From: Keith Henson (hkhenson@rogers.com)
Date: Sun May 21 2006 - 08:58:34 MDT
At 01:31 AM 5/21/2006 -0700, you wrote:
>As I am sure most of you are aware, Nick Bostrom created the Simulation
>Argument awhile back.
>
>This seems to be an SL4 concept, and I did not see hardly any discussion
>of it in the sl4.org archives.
>
>I wrote a critique of the Simulation Argument. I posted it on
>BetterHumans, but I did not get a big response. I think that the it was
>too high-level for that site, which tends to be more mainstream. Since I
>can't expect to reach Bostrom himself with my criticisms, I think this
>list would have people equally capable of responding to them.
>
>here it is, somewhat modified for this audience:
>
>******************************************
>The Simulation Argument is a rather intriguing argument made by Nick
>Bostrom. It states that:
>"At least one of the following propositions is true:
>(1) the human species is very likely to go extinct before reaching a
>“posthuman” stage;
>(2) any posthuman civilization is extremely unlikely to run a significant
>number of simulations of their evolutionary history (or variations thereof);
>(3) we are almost certainly living in a computer simulation.
>It follows that the belief that there is a significant chance that we will
>one day become posthumans who run ancestor-simulations is false, unless we
>are currently living in a simulation."
>(<http://www.simulation-argument.com/>http://www.simulation-argument.com/)
>--------------------------------------------------
You might be amused by a bit of history. The modern concept of simulations
goes back at least as far as _Simulacron-3_, 1964 by Daniel F. Galouye
which I read when it first came out in paperback. (It was required reading
for all those involved in _The Matrix_.)
"Probably influenced directly by Philip K. Dick's Truman Show-esque novel
Time out of Joint, Simulacron-3 can be rightly regarded as the first
description of virtual reality, even if the topic was already treated more
than two thousand years ago in Plato's allegory of the cave."
http://en.wikipedia.org/wiki/Simulacron-3
>I had heard this argument awhile back and, at first, thought there was a
>good deal of sense in it.
Stated in the above form it dates back to a specific time and place, the
1986 Artificial Life Conference in Los Alamos. Richard Dawkins was there
and it was the first time I met him. Hans Moravec was there and though it
was probably only the second time I met him in person, he has long been one
of my favorite people. At the time he had a draft of _Mind Children_,
thick as a city phone directory with him. We were outside the main lecture
room and Hans was rapping about the ever falling price of computation.
In one of those flashes that form long term memory, I interrupted him and
on the spur of the moment said, "Hans, do you realize how unlikely it is
that this is the first time we had this conversation?" Hans is one of
those truly bright people in the world, I am sure the blank expression on
his face was not a common sight! I went on to explain that at some point
in the future computation would be cheap enough to simulate history and all
the people in it, and not once, but over and over like reenacting civil war
battles or the Society for Creative Anachronism.
Therefore the chances were next to zero this was the first time we had this
conversation.
This conversation at that conference was discussed between Hans and me back
in the early 90s on the private Extropian mailing list. Hans expanded the
concept into a humorous article "Pigs In Cyberspace"
http://www.primitivism.com/pigs.htm
>Now I am looking at it from a more skeptical viewpoint. It seems like the
>basic premise is a common logical fallacy: a false dichotomy, or
>trichotomy in this case.
Considering it started out as a *joke* . . . .
>First and foremost, it seems to focus too much on humans. It is very
>likely that there are countless intelligent species, assuming the Drake
>equation is accurate.
Which it almost certainly is not. When Drexler figured out nanotechnology,
one of the first things he did was to look through a huge catalog of
unusual galaxies. He was looking for one where the stars were dimmed
behind a front of expanding technophiles, a galaxy that looked like a bite
had been taken out of it. He didn't find any. There is no evidence I am
aware of that any part inside our light cone is other than "wild
state." This is based on the assumption that matter and energy would
valuable to life forms and would not be wasted.
>It is also likely that among those species, some would be very strange.
Limited by physical reality and evolution of course.
