Borganization

From: Dan Fabulich (daniel.fabulich@yale.edu)
Date: Tue Aug 29 2000 - 19:53:22 MDT


My biological body, as such, will die. It may be the case that I can
store a copy of it elsewhere, and use that copy to reconstruct me when
my current body fails. But even if that happens, the copy will not be
exact; the bodies will be different.

The information stored in the body, of course, need not die. It can
be replicated exactly, backed up, preserved. The information has, if
not an infinite lifespan, then a very very long one.

If you think that you are your current body, if you think that you die
if your body dies, (even if there's another copy on hand or ready to
go,) then the fact that the information stored in your body can be
preserved will not lead you to think that you have a chance at
immortality.

With that having been said, people have taken great pains to achieve a
lot less and called that immortality. Immortality in one's works, for
example. In one's family line. In the memories of others. While I'm
not suggesting that you HAVE to call it "immortality" if I make a copy
of all the information stored in your body, I think you do have to
admit, it's the closest you'll ever be able to get.

This leads me to conclude that whether you believe that you are your
current body or not, your best chance at immortality in the material
world is the recording and storing of your body in a replicable form,
because your current body will die.

When we can figure out how to do this, we'll be right near
Singularity.

---------------

I've heard a number of different charts to a variety of different
things that might be called "Singularity." There are large deviations
between them; one of the largest rifts is the divide between those who
think that we should build an AI to figure out how to do the relevant
recording, (or, to be more precise, build a seed AI and hope that it
performs the relevant recording), and those who think that we should
try to upload ourselves without help from AIs; that, if we do employ
AIs to help us, we should only use those AIs stupid enough to be
completely controlled by us and to serve our bidding unflinchingly.

(We might also try a little bit of simple biological enhancement to
hurry us along either path.)

Either route will require a lot of thought, and a lot of intelligence.
We have market mechanisms and governments to coordiate this work for
now; this might be sufficient to bring us a seed AI before we blow
ourselves up, if seed AIs are pretty easy, but it's not enough if seed
AIs are very hard. Uploading is hard by everyone's estimates.

We'll get around to solving one or both of these problems eventually,
but as far as I'M concerned, this problem has a time limit. If this
problem isn't solved within my lifetime, I have no use for the
solution. If it isn't solved before we blow ourselves up, then it's
no use to anybody.

Other than biologically enhancing humans (non-PC, forbidden almost
everywhere), acquiring funding (either by performing useful work or by
begging for it, even begging the government), or actual lab work, what
more can be done to bring one of these two about SOON?

The answer is borganization, in which one or more intelligent people
get their heads together and think as a single entity of unified
intelligence. (Some cybernetics are probably required, but certainly
not the clunky stuff you saw on Star Trek.) The market and, to a
lesser extent, the government, is designed to coordinate our efforts
in this way, but it does so slowly and inefficiently. Direct neural
inputs and outputs coupled with even a mechanism as simplistic as an
e-cash market would lead to thoughts and conclusions many times faster
than our society can output. Everything would happen faster,
including our science.

If we're trying to build a seed, borganizing will get us there
faster. If we're trying to upload, borganizing will get us THERE
faster, too.

Is borganizing easier (or faster) than simply uploading, or than
building a seed? I very much think so in the case of uploading, and I
largely believe so in the case of building a seed. Uploading already
looks like it will be too hard a project to win the race against my
death. If it requires nanotech, then we risk that nanotech being
abused before it can be used effectively to upload me. If uploading
does not require nanotech, then the sort of technologies which would
be required for borganization seem to me to be enabling technologies
for uploading.

As for building a seed, I happen to hold a fairly pessimistic view on
the question of how long it will take before we get our seed; I worry
that we won't get it before we blow ourselves up. My pessimism aside,
it should also be noted that borganization, if not called
"borganization," is also a substantially lower future shock level than
seed AI, especially when you see it as a simple trend of accelerating
communications. This is mostly based on hunches, which is why I
suggest that I only "largely believe" that borganizing will be easier
than building a seed.

So, right now, I think borganization is my very best chance to survive
the death of my body. What do you think?

-Dan

      -unless you love someone-
    -nothing else makes any sense-
           e.e. cummings



This archive was generated by hypermail 2b29 : Mon Oct 02 2000 - 17:36:46 MDT