From: Billy Brown (bbrown@transcient.com)
Date: Mon Feb 21 2000 - 20:28:48 MST
Eugene Leitl wrote:
> Uh, do you you have actual numbers for this brash assertion? Are you
> claiming that essentially all patients are suspended after one hour or
> less after death has occured (not pronounced)? Don't think so.
Heavens, no. I'm merely claiming that it is hard to die in such a way that
you can't be suspended. If you die of an illness, you'll probably be at a
hospital when it happens. If you die in an accident, odds are there will at
least be paramedics on the scene by the time you go. Since they do check
for organ donor ID, that gives you an excellent chance of ending up frozen.
Of course, you could always get buried in an avalanche, or die while scuba
diving, or otherwise manage to evade the emergency medical care system, but
the number of people who go out this way is a very small fraction of the
total death rate.
> This sentence is insoncistent. Clearly there is physical damage that
> is irrepairable (say, incineration, or a week stored at RT) which
> destroys essentially all information about a particular individuum.
My point was simply that information preservation is the only thing that
matters. Damage that makes you look really messed up, but does not actually
destroy information, won't affect your chances of being reconstructed by
mature nanotechnology. This is an important point because information loss
is not visible to the naked eye, and it is quite easy to rack up lots of
nasty-looking but irrelevant incidental damage.
> > Now, what people tend to overlook is that the problem of deducing the
> > original information content of a scrambled brain is isomorphic to the
> > problem of deducing the information content of an encrypted
> message. Given
>
> Another assertion. Based on what?
I thought that one was obvious. Well, let's try a comparison:
When decrypting a message, you have an unknown body of information (the
message) that is known to be organized according to certain rules (the
language of the message). The information has been subjected to some
complex series of nonrandom transformations (encryption), and the object of
the exercise is to reverse those transformations and recover the original
state of the information.
When reconstructing the mind of a cryonicist, you have an unknown body of
information (the patient's memories and personality) that is known to be
organized according to certain rules (human neurology and psychology). The
information has been subjected to some complex series of nonrandom
transformations (ischemic injury and freezing damage), and the object of the
exercise is to reverse those transformations and recover the original state
of the information.
Do you see a problem with one of the comparisons here?
> Sure, you're already immortal. Why bothering with a suspension at
> all? Future science will reconstruct yourself from trajectories of
> atmospheric CO2. Or from a piece of apple pie, whatever is closer at
> hand.
Yeah, right. Thermal noise *is* random, and no known or projected
cryptographic technique would be capable of reversing it.
> Yeah, right.
> I knew cryonics was a religion, but I never saw it demonstrated
> so clearly.
Come now, Eugene, you can do better than that. We've got plenty of
experience with advanced cryptography to draw on, thanks to the cold war.
We've had people break seemingly solid codes time and again, by exploiting
the tiniest nonrandom features of the encryption technology. I'm merely
claiming that we can apply the same mathematical techniques to the task of
deducing the original state of a frozen brain.
Ischemic injury, AFAIK, is a very complex but highly ordered cascade of
chemical reactions. It seems perfectly plausible that you could simulate
the process in reverse to figure out what the original tissue looked like.
The same is true of freezing damage. Using cryptographic analysis is just a
different way of performing the same computation, like using a neural net
instead of a procedural algorithm.
Or maybe it was the rock-on-the-head comment that got you? But all we're
talking about there is low-velocity collisions - we could simulate that now
if the number of particles involved weren't so large. It only gets tricky
if there is enough energy involved to give you chaotic phenomena like
turbulence, and even that might not be a show-stopper. After all, we've had
at least one code broken because it used local barometric pressure readings
to seed its random number generator, and weather is pretty darn chaotic too.
Billy Brown
bbrown@transcient.com
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:26:57 MST