Re: Uploading

From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Sat Mar 09 2002 - 13:21:15 MST


On Sat, 9 Mar 2002, Richard Steven Hack wrote:

> Once again, we're assuming the speed of light is an absolute limit.

It would seem useful to believe for most purposes of discussion
that the current physical laws remain in effect for the future
of the universe. If they don't then the universe becomes very
unpredictable -- it might be a much "nicer" place, or a much
more difficult environment in which to survive. There is simply
know way of knowing. If one wants to start a thread stating
some arbitrary collection of physical laws, e.g. FTL travel, wormholes,
etc. then postulate how uploads should behave in such an environment
then feel free to do so. (I'll note the the new Star Trek prehistory
has included some of those elements so it isn't completely boring.)

> That may be true given current scientific knowledge, but that is not
> the same as being an absolute.
>
For the purposes of list discussion, I think its reasonable
to use something like Michael Shermer's skeptic scale. On that scale
I'd put the probability of nanotechnology based uploading in my natural
lifetime at about a 3 and for the younger list members at 1.5-2.
I'd then put FTL travel in the next century at an 8 and wormholes
of any significant use anytime in the future at a 9.

So for "planning" purposes it seems reasonable to assume that you
do hit some physical limits at optimizing the locally available
matter sometime over the next thousand years or so.

> A Transhuman has no need to replicate.

Only if you can absolutely guarantee that you trump the galactic hazard
function. Accidents happen to transhumans they just take longer.
As I discussed at Extro3, only a distributed replicated intelligence
has a good chance of surviving indefintely. Today, given my perspective
on how much more expanded I expect minds to become, I might argue that
a distributed mind with redundant intelligence support systems
might be more accurate.

> Now, it is possible that it may
> turn out that replication is a good move for some other reason, but
> replication is not identity (barring some tech that enables it to be
> identity), therefore replication offers no survival advantage to an entity
> bounded by space-time.

If one can design and build a point-source singularity weapon
(Shermer Skeptic Scale: 5) then it may be very difficult to detect one zooming
in from the Oort cloud at high velocity. You may not have the advanced
notice or the time to do a backup. So if you "confine" yourself to
a single dense location, you still retain a non-zero hazard function.
If, on the other hand, you distribute yourself over the solar system
its difficult to imagine what could cause a total system failure, so
the hazard function is much lower. *But*, you do think much more
slowly due to the propagation delays. If there is an "economy"
(in a traditional sense -- where the entities "value" something
that isn't universally available and will pay matter or energy
to get it), then the most distributed entities with the lowest
hazard functions are perhaps also the "poorest" (because they
probably produce the least quantity of novel information).

> As I say above, replication is not the issue. It is not clear that
> Transhumans need be competitive - that is *human* thinking (and low-grade
> human thinking at that).

Initially, clearly no. But if you allow yourself to make copies
that have autonomy or you allow minds (natural or artificial) to
self-evolve to a state where they decide to make copies and you
don't impose a solar system wide ban on copying (which may be quite
difficult to enforce given the feasibility of designing a self-evolving
seed without copying constraints) then you have a limited window
before you are once again up against the limits. Not only do you
have to outlaw unlimited copying, you have to outlaw unlimited
mind expansion as well. Can't have 1 mind getting 99% of the
available matter & energy and the rest getting 1% can we?

Fortunately our observational abilities may have increased to the
point where we might be able to discover how others may have solved
the problem and have some working examples of good strategies
to use ourselves. At least I hope that will be the case.

Robert



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:12:53 MST