From: Stirling Westrup (sti@cam.org)
Date: Sun Mar 19 2000 - 13:49:57 MST
Robert J. Bradbury wrote:
A bunch of fascinating questions, which I've had the temerity to try to
answer to the best of my abilities.
> 1) Are the sources of the gravitational microlensing observations
> universally distributed around our galaxy and are they engineered
> objects? [a subset of What is the Missing Mass?].
A partial answer can be found at http://www.cfht.hawaii.edu/News/Lensing/
which has released a map of the dark matter in a small patch of sky, as
deduced from carefully observed microlensing events. It looks awfully
fractal. Something as small as galaxy only shows up as a blob though, so
its hard to say what the local distribution looks like. If something
engineered this result, then they are waaaay scarey indeed.
> 2) What are the speed-limits to self-replication?
>
> 3) What are the speed-limits to evolution?
Are these questions about the absolute limits under any circumstance (at
which point I would say it looks to be around 1E-30 to 1E-40 seconds for
both questions), or is it a question about what parameters are critical to
setting a particular speed limit under particular circumstances. This is a
far more complex, and IMHO a far more interesting question.
> 4) How many really useful "biochemistries" (i.e. chemical systems
> that can support self-replicating self-assembling machines)
> are there and which of those can arise without conscious
> intervention? [A variant of "How big is the phase-space for life?"]
If you are limiting yourself to chemistry then the answer depends on how
big a difference is necessary to consider two biochemistries 'distinct'. I
saw a well reasoned estimate once (which I only vaguely recall, so take
the following with a grain of salt) that argued that carbon chemistry has
many structures used by life that perform functions that are hard to
perform in silicon or any other base for 'organic' forms. So at this level
the answer is 1. If you then look at the basic building blocks of life,
the amino acids, you find that there are hundreds, and the set earth life
uses is somewhere around 20. That gives a rough estimate of 5E20
biochemistries. So much for being able to eat alien food plants. At an
even higher level, the number of ways of encoding information for self-
replication purposes seems unbounded. So much for the 'everyone has DNA'
assumption of bad science fiction.
> 5) Are there feasible ways of creating alterverses, particularly
> those where you have rigged the fundamental physical laws and
> constants? [Interesting because I'm pretty sure we are doomed
> if all the baryonic matter decays and universe-tunneling is
> disallowed.]
The jury is still out on this one. All of the papers I've ever seen on the
matter (and there haven't been many) seem to require unwaranted
assumptions about the structure of space-time and/or the existance of
exotic (and purely theoretical) forms of matter.
> 6) What is the probability that an amoral, self-evolving, self-replicating
> AI/Alife can develop and "breakout" by accident? { After all, the
> computers and the net *are* an environmental niche... }
About the same probability that a chemist mixing amino acids in a lab will
accidentally create a plague-form capable of wiping out our species.
Note that this is a vastly lower probability than the chance that a
deliberately constructed, amoral, self-evolving, self-replicating AI/Alife
will "breakout" by accident.
> 8) What is the molecular size to which a human body must be reduced
> that molecular nanotechnology cannot reassemble you? [At some
> point you end up with too many 3D pieces that look identical, with
> too little positional information, to reassemble Humpty Dumpty.]
Oog. Good question. The answer is clearly related to the question of what
is the actual non-redundant information content of a human body. Clearly
humans are massively redundant with structures copied at all scales in the
trillions. I know I've seen estimates of this latter number somewhere, but
I'm drawing a blank.
> 7) What do SIs think about over trillions of years?
>
> 9) What are the optimal computing architecures for what SIs think about?
I think that both of these questions are fundamentally unknowable at our
level of technology. We can't even answer the (much simpler) questions of
what do humans think about over decades, and what is the optimal computing
architecture for this.
> And number 10, on RBs Top-Ten-List is:
>
> 10) Am I going to make it?
> [Because if Robin is right about the spoils going to the first, *or*
> Eliezer is successful in creating a fundamentally different AI, *or*
> the luddites rise up and drive us back a century or more, *or* *or*
> *or*... my feeling is that the odds aren't good. But(!), arriving is
> not so important as having an enjoyable journey.]
>
Yeah. Choosing a strategy for survival is tough. All I can think of to do
is to study as many relevant fields of knowledge as I can, so that I will
at least be well informed of my options at each point along the way.
--
Stirling Westrup | Use of the Internet by this poster
sti@cam.org | is not to be construed as a tacit
| endorsement of Western Technological
| Civilization or its appurtenances.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:27:31 MST