Robin Hanson, <hanson@econ.berkeley.edu>, writes:
> If there is any issue of a tradeoff of resources devoted to launching
> probes and resources remaining, selection would promote devoting the
> resources to launching probes. But if there is some remaining launching
> hardware after all accessible resources are converted to probes, the
> behavior of that hardware seems hard to predict. If there is any way
> that it could eventually get around to sending out more probes, then
> it seems selection would promote that. So we're really talking about
> situations where further probe launch seems impossible.
Presumably even in situations where further probe launches are worthwhile,
there will eventually come a time when the frontier has moved too far
away. Maybe it will be in a hundred years, maybe in a million.
What this should lead to is an "expansion zone" just behind the frontier
in which the contest for expansion is occuring. Here the replicators
would be engaging in the leapfrogging and other strategies which Robin
has described in an attempt to stay at the leading edge of the expansion.
The width of this zone would be the region in which the frontier was
still within range. It might be very large, if it were possible to
usefully reach the frontier from a great distance.
However wide the zone is, it seems that evolutionary considerations
similar to my earlier post would apply. Replicators in the expansion
zone would devote their resources to expansion. Behind the expansion
zone, behavior would be random and unpredictable.
> I think Zindell's SF novel (I forget the title at the moment) imagined
> probes inducing their stars to nova in order to help lanuch new probes.
> Under such scenarios, there isn't much of a launch system left to talk
> about.
Good example. This is from Zindell's series Neverness, The Broken God,
and The Wild. This has the same shortsighted emphasis on replication
without regard to the destruction being left behind.
An implicit assumption in these evolutionary models is that the
replicators are subject to natural variation. In practice however
replicators can harden themselves against variation to any desired degree,
albeit at a cost. Sentient entities may not wish to see themselves
reduced to mindless expansion. They would attempt to prevent variations
which eliminate their mentality.
If the universe is infinite, then any finite degree of protection will
eventually fail, and expansion at all costs will become the dominant
drive at the frontier. Only enough intelligence will remain to execute
whatever strategy is optimal for expansion. Any redundant or unnecessary
appendages will be discarded. What will be left will be "pure expansion",
without thought or care, a wavefront moving through space at nearly the
speed of light. What is left behind will be an artifact of how a replicator
programmed solely for optimal expansion at the frontier happens to behave
when thrown into a non-frontier environment.
Hal
Received on Mon Dec 15 22:06:18 1997
This archive was generated by hypermail 2.1.8 : Tue Mar 07 2006 - 14:45:29 PST