>So strange in-fact that one cannot make statements about the likelihood of
>what posthuman/post-ET civilizations would be like. Statement #2 reaks of
>a Modernist perspective that there is just one course that progress
>takes. Why does he lump all advanced civilizations into one group?
>The nature of the singularity makes the entire premises unknowable or
>irrelevant. It is possible that some of these post-singularity advanced
>civilizations progress so quickly that they breeze right through any times
>at which it might have been useful to create simulations... and continue
>to progress into some unknowable Omega Point. And others might take a
>completely different path that requires many simulations.
It is not "useful" in general to create historical simulations. People
reenact civil war battle for other reasons. I would find it very
distasteful to rerun the 20th century, death camps and all, but I can't say
that's a universal feeling.
>So we cant say anything about the likelihood of advanced civilizations in
>relation to creation of simulated worlds. Maybe only one civilization
>created them, maybe most, maybe none, it is impossible to know with even a
>shred of accuracy.
>It is also impossible to know if you take into consideration the lack of
>knowledge we have. It is quite possible that there are many different
>universes/dimensions/whatever that just make this whole notion of the
>Simulation Argument absurd. It is entirely possible that there is an
>infinite amount of nested universes/dimensions/simulations that there will
>always be one more realm that is just beyond our grasp and comprehension.
>In these other realms or far-away galaxies, who knows how things work? The
>most basic principles of time, cause and effect, logic, etc might be
>radically different. We have been learning that even some of the supposed
>"universal constants" actually change over time.
>This Simulation Argument borders on the metaphysical,
I wonder how many other metaphysical arguments started as a joke?
>and I side whole-heartedly with Wittgenstein in believing that no position
>at all is the best position on metaphysical matters. Sure, we can
>speculate and theorize but we should really do it whilst keeping in mind
>our extremely limited knowledge.
>Additionally, the conclusion drawn from the first three can be proven
>false using basic symbolic logic.
>
>"It follows that the belief that there is a significant chance that we
>will one day become posthumans who run ancestor-simulations is false,
>unless we are currently living in a simulation."
>
>It said at the beginning that "At least one of the following propositions
>are true." So that means that it is not a necessary conclusion that if one
>of the three is true, then the other two are false. All can be true, two
>can be true, or only one can be true according to that criteria.
>Therefore, according to the way his argument was phrased, numbers 1 and 3
>can both be true.
>Let's break the conclusion up into symbols:
>
>x = the belief that there is a significant chance that we will one day
>become posthumans (who run ancestor-simulations) is false
>
>y = we are currently living in a simulation.
>
>
>Here are statements 1 and 2 broken into symbols:
>(1) the human species is very likely to go extinct before reaching a
>“posthuman” stage; = x
>(3) we are almost certainly living in a computer simulation. = y
>
>
>Both x and y can be true simultaneously, as shown above, yet his
>conclusion states "x unless y."
>Thus, his conclusion is false.
No because the assumption is that it will require posthuman technology to
run a simulation. Also you left out that posthumans may just not run
historical simulations.
Then again, perhaps we are living in a simulation. You can't assume that
and relax, of course, because even if the vast majority of the worlds are
simulations, *one,* the first of them has to be rooted in physical
reality. For a positive outcome we have to assume we are not living in a
simulation.
There is also the argument that the entire universe is a simulation and we
see the simulation granularity as quantum mechanics.
Hope this helps,
Keith Henson
PS. About the same time I thought up the suicide lottery. Shared it with
only a few people for fear someone unstable would buy a lottery ticket and
a gun. Others probably invented it before and after I did. It is inherent
in "Many Worlds."
PPS. People living in simulations they could enter and leave was the
subject of the fiction I posted here a few weeks ago. I got a few off list
comments, but the lack of any on list responses to the implied rules for
the interaction of AIs and humans surprised me. Perhaps 7500 words exceeds
what people will read on a screen.
"Sex in the simulation had no biological consequences; producing food *out*
of the simulation and producing babies *in* the simulated world being the
two built in limits Suskulan had no desire to break."
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT