X-Message-Number: 12775
From: Thomas Donaldson <>
Subject: about repairing brains, by nanotechnology or other
Date: Mon, 15 Nov 1999 12:10:44 +1100 (EST)

Some comments for George Smith:

First, as Charles Platt has already explained, our genes give a general
notion of what we're like, not an exact blueprint. This means that we'll
first have to do lots of investigation of the damaged remains of a person
before we seriously try to revive them. The nature of the damage, and
its repairability, will differ with each person: something likely to make
such investigations even more complex.

Naturally we'd be almost entirely interested in the BRAIN of this damaged
person. I'm happy to accept that almost all the BODY of this person (given
that his/her DNA survives without damage) can be recreated much more
easily. That part probably doesn't need even any advanced nanotechnology;
we're almost there already. But it's the patient's BRAIN that is most
important.

Second, if our nanodevices are to repair someone's brain, they just won't
work if they're constantly being damaged by external chemistry. We'd need
to use such devices, or any other means, to work out first just what the
damage IS ... and it's likely to differ between patients. If the devices
keep being destroyed, we'll find it hard even to do this first step. 

Incidentally, any such destruction can be repaired much more quickly if
these devices act more like simple machines and less like complex ones.
Not only would more complex machines be more subject to damage, but they'd
also be harder to make, introduce into the patient's brain, and guide. 
That's one major reason why I think that lots of simple machines in some
kind of solution (note that I did not say water) would be more efficient
in doing any repair. And a half-done repair would be totally useless, and
in some cases even harmful.

Our own cells and tissues contain lots of methods to defeat damage due to
various chemicals, some of them simple wastes, others (perhaps much more
important) chemicals which our brain actually uses but which can also 
cause destruction in the wrong locations or concentration. Artificial 
"enzymes" to prevent such destruction would be needed for any attempt at
repair. And you (and others) do ask a reasonable question when you ask
why enzymes seem so complex. The reason is simple: they work with other
chemicals, just so that they don't ALWAYS work. If they worked with no
control at all, then they'd cause damage, too. Enzymes aren't just
catalysts, though they ARE catalysts. They also contain various structures
which allow their workings to be controlled. (I do not know how much
that structure can be realistically stripped down, but it's very clear
that we need much more than just the active site of the enzyme).

Repair of a patient's brain will be far from simple... and moreover, even
now, there are plenty of things about how our brains work that we do not
understand. Those things are far from trivial: how is our truly long term
memory stored? (We're getting there, and we know a lot more than 10 years
ago, but we still lack a complete story). Before we rush in to do any
repairs, we first need to learn more about human brains IN GENERAL.

One proposal, not totally without merit, suggests that instead of
repairing a patient's brain we instead "read it off", and then create
another with the same memories. This is very easy to say but likely to be
hard to do. The problem getting in the way is that we're not reading off
the brain of a healthy person; we're trying to do that for a very damaged
brain. And so, rather than deal with the issues involved in repair, we
must instead have nanodevices which deal with the issues involved in
working out from the damaged brain just what it was like before its
damage. In practice that will be at least as complicated as direct
repair: if our memories lie in the connections between our different
neurons, any repair must work out what those connections WERE (including
the possibility that some neurons were completely destroyed). Among other
problems, such calculations require the connections of ALL neurons in
order to succeed. They could not be done simply by nanodevices working
inside the brain, though each nanodevice could find SOME connections.
Working out how the brain had its connections would have to be done
separately outside the brain. Moreover, if those calculations gave
several possibilities, the whole operation would have to be repeated
as our nanodevices went in again to look for more possible clues. And
clues, as you should know, won't have any special sign saying so: what
can be a clue in one case could be irrelevant in another.

I could go on here but won't just now. The main point is that a repair
system for human brains will be needed, and it should not be designed just
like we might design nanocomputers (which are far simpler). Ultimately
I even believe that so far as someone CAN BE repaired, we'll work out
how to do it. But even with very advanced nanotechnology (of whatever
kind) it's unlikely to be an easy problem, and will require time and 
specially designed devices.

			Best and long long life to all,

				Thomas Donaldson

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=12775