X-Message-Number: 3118
Date:  Mon, 12 Sep 94 03:50:41 
From: 
Subject: CRYONICS -- Determinism


Determinism: some more thoughts
Mike Perry, 12 Sep. 1994

Bob Ettinger and others have commented recently on "determinism" 
vs. "nondeterminism." I'd like to summarize my understanding of the 
main points, then offer some thoughts that others may not have 
considered yet.

In a deterministic universe, the future is completely determined by 
the present. In a nondeterministic universe there are "random" events 
that cannot be predicted from the previous condition of things, e.g. 
some future event could happen that could not be foretold in 
advance, even with perfect information about the state of everything. 
Some (and Bob in particular) find the notion that truly random events 
happen unscientific and (as I gather) essentially mystical, something 
then that would be disallowed by the premise that events can be fully 
explained by some scientific theory. Bob in turn is confident that the 
scientific rather than the mystical holds sway in our existence, thus 
the universe will eventually be found to be fully deterministic. An 
additional property, not guaranteed by determinism, would be what 
could be called "reverse-determinism"--that the past (rather than the 
future) is completely determined by the present. As has been pointed 
out the one does not necessarily follow from the other; however Bob 
has expressed some confidence that even reverse-determinism will 
hold. If this is true it will mean that every person who ever lived can 
be resurrected (that is, we should be able to reassemble the particles 
that composed any past individual in the correct order, or at any rate, 
recover a complete description from which a functioning duplicate 
person could be built). It would also mean that cryonics, in an 
ultimate sense, is superfluous, though it still would presumably be 
advantageous because you would come back *sooner*, and perhaps 
gain other advantages, if your remains had not disintegrated.

Sometimes the opinion is also expressed that *unless* a sufficiently 
complete description of a person can be recovered from surviving 
information, that person is gone forever or cannot be resurrected. For 
those who doubt reverse-determinism, then, cryonics may be the 
only hope for a life after clinical death (though I have a different 
viewpoint).

In any assessment of reality, of course, the final arbiter is reality itself. 
There was a time when the universe did seem unquestionably 
deterministic, but in the 20th century, quantum uncertainty seems, 
on the face of it, to abolish that possibility. A leading theoretician, 
John von Neumann, "proved," on the basis of quantum mechanics, 
that uncertainty was built into the fabric of reality, and could not be 
eradicated. However, eventually this "proof" was shown not to be 
airtight, and under certain conditions determinism could still hold. 
One other property had to be sacrificed, however, that of locality. 
Locality means that a given event is not influenced by events that are 
happening more-or-less concurrently, but at great distances away. 
With nonlocality you have "spooky action at a distance" or 
interactions of concurrent processes that are arbitrarily far apart. 
(The allowable interactions are not arbitrary, however, and do not, for 
example, allow passing of messages back and forth at greater than 
light speed.) By now the spooky action at a distance has actually been 
verified (to most persons' satisfaction), as in the Aspect experiments 
with paired, correlated photons.

The universe, then, may be deterministic. It does not follow, 
however, that we can, even in principle, make arbitrarily accurate 
measurements of some sort, and predict gigantic chunks of the future 
to subatomic precision, or similarly retrodict the past. Where hidden 
variables are involved, they may remain forever hidden. To make a 
prediction it is not sufficient that the determining information exist, it 
must also be available to the observer. It is possible then that 
determinism holds in some absolute sense, avoiding the imputation 
of mystical underlying causes, but in practical terms, the universe is 
randomizing nonetheless. This state of affairs I'll call "inaccessible 
determinism."

I've now surveyed some opinions of others, and will offer some of my 
own. Some of these are rather wild speculations; I hope they will be 
interesting enough to be worth a read.

In general, I lean toward inaccessible determinism. I am optimistic, 
however, on the possibilities of ultimately resurrecting persons from 
the remote past, though guardedly. It's better to be frozen than not, if 
you have the choice. Some interesting possibilities may exist for 
reconciling seemingly incompatible world views such as Many 
Worlds vs. One World. 

On the face of it, it seems too tall an order to me that the entire past 
history of the cosmos, down to subatomic dimensions, could ever be 
retrodicted. I know of no procedure, for example, that seems 
remotely capable of mapping the molecular structure of an ice cube, 
after the ice has melted. Basically, whatever information is 
straightforwardly captured in written records, fossils, etc., is all we're 
ever going to get. No doubt with nanotechnology a great deal of 
previously unknown information will come to light, and it will be 
exciting times, but I doubt if such things as, say, even the names my 
1,000th-generation ancestors will be recoverable. (This would take us 
back to around 20,000 BC, well before the invention of writing!) The 
only hope for resurrecting a person once obliterated, as I see it, is a 
series of lucky guesses that restore the necessary information. This is 
a very, very remote possibility, by any ordinary standards. For 
example, if 10^12 bits are sufficient to characterize a person, and we 
have nothing else to go on, it will take 2^(10^12) guesses to arrive at 
the correct sequence of bits from which a functioning duplicate 
person (equivalent to the original and qualifying as a resurrection, in 
my view) could be constructed. This seems preposterous, but is far 
from ruled out if time and space are infinite. We may never know if 
we have found the correct sequence, but that does not rule out 
finding it anyway. There is more to say on this subject in view of a 
Many Worlds cosmology.

Basically, the viewpoint of Many Worlds is that the universe as we 
know it is far from all there is--there are in fact parallel universes in 
which events are happening that are similar but not identical to our 
own. Whenever a "random" event happens, even on the smallest 
scale, the whole universe splits into near-identical copies in which 
this same event happens in different ways, enough universes to 
completely exhaust all the possibilities. So in reality there is no 
randomness, but metaphysically this is preposterous! What process 
could give rise to whole universes on such short notice? On the other 
hand, who said reality had to fit our preconceptions? Despite the 
metaphysical difficulties Many Worlds has some attractive features, 
e.g. observations can be explained without invoking an observer that 
not subject to quantum processes (a bug with most other theories). 
There are also ways of effectively getting Many Worlds without the 
incessant creation, e.g. supposing infinitely many worlds *already* 
exist side-by-side, some of which are so closely correlated as to seem 
identical, until a "random" event happens differently in different 
worlds, effectively splitting them apart. In addition to splitting apart, 
it appears worlds can be joined together, by what we perceive as 
information losses, e.g. the melting of an ice cube. Suppose at the 
start there are many similar worlds, each with an ice cube. On the 
macroscopic level the cubes appear identical but ultrastructurally 
they are different (and detectably so), effectively separating each 
world from the others, Melt the ice, however, and this distinguishing 
information is (arguably) lost, effectively joining the worlds, or some 
of them.

How would Many Worlds affect the prospects of resurrecting 
previously obliterated individuals? Mainly, by enlarging the 
possibilities for success, and by greatly reducing the time expected to 
achieve any given resurrection. This would follow because (a) more 
than one historical timeline is authentic, and (b) more than one 
resurrection process would be expected to be in progress, if we ever 
reach the point of trying one ourselves. Basically, every world has its 
own, distinct historical timeline. A case can be made in fact that 
substantially every possible history is happening in parallel. So not 
many guesses may be needed (perhaps only one) to recreate a person 
from *some* real past. (And with Many Worlds, there is not just one 
version of "our" past but more than one with equal claims of 
authenticity.) On the other hand, if a resurrection project is ever 
started (as might be the case in a few centuries or millennia) one 
would expect a great profusion of similar processes in other worlds. 
Any given, real person, once obliterated, could then expect to pop up 
somewhere amidst all the attempted reconstructions. To me it would 
seem sensible, not to just try for individuals, but for entire histories of 
individuals stretching far back into the past. More bits to guess, but 
the same considerations would apply. It would make sense (the first 
time around at least) to reconstruct a history that perfectly fit the 
surviving records, which by then we could assume had been totally 
exposed through nano-archaeology.

A resurrection of this sort, i.e. involving lucky guesses, is different 
from one that could be carried out straightforwardly, if we had 
enough information to start with (as we hope will happen with 
cryonics, for example). With lucky guesses we will never "know" 
when we have the person; such persons will be separated 
irremediably from verifiable history. To me that is significant. I 
would rather be part of verifiable history if I could; this essentially is 
what survival is all about. I would settle for an alternative if I had to 
(i.e. if I died and couldn't be frozen); better to be present in some 
sense than not at all, but then I'd want to survive for the *rest* of 
eternity, at least. So it's better to be frozen, than obliterated. Another 
reason for being frozen is that you'd certainly expect to come back 
sooner, even if the parallel resurrections were eventually going full 
tilt, as I've indicated.

There are other interesting possibilities. Those that did the 
resurrecting would presumably be strongly immortalist posthumans, 
who we can assume will come to dominate the future. If I were such 
a being, I might create the *potential* for regenerating an entire, 
multigenerational population of earlier beings, by generating the 
necessary information, long before I'd want to use that information to 
actually resurrect some specific being. The ones I'd be most interested 
in resurrecting early on, by far, would be rationally, strongly 
immortalist themselves. (A few other attributes would help too, e.g., 
decency, kindness, love of fellow beings, etc.) I think I would have a 
lot of empathy with that sort of person, even if they were primitive 
by my standards, but considerably less interest in those who really 
didn't have it together that way. If this is any indication of what the 
future actually holds in store, and you would like to be resurrected 
relatively soon, you'd better be strongly, rationally immortalist right 
now. This means, of course, being a signed up cryonicist! If you aren't 
that way, then you may reappear someday, but rather remotely, and 
in some latter-day counterpart of a mental hospital, where you might 
have to spend a long time (subjectively at least, maybe in actuality) 
before you could reasonably take your place in the world.

Many Worlds is a viewpoint I favor. Partly it's just because I like it--it 
seems to provide the best scenarios of resurrection in the tough 
cases, i.e. after obliteration, assuming the hidden past will never be 
recoverable in the usual sense, which I think is likely. But also it 
actually seems to be viable from a physics point of view, in ways 
most alternative interpretations of reality are not. (Sometimes the 
complaint is raised that no experiment can distinguish between 
Many Worlds and alternatives, but while this has been true up to 
now, there are some apparently viable future possibilities.) Some 
alternatives exist to Many Worlds, however, that avoid the problem 
with the observer I mentioned earlier. One of these that has attracted 
interest recently is a theory of David Bohm (*Scientific American* 
May 1994 p. 58). This is a fully deterministic theory, but has 
nonlocality. It is a One World, not a Many Worlds theory. It is also, 
relative to other, comparable quantum theories, quite simple, which 
itself is a reason to give it  attention. 

So now I'll offer another wild speculation: the (possible) 
reconciliation of a One World, nonlocal theory with Many Worlds. 
Basically, while there is really One World, there are Many Large 
Pieces of that world, connected by nonlocality, and interacting in 
such a way as to effectively constitute Many Worlds. Our known 
universe would be one Large Piece, for instance. Somewhere there is 
something else very similar to that, and somewhere else something 
else very similar, etc. In all of (possibly infinite) time and space there 
are many possibilities. Sufficiently similar pieces would interact 
nonlocally, that is, instantaneously at a distance. The how and the 
why (or the whether, for that matter) I don't know, but I think (and 
hope) this bears looking into.

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=3118