X-Message-Number: 8536
Date: Thu, 4 Sep 1997 12:42:14 -0400
From: yvan Bozzonetti <>
Subject: Answer to T. Donaldson on uploading.

In Message #8519, Thomas Donaldson said:



>About "uploading":
>
>1. If we don't understand how brains work, then whatever pictures or other
>   information we store might easily miss some very critical things. For
>   instance, it's only very recently that neuroscientists realized that 
>   some gases such as NO and CO could act as transmitters. So just how
much
>  are we losing when we "store" a brain in a computer?
>
We know our brain works with electrochemical signals, even if we don't know
all of them. Now, each electrochemical reaction produces an electric
activity, so we can spy on it with micro-electrodes and copy its results in
an analog or digital system even if we don't know how the biological brain
get the same result. We don't need to know everything from a bird to build
flying machines.
 
>2. I most certainly agree that we must still think about what to do with
>   damaged patients. HOWEVER in that case, the problem of storage is still
>   worse, since we have many different kinds of damage which will affect
the
>   brain of a patient in many different ways. Just what information to
store
>   and what to ignore, and in what form to store it, becomes important.
The
>   fact that some structures, features, or chemicals are NOT there becomes
>   just as important as their presence. The setting: what was done to the
>   patient in the past, as compared to their present state, also becomes
>   important.
>
More time is lost between death and cooling, less you know about past
destructions. A simple system, using  for example dry ice may be more
harmful at first look but it freeze a near living state with some known and
reproducible kind of chemical scrambling. I think with different "surface
coating" we have the same basic view.

>   While we will probably use very powerful computers to unravel all this
>   about a damaged patient, I would prefer to filter (and possibly lose)
>   information at the OTHER end rather than storing the damaged brain 
>   now. Freezing it, with or without cryoprotectant, is one way to store
>   it without making assumptions about what was there when the patient was
>   alive and undamaged. (I make no assumptions here about future methods =

>   for storage. But we must store people now).
>
I think of uploading for running brains on computers, not as a way to store
the brain information today. So, once more we agree.

>3. About the problems of uploading: first of all, as I understand it, the
>   present imaging technology might give a very gross view of a brain,
>   but hardly do so at a microscopic level.

Nuclear magnetic resonance  with high field gradient can see at
sub-cellular level on short distance. to use it would need to slice the
brain, a destructive process I don't follow, even if it is technologically
possible today.

> X-rays don't obviously fix
 >  that problem, since their energy may well damage or change what we are
 >  trying to look at ... assuming in the first place that we could
actually
 >  do the focusing required. We may well need a quite different
technology,
 >  perhaps a system which inserts multiplying bacteria into the brain,
each
 >  one of which attaches to a neuron, maps out its connections, and then
 >  leaves to report its results ... after which a computer puts them all
 >  together (that is a VERY simple statement of something which would have
 >  to be much more complex! and I also ignore many problems with it here).
>
Here we disagree. I think X-ray astronomy may be a good ground to produce
the optical quality we need. About radiation damage, you could see an
article published one year ago in Scientific American about quantum
non-demolition (sorry, I have not the magazine at hand to give the precise
reference!). With it, we can reduce as much as we want the radiation
damage. X-ray scanning don't need any thawing and would not add significant
destructions beyond what is done by poor cryonics, so you could use after
invasive technologies of the nanotech kind.

>4. If you think such patients will require uploading to be revived, I'd
>   point out that even that is far from obvious. Yes, we'd need to work
>   out the structure and chemistry of their damaged brain, in detail. We
>   would also need a computer to figure out the undamaged structure and
>   work out a plan for repair. Whether we then choose to recreate a brain
>   with the former structure and chemistry, or fix the one we have, looks
>   to me like a choice which would depend a lot on several different 
>   things: just how much and how damaged was this brain? What kinds of 
>   brain repair are now possible (ie. when the repair is done).
>
Uploading has two advantages: It need very powerful computer, but other
technological needs will produce them. X-ray scanning may be produced for
other purpose too, from astronomy to electronics componement production.
And here, investments can reach many billions $ per year. When line
definition in computer processors will fall beyond 100 nanometers, X-ray
optics will be the only workable solution. That is not to say we have
nothing to do to see uploading, but at least the basic technology will be
produced in the twenty years to come.

>   As a subscriber to PERIASTRON, you know already that lots more repair
>   potential exists in brains than neuroscientists formerly thought. Just
>   what boundaries there might exist to affect our choice is something
>   which should be settled AT THE TIME OF REPAIR.
>
I think that remark is far more important for the Prometheus Project level
of technology. Self repair is indeed important if the brain (and the full
body) can be started anew as a biological system. That possibility is good
for PP conservation quality, I think it is not if we must count on
nanotech.


Yvan Bozzonetti.

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=8536