From: DarkVegeta26@aol.com
Date: Sat Dec 08 2001 - 23:01:21 MST
I have to agree with this scenario...it is true that limits we put on
ourselves seem to fade away with time...sociological, cultural and ethnic
limits, technological limits...limits on productivity, on happiness, on
intelligence, limits like death. Consistently we have seen this limits
crushed by the unstoppable rush of the future and hyperexponential growth.
Like Eliezer, I suggest that the only limits we have our the ones in
our minds (as usual), and the "ultimate" or "absolute" "limits on reality"
are not very absolute at all. <A HREF="http://www.frc.ri.cmu.edu/~hpm/project.archive/general.articles/1998/SimConEx.98.html">Perhaps even it is our brains (or bodies)
which creates them in the first place.</A> It would be inconsistent for the
asymptote-approaching hyperexponential curve of progress to suddenly screech
to a halt at any limit less than perfection...the speed of light is
relatively slow to the potentially infinite "velocity" within a virtual
reality. I find the idea of an infinite-dimensional, law-free, transcendent
world jailed within a 3+1D substrate hard to reconcile with the
Singularitarian vision of realizing more human-viewpoint-tangible
achievements like overcoming poverty or suffering.
Throughout the history of the universe, information processing has
been accelerating and compactifying itself, becoming more and more efficent,
using less and less resources to accomplish the same developments (generally
to create its successor). John Smart is a proponent of this view. The
currently leading Singularity/Sysop view seems to be that a Singularity will
occur on the Earth or Solar Sytem scale, with a "shockwave" (perhaps of Von
Nuemann probes?) traveling outward at the speed of light with the intention
of colonizing the entire universe. At this rate, it would take billions of
years to colonize the entire universe after the initial Singularity, a
gigantic length of time...especially for transcendently accelerated minds. I
find this unlikely. A more likely scenario seems to be for a Transcendent AI
to develop and master ontotechnology a very short quantity of time after
mastering the "lower" technologies such as nanotech or conversion of matter
to energy. Think what would happen if you could put time machines and
wormholes everywhere in the universe. You transcend time and space. The
entire universe could be swallowed in an instant. The True State of Reality
would be realized.
The trend of concrescence is another trend which is rapidly
accelerating nowadays...cultures and races merge...along with their
respective beliefs and ideologies...schools of thought engage in a dynamic
balance until a unification or conclusive outcome is acheived and progress is
made. Einstein discovered that space and time were the same thing, as were
matter and energy. Imagine a civilization occupied by consituents with
technology so advanced that change between matter and energy or time and
space is as simple as "changing" the shape of a fluid by pouring it into a
differently shaped container. Their distinction between "time" and "space"
would be obsolete in the same manner that distinctions between different
races are becoming more and more obsolete today. In a virtual reality, mass,
acceleration, velocity, energy, etc will no longer be distinct values. I
hypothesize that a single value, "computational capacity", will dictate the
properties of the "universe", some of its various forms being "dimension",
"mass", "velocity", "space-time", "matter-energy", and so on...including ones
we couldn't even guess of now...
In the spirit of Eliezer's "Comments on Vinge's Singularity"; "the
objections fail to consider that we can cheat." Everything falls in the face
of intelligence, including what we now consider the "Ultimate Unbreakable
Limits". Rather than Sysop Scenarios or Unix Scenarios...perhaps we should
be considering "Ontotechnological Infinity Scenarios", or the like...I
apologize for the length of this message...
Michael Anissimov
San Francisco
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT