_Permutation City_ by Egan
john baez
>In article <49g9nc$7...@usenet.rpi.edu>, dam...@lugnut.stu.rpi.edu (Damien
>Neil) wrote:
>> What Durham did was construct an initial state for Elysium, run it for
>> several minutes on computers, and trust the universe to do the rest.
>> I'm a bit unclear on why it was necessary to `initialize' it in this
>> fashion; it should have already have existed independantly.
>Maybe because it was clearly artificial in origin. If it had a
>self-consistent origin, it would already have existed, by finding itself
>in the dust. But in order for something that was created to exist,
>someone has to create it.
No, I think the idea is that every pattern [I would add, of sufficiently
low algorithmic complexity] does "exist" in a sense.
>Besides, Durham clearly _wanted_ to do it. Much the same way that a man
>with only 8 months to live might still want to father a child that he'll
>never see as anything but a sonogram.
Yes, it seemed clear that this is the reason. After all, he killed
himself afterwards; he needed the emotional link to his copied self that
he could only get by seeing it, to have the nerve to kill himself. Of
course, even so he was crazy to kill himself, since:
If one had the appropriate personality, one could simply go ahead
and kill oneself, if one knew that a pattern isomorphic to oneself existed
somewhere. But in practice our personalities have evolved in such a way
we don't do this; we want ourSELVES to exist, not some isomorphic copy
encoded somewhere in the universe to exist. To be content with the
abstract existence of some copy encoded into the universe, completely
inaccessible, is evolutionarily unfit, i.e. "crazy".
Still, there are some subtleties. As Heraclitus noted, our self 10
minutes later is not really the "same" as our self now, it's just
approximately isomorphic, and fairly nearby in spacetime. (Heraclitus
said "you can never step in the same river twice." One reason is that,
water having flowed on, the river is never the same again. The subtler
reason is that YOU are never the same again.) There's no precise line
concerning how good the approximate isomorphism needs to be, and how
close the proximity in spacetime needs to be, for us to feel happy about
this. Science fiction and philosophy explores this theme a lot... Egan
just found a rather spectacular limiting case.
john baez
I liked this book a lot, because my friend Bruce Smith and I used to
discuss the dust hypothesis back when we were crazy college students.
(Now that I'm a crazy college professor I think about still odder
things.) I wrote something about this in my column This Week's Finds in
Mathematical Physics a while back:
1) Permutation City, by Greg Egan, published in Britain
by Millenium (should be available in the U.S. by autumn).
There is a lot of popular interest these days in
the anthropic principle. Roughly, this claims to explain
certain features of the universe by noting that if the universe
didn't have those features, there would be no intelligent
life. So, presumably, the very fact that we are here and
asking certain questions guarantees that the questions will
have certain answers.
Of course, the anthropic principle is controversial. Suppose
one could really show that if the universe didn't have property
X, there would be no intelligent life. Does this really count
as an "explanation" of property X? People like arguing about
this. But this question is much too subtle for a simple-minded
soul such as myself. I'm still stuck on more basic things!
For example, are there any examples where we *can* really show
that if the universe didn't have property X, there would be
no intelligent life? It seems that to answer this, we need
to have some idea about what we're counting as "all possible
universes", and what counts as "intelligent life". So
far we only know ONE example of a universe and ONE example of
intelligent life, so it is difficult to become an expert on these
subjects! It'd be all to easy for us to unthinkingly assume that
all intelligent life is carbon-based, metabolizes using
oxidation, and eats pizza, just because folks around here do.
Our unthinking parochialism is probably all the worse as far as
different universes are concerned! What counts as a possible universe,
anyway? Rather depressingly, we must admit we don't even know
the laws of *this* universe, so we don't really know what it takes
for a universe to be possible, in the strong sense of capable of
actually existing as a universe. We are a little bit better off
if we consider all *logically possible* universes, but not a whole
lot better. Certainly every axiom system counts as a logically
possible set of laws of a universe --- every set of axioms in every
possible formal system. But who is to say that universes must have
laws of this form? We don't even know for sure that *ours* does!
So this whole topic will remain a hopeless quagmire until one takes
a small, carefully limited piece of it and studies that. People
studying artificial life are addressing one of these bite-sized
pieces, and getting some interesting results. I hope everyone
has heard about Thomas Ray's program Tierra, for example: he
created an artificial ecosystem --- one could call it a "possible
universe" --- and found, after seeding it with one self-reproducing
program, a rapid evolution of parasites, etc., following
many of the patterns of ecology here. But so far, perhaps merely
due to time and memory limitations, no intelligence!
*One* of the cool things about "Permutation City" is an imagined
cellular automaton, the "Autoverse", which is complicated enough to
allow life. But something much cooler is the main theme of the book.
Egan call's it the "Dust Theory". It's an absolutely outrageous
theory, but if you think about it carefully, you'll see that it's
rather hard to spot a flaw. It depends on the tricky
puzzles concealed in the issue of "isomorphism".
Being a mathematician, one thing that always puzzled me about
the notions of "intelligent life" and "all possible universes"
was the question of isomorphisms between universes. Certainly
we all agree that, say, the Heisenberg "matrix mechanics"
and Schrodinger "wave mechanics" formulations of quantum mechanics
are isomorphic. In both of them, the space of states is a Hilbert
space, but in one the states are described as sequences of numbers, while in
the other they are described as wavefunctions. At first they
look like quite different theories. But in a while people
realized that there was a unitary operator from Heisenberg's space of
states to Schrodinger's, and that via this correspondence
all of matrix mechanics is equivalent to wave mechanics.
So does Heisenberg's universe count as the same one as
Schrodinger's, or a different one? It seems clear that they're
the same. But say we had two quantum-mechanical systems
whose Hamiltonians have the same eigenvalues (or spectrum);
does that mean they are the "same" system, really? Is that all
there is to a physical system, a list of eigenvalues??? If we are going to go
around talking about "all possible universes", it would probably
pay to think a little about this sort of thing!
Say we had two candidates for "laws of the universe", written down as
axioms in different formal systems. How would we decide if these were
describing different universes, or were simply different ways of talking
about the same universe? Pretty soon it becomes clear that the issue is not
a black-and-white one of "same" versus "different" universes. Instead, laws
of physics, or universes satisfying these laws, can turn out to be isomorphic
or not depending on how much structure you want the
isomorphism to preserve. And even if they are isomorphic, there
may not be a "unique" isomorphism or a "canonical" isomorphism.
(Very roughly speaking, a canonical isomorphism is a "God-given
best one", but one can use some category theory to make this
precise.) If you think about this carefully you'll see that our
universe could be isomorphic to some very different-seeming ones,
or could have some very different-seeming ones `embedded' in
it.
Greg Egan takes this issue and runs with it -- in a very interesting
direction. Everyone interested in cellular automata, artificial life,
virtual reality, or other issues of simulation should read this, as
well as anyone who likes philosophy or just a good story.
-----------------------------------------------------------------------
However, in the above stuff I didn't want to give away any of the ideas
or plot, so I confined myself to very general remarks. Now I can
finally talk about the actual cool stuff. I'm coming into this
discussion late, having been notified of its existence by
b...@areaplg2.corp.mot.com, so please forgive me if I repeat some stuff
people already said.
So:
>And the actual scheme and operating principles of Elysium, the
>Permutation City, are completely beyond me. How does one go about
>creating an infinite state machine on a finite amount of silicon (or
>whatever) semiconductor transistors? Mathematically, what Egan says
>might work, in an infinite space, but the characters don't _have_ an
>infinite space, no matter what they may choose to believe. They have
>finite machines to work with.
Think of it this way. Suppose you start running a program that
simulates a person at the rate of 1 second virtual time per 1 second
real time. Of course the simulated person, or "copy", would feel fine.
But now suppose it was done at 1 second of virtual time per million
years of real time. Of course the copy would feel just as fine... only
his interactions with the outside world would be screwed up. This is
what Egan dramatizes at the beginning of the book.
Now suppose that the copy was simulated for a while, then the computer
crashed, and then the simulation was continued a few days later after
the backup tapes had been used to get back to the same point in the
simulation. Of course the copy would not notice this, as long as he
wasn't informed of it, or wasn't busy interacting with the outside world
when the crash occurred. (If he *was* interacting with the outside
world, he would simply notice a "glitch" where all of a sudden the
outside world jumped forwards in time a few days.)
Now suppose that there was a disaster and all our advanced computer
technology broked down, so that instead of being run on a computer, the
copy could only be run means of a big bunch of people in a room, passing
notes to each other according to some pre-established rules. (Of course
this is ridiculously impractical, because it would need too many people.
But just pretend!) The copy would still feel fine. Even if people lost
some of the notes and then found them again, or went on holidays over
Christmas, the copy would still feel exactly the same as long as the
overall pattern of information processing was the same.
In short, all that a copy needs to feel perfectly happy is that some
version of its characteristic patterns of information processing be
carried out, somehow, somewhere. Or in other words, the subjective
experience of two isomorphic systems must be the same... or more
precisely, isomorphic.
But now Egan pushes this idea to its seemingly paradoxical extreme. He
notes that the universe is a tremendously big and complicated system,
with zillions of atoms whizzing around in complicated patterns, so that
(he argues) there is a isomorphic copy of ANY system embedded in it
somewhere... at least up to a certain level of accuracy. It's sort of
like Borges' Library of Babel: the complete collection of all possible
books (of a certain length or less) contains *any* book in it *somewhere*.
(Well, any book of less than a certain length, anyway.)
So when Elysium gets started up and then shut down, its patterns
of information processing still occur *somewhere* in the universe, at
*some time*. (Possibly all scrambled up in space and time.) So the
subjective experience of its inhabitants continues.
Now, the only actual loophole is that maybe there is not enough stuff
going on in the universe to find lurking within it, encoded, the
patterns characteristic of Elysium.
This loophole is subtle because it is related to the question of what
we really mean by saying some pattern contains an encoded version of
another pattern. Given codes of arbitrary subtlety, it is hard to set
any fixed limit on this. (Try Stanislaw Lem's His Master's Voice for an
exploration of this theme.) However, there is a mathematical subject
called algorithmic complexity theory that addresses this issue, and I
would argue that there is a reasonable sense in which we can sometimes
say "pattern X does not contain within it an encoded version of pattern
Y." So if the universe is finite, it's possible that it simply could
not contain the Elysium pattern encoded within it, once Elysium got
sufficiently big.
However, I don't think Egan ever mentioned this possibility, and I don't
think *this* is his explanation of what happened at the end of the book.
>I am completely boggled by the whole premise. Could someone please
>explain to me why Durham's scheme promised literal immortality, or
>why he thought it could never be shut down? Eventually, at a worst
>case scenario, the sun dies, the power systems on the chips wear out,
>and everything stops.
No, Elysium is not being run on the chips except for the first few
seconds; after that it's being run on patterns hither and thither in the
universe. It doesn't even need to run forwards in time; as long as some
version of its patterns exist in the universe, it is in some sense
"there". Of course, if we can't interact with it, that doesn't do *us*
any good. Usually when we simulate something we want to interact with
it, and in that case the kind of simulation afforded by the "dust
hypothesis" is completely useless. That's why the whole idea of
Permutation City is a big, elegant, beautiful, mathematically
fascinating joke.
Stephen Harris
>1) Permutation City, by Greg Egan, published in Britain
>by Millenium (should be available in the U.S. by autumn).
>this. But this question is much too subtle for a simple-minded
>soul such as myself. I'm still stuck on more basic things!
Certainly. Such as Topos Theory. Thanks for the reminder about
the book and for interesting #68 and #70 especially.
>all intelligent life is carbon-based, metabolizes using
Yes, look at those silicate AI life forms on Space Above
and Beyond whose only redeeming virtue is gambling.
I suppose I will read this because of Dust Set and Cantor.
This had better not turn out to be like NonCommutative
Geometry which you recommended and cost $70 and
takes me two hours to read a single page while having
a mound of reference books like Wizard's First Rule.
Fractally yours,
Stephen
Mary K. Kuhner
>The idea of isomorphic universes first occurred to me when I was six
>years old! The question I asked myself was this: when I look at
>something red, how do I know that other people see it as red? For all
>I know, what I see as red, they see as blue, and what I see as blue,
>they see as yellow. Is there any way to prove that other people see
>the same colors that I do?
>Anyways, I eventually arrived at the conclusion that since there's
>no way to distinguish these isomorphic universes, it's pointless
>to even think about the issue except for amusement purposes.
Went through the same set of arguments, but in junior high, I think (I
recall really annoying a teacher with them--she couldn't grasp what I
was saying, and kept repeating "Red is red.")
Whether the question is resolvable depends on how you define "seeing".
We could ask whether the same general neuron patterns occur in your
brain as in mine when a red object is presented, and for an objective
definition of "seeing" that might settle the issue. I can't come up
with any way to compare the subjective experiences. Even telepathy
seems inadequate--there's a nice story by Hal Clement pointing out that
in order to comprehend someone by telepathy you need to make a mapping
of his thoughts onto yours (otherwise they are just meaningless brain
waves, since he does not store his thoughts exactly as you do due to
differences in both brain structure and experience). This mapping
could, itself, introduce a red/blue inversion.
Someone in my department is researching the genes that produce red and
green sensitive visual pigments, which due to a quirk of their structure
are very prone to produce hybrid red-green genes (this is a major force
behind red/green color blindness). It turns out that a substantial
proportion of "normal" males do not see the same set of colors as the
rest of us, though they have enough red/green discrimination to pass
color blindness screens and generally have no idea that they are
different from anyone else. Their green-sensitive pigment may pick
up much closer to the red than it should, for example. I wish I knew
what that made things look like!
Mary Kuhner mkku...@genetics.washington.edu
--
I do not receive posts from the following systems because they tolerate
abuse of Usenet: interramp.com psi.com scruz.net
If you wish me to see your message anyway please use email.
david petry
>Being a mathematician, one thing that always puzzled me about
>the notions of "intelligent life" and "all possible universes"
>was the question of isomorphisms between universes.
>[...]
> But say we had two quantum-mechanical systems
>whose Hamiltonians have the same eigenvalues (or spectrum);
>does that mean they are the "same" system, really?
Here's a story that some people might find amusing.
The idea of isomorphic universes first occurred to me when I was six
years old! The question I asked myself was this: when I look at
something red, how do I know that other people see it as red? For all
I know, what I see as red, they see as blue, and what I see as blue,
they see as yellow. Is there any way to prove that other people see
the same colors that I do?
Obviously, it's not enough to ask people what color they see. Even
if someone sees blue where I see red, he has probably learned to
associate the word "red" with the color he sees as blue, so he will
answer any question about the colors he sees the same way I would
answer the question, even though he sees different colors!
Like I said, this issue had me quite perplexed when I was six years
old. I couldn't convince myself that the question was completely
meaningless. For one thing, at that age anyways, colors had a
certain "feel" to them. For example, I really liked red, but blue
kind of turned me off. So maybe the people that liked blue but not
red actually saw blue where I saw red and vice versa.
john baez
>The idea of isomorphic universes first occurred to me when I was six
>years old! The question I asked myself was this: when I look at
>something red, how do I know that other people see it as red? For all
>I know, what I see as red, they see as blue, and what I see as blue,
>they see as yellow. Is there any way to prove that other people see
>the same colors that I do?
>Like I said, this issue had me quite perplexed when I was six years
>old. I couldn't convince myself that the question was completely
>meaningless. For one thing, at that age anyways, colors had a
>certain "feel" to them. For example, I really liked red, but blue
>kind of turned me off. So maybe the people that liked blue but not
>red actually saw blue where I saw red and vice versa.
Colors do have a certain feel to them. Part of the reason is that
the realm of color is not completely separate from other realms of life.
Red is "hot" at least in part because lots of hot things are red, and
it connotes "danger" at least in part because blood is red. Colors may
also have a certain feel to them because we are hard-wired in a very
particular way when it comes to seeing color... e.g., it's probably no
accident that we think of primary colors as being purer and simpler
than things like mauve, khaki, or auburn, because our eyes have (at
least) 3 color receptors in the cones. If we think of brown as somber,
it probably related to the fact that there are no bright brown lights.
And so on. It would be fun to understand how the hard-wiring of our nervous
system, when it comes to color, evolved to be functional for survival
purposes. E.g., is there a basic built-in mechanism that connects "red"
with danger? Lots of poisonous insects are brightly colored, so even
dumb predators must make specific, non-arbitrary, connections between
colors and pleasant or unpleasant sensations. (This reminds me of an
interesting article about a certain species of spider that spins colored
webs, but that would be too big of a digression here.)
In other words, it's not as colors themselves are arbitrary signifiers
in the sense of semantics. Red really plays a differently structured
role in the world than green does; there's no precise symmetry between
them.
On the other hand, if everyone in the world decided to call "red"
"green" and vice versa, things wouldn't be too different.
So this stuff can be pretty tricky.
>Anyways, I eventually arrived at the conclusion that since there's
>no way to distinguish these isomorphic universes, it's pointless
>to even think about the issue except for amusement purposes.
Well, it is important to ponder if you are a mathematician, because the
question of isomorphism is not a simple yes-or-no thing. Things are
typically isomorphic, or not, *in some particular way*, and there are
many situations where there is a multiplicity of interestingly different
isomorphisms. In a sense that's what group theory (or groupoid theory)
is all about. Also, whether two things are counted as isomorphic
depends on the context, that is, the "category" in which you regard them
as objects. E.g., a circle and a square are isomorphic to a topologist,
but not to a geometer. Keeping track of all these subtleties and
distinctions is pretty important in mathematics and physics... but I
suppose that counts as "amusement purposes", since it's all very fun.
Bronis Vidugiris
)This loophole is subtle because it is related to the question of what
)we really mean by saying some pattern contains an encoded version of
)another pattern. Given codes of arbitrary subtlety, it is hard to set
)any fixed limit on this. (Try Stanislaw Lem's His Master's Voice for an
)exploration of this theme.) However, there is a mathematical subject
)called algorithmic complexity theory that addresses this issue, and I
)would argue that there is a reasonable sense in which we can sometimes
)say "pattern X does not contain within it an encoded version of pattern
)Y." So if the universe is finite, it's possible that it simply could
)not contain the Elysium pattern encoded within it, once Elysium got
)sufficiently big.
)
)However, I don't think Egan ever mentioned this possibility, and I don't
)think *this* is his explanation of what happened at the end of the book.
The autoverse inhabitants seemed very disturbed by the essentially
infinite nature of Elysium. (Technically, I suppose it's always finite,
I don't think the growth rate was even exponential, just polynomial,
not positive).
I'd say that counts as "mentioning the possibility", I'm not sure it's
sufficient to say that that's what Egan thinks is "the" explanation.
Especially since the fountain inhabitants don't seem to have been
affected! As I mentioned in another post, I find it hard to explain why
Elysium was affected and the fountain inhabitants not affected. We get
several different views of what people think is going on, all with
incomplete information. (The fountain inhabitants seem not to be aware of
the Autoverse concerns of Elysium, the autoverse and Elysium people are
totally unaware of the fountain people at all, etc.
)>I am completely boggled by the whole premise. Could someone please
)>explain to me why Durham's scheme promised literal immortality, or
)>why he thought it could never be shut down? Eventually, at a worst
)>case scenario, the sun dies, the power systems on the chips wear out,
)>and everything stops.
)
)No, Elysium is not being run on the chips except for the first few
)seconds; after that it's being run on patterns hither and thither in the
)universe. It doesn't even need to run forwards in time; as long as some
)version of its patterns exist in the universe, it is in some sense
)"there". Of course, if we can't interact with it, that doesn't do *us*
)any good. Usually when we simulate something we want to interact with
)it, and in that case the kind of simulation afforded by the "dust
)hypothesis" is completely useless. That's why the whole idea of
)Permutation City is a big, elegant, beautiful, mathematically
)fascinating joke.
I've got some basic problems with this viewpoint. Basically, I'm
used to a world where if I drop a brick on my toe, it will *hurt*.
Now with quantum mechanics, one can no doubt come up with certain
evolutions of the wavefunction of the brick and my toe and my brain that
don't result in something that's feeling "hurt". But there is a formalism
that predicts that the probability that this will happen is low. Really
low. And the low probability means that it just doesn't happen.
I don't quite see how the dust idea is compatible with everyday
notions of probability, and of bricks hurting when I drop them on
my toe(s). Perhaps there is, but it seems to me that the
number of incosistent "states" in the dust which are basically
chaotic should be "greater" than the consistent states in which
dropping bricks on one's toes always produces pain. I can't justify
this with any sort of math, it's my gut feel.
The problem with the dust idea is that it explains everything.
It seems too much like saying: "Well, assume 1=0", then prove
<whatever you want>. IMO.
Possibly this is the whole point of Permuatation City, I suppose,
but in that case I'd have to say I preferred the Niven story
about the suicides (think it was called "All the Myriad Ways".
If the point is that there is no point, I'd say that this point
would best be made in a short story :-).
I can't help but think (or perhaps hope) that Egan had a more
consistent idea of what caused the collapse of Elysium (while
not affecting the fountain inhabitants) than the justification
that "anything is possible with dust". Possibly I'm looking
for too much, maybe there never was intended to be an answer
of the form I'm thinking of.
Anyway, if it's there, so far I've missed it.
Bronis Vidugiris
Mary K. Kuhner <mkku...@phylo.genetics.washington.edu> wrote:
)Someone in my department is researching the genes that produce red and
)green sensitive visual pigments, which due to a quirk of their structure
)are very prone to produce hybrid red-green genes (this is a major force
)behind red/green color blindness). It turns out that a substantial
)proportion of "normal" males do not see the same set of colors as the
)rest of us, though they have enough red/green discrimination to pass
)color blindness screens and generally have no idea that they are
)different from anyone else. Their green-sensitive pigment may pick
)up much closer to the red than it should, for example. I wish I knew
)what that made things look like!
Very interesting!
Offhand, I'd say the difference in appearance would show up as
a different map of color-space. I.e. if you present such a person
with a monochromatic "rainbow" color and ask him to match it with
red, green, and blue light, he'd choose different amounts of RGB
than you would to match the color.
Actually I think that the monochromatic colors wouldn't be good
ones to use, you can't really match all fully saturated colors
with RGB, so it would be better to give him non-saturated
colors (mix in some white light) and compare his responses
to yours.
(The rainbow colors really *can't* be captured fully on pigment or
on the monitor screens, the maps of color space are actually
a little convex from what I've seen.)
Naturally such a person might have different ideas about what
colors co-ordinate, as he's working from a different map.
Some entity who was capable of seeing the colors as a complete spectrum
might well be very confused by human ideas of color - a violet
light is spectrally different from a mixture of red and blue (purple)
lights, it's only the limitations of the human eye that cause us
to imagine that they are similar or the same.
Lee Rudolph
>But
>if it's a pageturner you're looking for, Egan's Permutation City beats
>Connes' Noncommutative Geometry hands down.
--ba...@guitar.ucr.edu (john baez)
Lee Rudolph
Stephen Harris
You might like this book by Terry Goodkind.
>Two hours a page... and several hundred pages long... so just THINK how
>many hours of pleasure you get for a mere 70 bucks!! What a deal! But
Now I realize what qualifies you as a professional mathematician. After
thinking it over I think you deserve a reward. Do You have a fund that
I can contribute to? <grin> I have just finished Weyl's book on groups
and quantum which I read to get further into non-com geom. Good!
>if it's a pageturner you're looking for, Egan's Permutation City beats
So you get the right idea, I went out and bought it immediately after
reading your post. Only $5. Already fun. I read Quarantine also.
He must have a pretty good background. About red/green and isomorphism
Someone who selects colors for dyeing fabric needs to have a much
greater range of color than a plumber. So I agree about context. Perhaps
two people could pass a hundred color test. So might be considered
equal with color differentiation. But given a test of a thousand shades
might very seperate the wheat from the chaff. So the area where symmetry
is important depends upon the question. Likewise for tones and conducting
an orchestra a different category than tuning your favorite am radio station.
Its been Surreal,
Stephen
john baez
>I suppose I will read this because of Dust Set and Cantor.
>This had better not turn out to be like NonCommutative
>Geometry which you recommended and cost $70 and
>takes me two hours to read a single page while having
>a mound of reference books like Wizard's First Rule.
