But always smaller than their organic counterparts!


>> 3. Nanotech devices do consist of about 90 different kinds of parts.
>>    HOWEVER so does everything else. We call those parts ATOMS, and however
>>    we make our device we must use some form of chemistry to do so. One
>>    feature of chemistry is that our atoms don't necessarily stay where
>>    we put them.
>>    Sure, with nanodevices we might make stable chemicals
>>    which would not have been easy to make otherwise (though chemists even
>>    now work on ways to synthesize the wildest chemicals). Even so, that
>>    chemistry provides one way in which external influences can break
>>    a nanotech device ... perhaps, when we actually have them, the most
>>    common way. Not every atom happily combines with every other, and
>>    keeping a chemically active molecule from doing what it wants can
>>    be certainly be done, but once more requires extra atoms solely for
>>    that purpose. Protecting our device from other active chemicals
>>    nearby isn't easy either, and adds more to the device when we do
>>    it. Among other problems, not every chemical is stable at the
>>    temperatures, light levels, and vibrations we might want to use it.

Certain failure rates are acceptable. At least the chemical structures in 
nanodevices will be a lot more stable than those used in organic systems. 
Compare the structure of diamond to proteines. Especialy, the failure rates 
because of internal structures being not stable, will only be a very minor 
cause of errors in nanomachines.

>> These 3 points should be thought about whenever we try to build a real,
>> working nanotech device. I personally suspect (but no, I cannot now
>> PROVE) that our biotechnology has the form it has (lots of very small
>> nanotech devices working in a fluid, rather than put together into
>> larger machines) comes exactly from means to avoid these problems.
>> (And please note that I'm not saying these machines must be the same
>> as enzymes, or that the fluid is water).

To me it seems that a system constructed with an internally atomically 
precise predetermined environment, and furthermore constructed of 
very strongly bonded molecules (much stronger than organic molecules) 
must be a lot more stable, efficient, and robust then a system which is 
based upon statistical behavior and sloppy loose chemical bondings.


>Now that we KNOW we can build a nanocomputer to read and direct the DNA
>blueprint, the rest is just a matter of proceeding.

Uhh ?

>> So these are some points for those who think nanotechnology alone will
>> solve all our problems.

Nanotechnology will not solve all our problems. But it will solve most of our 
material problems!

>Thomas, I thank you for your efforts to try and identify what could go wrong
>with the nanotechnology effort to make cryonics work (as well as radically
>improve our entire way of life).  It's just that it seems clear to me that
>these particualr points have been made obsolete in view of the July 20, 1999
>breakthrough and the coming completion of the Human Genome Project.  I also
>believe that Drexler's books and other discussion have addressed these
>issues extremely well.

As I said before these are no breakthroughs concerning nanotechnology, i.e. 
artificial molecular machines.

Grtz,
>Hkl
=======================================================================>
>Hkl  ------> Technology & Future at             http://www.dse.nl/~hkl
Transcedo --> Dutch Transhumanist Society      http://www.transcedo.org
Because the future is where we will spend the rest of our lives ...

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=12760