From: David McFadzean (david@lucifer.com)
Date: Wed Dec 03 2003 - 07:50:30 MST
The origin of Moravec's "pigs in cyberspace" meme concerning the
simulation argument...
Source: CRYONICS Volume 13(12) DECEMBER, 1992 Issue 149
URL: http://www.alcor.org/cryonics/cryonics9212.txt
Vector: hkhenson
Future Tech
by H. Keith Henson
Time and again?
The Extropian mailing list is always a source of ideas for these
columns -- though with the traffic sometimes running to 100 messages a day
it can be overwhelming. A substantial percentage of the people on the list
are signed up with Alcor or ACS.
Someone posted a speech by Hans Moravec (author of "Mind Children" and
well-known robotics researcher) which concluded with a section about a
distant and processor dominated future. A list regular picked out this
part near the end.
"If these [future] minds spend only an infinitesimal fraction of their
energy contemplating the human past, their sheer power should ensure that
eventually our entire history is replayed many times in many places, and in
many variations. The very moment we are now experiencing may actually be
(almost certainly is) such a distributed mental event, and most likely is a
complete fabrication that never happened physically."
. . . and asked if Hans was serious. And if he was, how did he arrive
at "almost certainly?"
I replied:
"This is one of those "Ha, Ha . . . . only seems unavoidable." Chances
are he thought it up himself, but I have long expressed (and may have
mentioned it to Hans) the concept that our remote decedents will simulate
history with many variations just as we have programs today which simulate
the early evolution of life, or the Society for Creative Anachronism
reenacts Medieval battles. Given a universe almost entirely converted to
computational and memory elements, simulating the 20th century down to and
arbitrary level of detail (well below that noticeable to the agent/actors
inside it) would be a project for the Speilbergs and Lucases of the day.
"In fact, it seems inevitable that they would rerun it more often than
we have rerun 'Casablanca.' Given a lot of time, they would run it vast
numbers of times -- making the chance that this is the very first and
'real' running of history negligible.
"To an actor inside such a reenactment, there is no way to tell it from
the 'real thing.' I try not to let such a thesis effect my actions,
because we have to assume that this may indeed be the first time, and even
if it is not, we should do the best we can with the cards we have to play.
But it still bothers me when I catch the stage hands out of the corner of
my eye changing the scenery. :) "
Another reader took this last line a little too seriously -- an ever-
present danger in this medium. My reply to him was:
"Before this drifts off in a direction where I am accused of being more
mystical than is justified, let me repost the sentence:
> But it still bothers me when I catch the stage hands out of
>the corner of my eye changing the scenery. :)
"The grin emoticon [:)] is meant to negate the last sentence!
(Emoticon = emotional icon.) Perhaps I should have used an alternate
emoticon, :^), tongue-in-cheek, because this is certainly my feeling about
this subject. I do find it rather amusing (for all its complete
uselessness) that what we see as ultimate high tech folds back on itself to
generate such a weird mystical "inevitable" conclusion. I can take it one
step further: If we miss making it to the high tech future (especially if
it is for this stupid reason) then the reality we have is all we get -- and
our times will never be played again. So believe in, take care of, and try
to improve the reality you have -- it may be all there is."
The thread went one more round with someone asking why I consider this
a stupid view. My reply was:
"The view is not stupid, as I mentioned, it seems sort of hard to
avoid. But it is stupid to turn this view into a reason not to work on
getting to an interesting future (i.e., because we might already be
there)."
Hans also replied to my first posting:
"Keith and I were both at the first Alife [Artificial Life] conference
in Los Alamos in 1987. I had with me an early draft of "Mind Children."
One of the newest ideas in it (evolved out of some time-travel
speculations) was resurrection by future super-archeologists, and the
implication that this very moment might be a historical reconstruction. I
mentioned this to Keith in passing while we were having a mild argument
about something else (maybe the merits of cryonics: it was too low tech
for my taste: resurrection by such a crude route was likely to by painful
and imperfect, so I was willing to wait, even a long time, for a more
sophisticated solution, one that could work from more diffuse data than a
frozen body, like the traces you constantly leave as you live -- patience
comes easy to a dead person!). Anyway, as soon as I described the
resurrection idea, Keith said, yeah, yeah, I've always thought this moment
exists in millions of instances. Outbid, I quietly nodded my head, with
the idea fixed more firmly than before. Since then I've had several
interactions with Frank Tipler, whose Omega Point speculations are way
beyond my own. He's writing a popular book expounding on it, and its many
interesting implications. I use OP, in an agnostic spirit, to finish out
my own forthcoming book. -- Hans Moravec"
Some of you may recognize these
ideas as rather kindred to those of Mike Perry. I replied:
"It is trivial to work up counter-arguments. For example, while our
motivations seem to include a strong component of interest in history, this
might not be true of our future selves where we have messed with our
motivations. Perhaps reconstructions of the past would be so painful to
the inhabitants of the future that there would be very strong social
pressure not to do it. (See the end section of Marc Stiegler's "Gentle
Seduction.")
"In reference to cryonics being low tech, as one on the "wetwork" team,
this end of it sure is painful and imperfect -- though certainly no worse
than the only available alternatives! There is, however, no reason to
believe that the other end of the process should be painful, and it should
be perfect to the limit of the available information in the frozen patient.
"As far as working from the traces left behind -- well, maybe. I could
imagine a process where some ambitious grad student was trying to make a
minimum error "reconstruction" of the historical Hans at the point he
finished 'Mind Children.' So he simulates Hans and the complete
environment in which he grew up, does a comparison between the original
book and the reconstruction's version and iterates the process till there
are few or no text differences. I hope the temporal version of the Humane
Society would make the discard process painless, but how many versions of
Hans would have to be discarded before this process converged? (Assuming,
of course, that it would converge!) Of course, the process would have to
be a joint reconstruction of editors, authors, and (in many cases) the
typesetters who introduced some of the typos.
"I suspect, however, that the above process is unworkable no matter how
many resources are poured into it. Chaos makes it impossible to predict
beyond certain horizons in the future direction. The inverse of this
should make it impossible to tell which of a multitude of pasts led up to
the present.
"Following [list member] Perry Metzger's lead, I won't preach either.
I think our world will be less interesting for the decisions of Heinlein,
Moravec, and innumerable others who turn down the cryonics option, but it
is their decision. All I can do is be appreciative of those who are trying
to make it."
There was lots more to these threads, about 15 times as much material
as I have put into this column.
Next time I might report on building and debugging the controlled neuro
cooler -- especially if we get it working right. And some time I should
write a column on reworking big planets, stars, and black holes into
habitats.
--- To unsubscribe from the Virus list go to <http://www.lucifer.com/cgi-bin/virus-l>
This archive was generated by hypermail 2.1.5 : Wed Dec 03 2003 - 07:50:31 MST