[extropy-chat] Fw: [atlantis_II] Re: fermi's paradox: m/d approach

[Robert J. Bradbury]

Responding to the most recent comments by Dennis forwarded
by Daniel...

> Sent: Monday, January 05, 2004 5:26 PM
> Subject: [atlantis_II] Re: fermi's paradox: m/d approach

> No investment in discovery?  A singular thought pattern
> which doesn't experiment or take risks?  No diversity
> of opinion concerning value.  Again it doesn't sound like
> much of a brain.

:-;  But there *is* potentially an investment in discovery.
MBrains, using 1% of their mass resources, can have
100 billion telescopes the diameter of the moon.  These
in turn can be laid out across distances of AU's to LY
and it may be possible to link them together into one
f***ing big interferometer (the configuration for the
NASA TPF mission and the ESA Darwin mission to view
extraterrestrial planets).

An MBrain can see just about *everything* within the
speed of light constraints.  In contrast if one wants
electromagnetic communications one effectively gets
only 0.5c and the speed for probes is typically cited
as 0.1c (there) and then c after you build something
capable of sending back information (still at a trickle
of what is available).  Go look at the data rates on
the current Mars probes -- highest speed is 128K bits.
Its going to take a week to get the current pictures
already taken back to us.  Do we communicate with
the probes?  Not really -- they are essentially using
their own intelligence due to the 8 minute delays
in signals to/from the probe.

If one wants probes to survive across interstellar
distances one has to build self-adaptation into them.
If one builds in self-adaptation it may be hard to
retain loyalty.  Particularly because any intelligent
entity is going to realize after 1000 years of travel
time to a star 100 l.y. distant, that whatever one
was once loyal to is unlikely to exist.  An MBrain
could have completely rearchitected itself, dispersed
itself, altered its fundamental purpose, etc. during
that time period.  So one has an extremely fine line
between sending out a probe capable of surviving and
sending out a probe capable of ones enemy.

> Worth something?  Whole libraries can be sent on
> a laser signal for next to nothing.  What makes you
> think those >2^50 bits aren't mainly archives of
> Alien Slug Porn at high resolution or something
> equally useless to someone else.  Value is in the
> eye of the beholder.

Yes, the value is in the eye of the beholder and
establishing a common dictionary between civilizations
that have gone off on very different vectors may be
very difficult.  But you are not paying attention
to the speed and costs of transmitting data across
interstellar distances.  The Library of congress
contains ~20x10^13 bits, the Web is probably slightly
more than that now, all current human memory is
perhaps 1.2x10^19 bits.  This is peanuts compared
with 2^50 bits.

Now, we are not talking a fiber cable between two
cities, we are talking a laser signal across interstellar
distances.  So, I'd suggest you sit down and calculate
the beam power required to have even a single photon
hit the receiver given dispersion and dust interference
across interstellar distances.  While I am not an
expert in these areas I would like to suggest the
cost is not "next to nothing" and the time required
to transmit even the small amount of data in the
Library of Congress is millions if not billions
of years unless one gets very clever.  (Advanced
civilizations may be able to be clever -- but there
are still limits and cost tradeoffs).

> I can only go by what little I have read.  No doubt the
> fantastic unrealistic claims get more press than those
> who are serious.  I will have to read more by those who
> are thought to be serious.

I urge this -- Drexler first proposed the gray goo scenario
in Engines of Creation (which is online at www.foresight.org)
and Robert Freitas has quantified the risks and proposed
solutions:
  http://www.foresight.org/NanoRev/Ecophagy.html

Regarding gray goo and ecophagy:
> I would think that all these claims depend on the efficient
> mining of resources from the environment to enable
> self-replication. Something yet to be demonstrated and
> unlikely to be as simple as postulated.

No argument that general resource extraction will be complex
and that self-replication is simple.  But we know that
microorganisms manage self-replication with something like
350-450 genes.  That isn't a huge number given the complexity
of things that humans have built (the Apollo command module
for example).  Similarly the basic needs for gray goo such
as elements (C, N, O, H, Si) are quite available in the
environment (literally in the air, water and soil) and
energy is clearly available from the sun.

> Some of the more
> fantastic claims I've heard clearly ignored the necessity of
> mining diverse resources and the actual energy costs and
> heat losses incurred in doing so.  There are many biological
> systems which are extremely efficient.

I don't dispute this -- the ATP synthase molecular complex
in the mitochondria may have an extremely high energy conversion
efficiency (80-90%?).  Yet the overall conversion of solar
energy into useful energy by plants is on the order of 1-4%.
Solar cells are now pushing 30-40%.  So it is clear that
the energy conversion mechanisms that may be employed by
gray goo may be more efficient than natural green goo.

Regarding the elimination of superintelligences (MBrains, etc.):

> Or simply: enough hydrogen bombs hurled at them fast
> enough for long enough, or swarms of pellets fired at them
> for long enough from all directions, or enough anti-matter
> hurled long enough, or destroying it while it is small by
> any number of means, or setting off nearby stars to create
> lethal neutrino showers, or orienting parts of the blast from
> supernovas, or hitting it again and again with solid objects
> traveling near the speed of light, and so on.

Yes, there are possible paths to attack megaconstructs.  The
question then becomes could one launch any of these types of
attacks undetected?  And could one detect revenge-intelligences
deployed so as to be undetectable (in suspend mode for thousands
of years in cold/dark interstellar space or in tight orbits
around stars so as to be masked by stellar radiation (used
by Linda Nagata in Vast I think) or hidden within what otherwise
seem to be harmless bodies (asteroids, comets, moons, etc.)?

I bring to mind, the infamous saying by Kahn in one of the
Startrek episodes/movies -- "Revenge is a dish best served cold"!

> Exactly my point - WoMD will destroy the big buck brain
> before it gets very big.  Who says berserkers of one kind
> or another aren't already out there causing the
> Fermi Paradox?

The delay in the speed of light.  Using our civilization as
an example we make the transition from "primitive" to the
singularity in something like 2000-4000 years (depending on
how one defines "primitive").  We make the transition from
"non-scientific" to "scientific" in ~300 years.  So that
imply a requirement for advanced civilizations to be watching
us in very close proximity (10s to 100s of l.y.) to be able
to detect our development and send destructor-bots to eliminate
us.  Once the singularity takes off it will be too late.
It

Regarding the detection of dark, cold, distant MBrains:

> Everything is visible to spread spectrum impulse E&M.  You
> can't hide anything of that size if someone cares to look.
> If I send out trillions of spread spectrum impulse probes I
> will find anything I care to look for.  If you are large and
> lumbering once your found your dead - given sufficient
> time.

I'm not sure that I understand this claim fully.  *But*
taking it at face value if you are inside the galaxy and
an MBrain is orbiting 100,000 light years outside the galaxy
by the time you receive any signals back from such probes
you have minimal confidence that the MBrain will be where
you expect it to be.  All one has to do is execute random
course change maneuvers.  To develop an effective attack
on something with a usable mass of 10^26 kg and an energy
availability of 10^26 W most likely requires the massing
of similar or greater masses and energies.  It seems really
difficult that such activities would go undetected.  So
one could disperse the intelligence across an extremely
large volume of space forcing an increased expenditure
of resources to track the subcomponents down and eliminate
them.  It sounds like the only way to be successful is
to start out way ahead in the game.  And how far "ahead"
one can get appears to be limited by the quantity of
intelligence one can gather given limits-of-physics
delays on communication times.

> I would enjoy being informed in those areas you feel I behind in.
See:
  http://www.aeiveos.com/~bradbury/JupiterBrains/index.html
  (a background)
  http://www.aeiveos.com/~bradbury/MatrioshkaBrains/index.html
  (more or less a guide to recent thoughts/papers)

> The other question is: why is my solution to the Fermi Paradox
> less plausible - since it does not depend on new science,
> big brains, or unknown nanotechnology?

If by this you mean that civilizations self-destruct, I don't consider
it less plausible.  We have been and continue to be at risk of
self-destruction and/or continually knocking ourselves back into
pre-technological states.  *But* we are also at risks from natural
hazards (from earthquakes to global warming to gamma ray bursts).
Those may knock us back as well.  Once intelligent civilizations
become aware of such hazards they will presumably develop strategies
to protect themselves from such.  Finally one has to face the
potential of intelligent external hazards.  I deal with this in
two ways.  First, one runs the risk that an intelligent remote
outpost can always turn on you (look at the United States vs.
England).  This kind of makes one think twice before one runs
off colonizing remote outposts.  Second, there is questionable
benefit to the expenditure of resources to develop remote
outposts given the quantity of information and the speed
of evolution within local entities and the fact that even
if the remote outpost discovers/invents something "wonderful"
it may be too expensive to return it to the founder.

I'm not saying that these are hard and fast rules -- but
I am saying that the delays imposed by the speed of light
and the speed of interstellar travel need to be taken into
account when thinking about these problems.

Robert