Re: poly: Solar system development

From: Ralph C. Merkle <>
Date: Sat Dec 13 1997 - 16:28:42 PST

Some comments on the colonization of the universe:

1) That which breeds faster and moves faster and can hold what it seizes
2) Breeding rates greatly in excess of human are feasible, speeds close to
C (with some debate about exactly how close) are feasible, military
machines greatly exceeding human capabilities are feasible.
3) Colonization of the universe by some form of machine intelligence at
C-epsilon will be initiated in the not too distant future (likely some
modest number of decades).
4) Assuming that the machine intelligence has sufficient wits to deploy
the resources it acquires in a way which lets it defend what it has
acquired from entities which have much less in the way of resources, this
will establish the initial state of the entire universe.
5) If we play our cards right, the machine intelligence will be compatible
with the continued growth and development of lovable us. If we don't....

There are some 10^22 stars in the universe. Presumably, there is some
not-too-far-from-optimal colonization strategy. Ideally, we'd like to
simply launch 10^22 seeds directly at the 10^22 solar systems at C-epsilon.
 Mass/energy limitations in the launching solar system might preclude this
direct approach (as well as uncertain information about the current state
of affairs near the destination), in which case we'd be forced into a
multi-hop strategy (where each hop involves: launch a wave of seeds, each
seed converts the available mass in the region where it lands (presumably a
solar system) into more seeds and launchers for same, then launches them).
I don't think many hops would be required to reach 10^22, as the "fan-out"
from each hop can be very high.

Example: a "first wave" is launched from our solar system at 99.9% C, and
is aimed at stars approximately 1,000 light years from us. After a
one-year delay it will no longer be possible to catch this first wave, even
in principle (assuming faster than light travel is in fact infeasible --
which seems to be the case). To beat them, you'd have to launch at 99.99%
C and aim at targets significantly beyond 1,000 light years away.

Obvious questions:

how small can epsilon be? Is it really limited only by the energy
available to accelerate the seeds?

At some sufficiently high speed, the "vacuum" of space starts to look more
like pea soup, and will result in thermal heating and other damage to the
probe. How can this be minimized? Simple approach: a very long tungsten
rod ahead of the seed as a shield. Given the density of mass in the
universe (something like one atom per cubic meter if memory serves -- which
it might or might not), how fast can we go using a simple "heat shield"

Can we go faster if we adopt a more complex (self repairing) approach?

Can we divert charged particles by magnetic/electrostatic methods?

Should we try to "dodge" obstacles at relativistic speeds? How?

Should we deploy fewer seeds at higher speeds? The limit of this trend
would be a handful of seeds launched at the highest possible speed at the
farthest possible targets using every available joule of energy in the

After landing, a seed will have to spend some time converting the locally
available mass/energy into launch systems and more seeds. How fast can
this be done? Doubling rates on the order of hours should be feasible, if
not faster. But acceleration and deceleration will take time. How much time?

Having launched the next wave, a seed will have to defend the territory it
has acquired. This presumably will require some form of machine
intelligence (upload, AI, some combination). How effective a defense is
feasible? Will we reach a limit where the available mass/energy is
deployed in a not-too-far-from-optimal fashion, and cannot be seized by an
opponent with significantly less mass/energy at his disposal? Or will ever
smarter machines develop ever more devious strategies for taking what is
already held by not-so-smart machines?

At present, the percentage of the population that takes this kind of
discussion seriously is vanishingly small. Does this imply that some
relatively small organization might seize the entire universe by the simple
expedient of launching seeds at stars that nobody cares about because they
are "too far away to matter" (e.g., 1,000 light years)? If people realize
what's happening, presumably only larger organizations will have a
reasonable shot at initiating this process. Should this be encouraged?

This "space rush" looks like a winner-take-all (and I *do* mean *all*)

Received on Sun Dec 14 00:22:00 1997

This archive was generated by hypermail 2.1.8 : Tue Mar 07 2006 - 14:45:29 PST