Two hours a page... and several hundred pages long... so just THINK how
many hours of pleasure you get for a mere 70 bucks!! What a deal! But
if it's a pageturner you're looking for, Egan's Permutation City beats
Connes' Noncommutative Geometry hands down. (By the way, Egan's "dust
theory" has nothing much to do with the "dust sets" in fractal theory,
so don't read it for *that* reason. There are many characters named
after mathematicians, though.)
john baez
b...@areaplg2.corp.mot.com (Bronis Vidugiris) writes:
>John Baez writes:
>)So if the universe is finite, it's possible that it simply could
>)not contain the Elysium pattern encoded within it, once Elysium got
>)sufficiently big.
>)However, I don't think Egan ever mentioned this possibility, and I don't
>)think *this* is his explanation of what happened at the end of the book.
>The autoverse inhabitants seemed very disturbed by the essentially
>infinite nature of Elysium. (Technically, I suppose it's always finite,
>I don't think the growth rate was even exponential, just polynomial,
>not positive).
>I'd say that counts as "mentioning the possibility", I'm not sure it's
>sufficient to say that that's what Egan thinks is "the" explanation.
By the way, I don't like or especially understand the ending of the
book, either. I could say what I think Egan thinks, but Michael Weiss
has complained that he doesn't want me giving away the ending, so I
won't; in any event, I don't find it as convincing as the rest of the book.
>I don't quite see how the dust idea is compatible with everyday
>notions of probability, and of bricks hurting when I drop them on
>my toe(s). Perhaps there is, but it seems to me that the
>number of incosistent "states" in the dust which are basically
>chaotic should be "greater" than the consistent states in which
>dropping bricks on one's toes always produces pain. I can't justify
>this with any sort of math, it's my gut feel.
Well, as far as I see it, there are a couple of ways around this
problem. (Now I am forgetting about Egan entirely, and drifting off
into physics, or perhaps metaphysics, so I am setting the followups to
sci.physics.)
Temporarily ignore quantum mechanics. Pretend we're in a classical
universe.
Okay. So, we inhabit a particular universe. But imagine
describing the entire history of this universe using binary digits,
following a known pre-specified scheme. Now we have this big list of bits
which contains all the information about the universe. It might be an
infinite list, by the way... that doesn't bother me.
Since this big list of bits encodes all the information about the
universe, and since we know the coding scheme, we are in some sense able
to reconstruct the universe from this list of bits... so in some sense
having this list of bits is equivalent to having the universe.
Of course, somehow in the course of the discussion we have worked our
way out of the universe! So if you want, imagine we are now in heaven,
sitting in God's library, in front of the volume labelled "Universe
Number 3,151,592: Complete Printout". In some sense, this book *is* our
universe. We should probably be more careful and say the information in
the book is isomorphic (via the coding scheme) to our universe.
Now there are a few questions we can wonder about, which have to do with
how we should treat the notion of isomorphism. Does the fact that the
information in the book is isomorphic to our universe imply that there
are conscious living beings "in the book"? Certainly, if we believe
everything about consciousness is physical, and everything physical
about our universe is in the book, then everything about all the
conscious beings in our universe is in the book. But are those patterns
in the book which are isomorphic to our thoughts and feelings,
themselves thoughts and feelings? That's what I mean to say when I ask,
"are there conscious living beings in the book?"
If you say no, you either think there are nonphysical aspects of
consciousness which the book is leaving out (in which case I have
nothing further to say to you, since our models of reality are too
different), or you need to explain why an isomorphic copy of a conscious
being is not conscious! (Or else you need to pull some other trick.)
If you say yes, then consider this. Suppose you pull the book out of
the shelf and, when God isn't looking, you burn it! Clearly this is
going to get you in serious trouble, but my question is this: what
happened to the conscious beings, that were in "in the book"? Did
they burn up? Clearly they didn't feel the heat of the flames and turn
into ashes, since they were just abstract patterns anyway. But are they
"gone" now?
Does the fact that they now only exist "in the past" in heaven have any
effect on their own self-perceived existence?
More generally, in what sense must an abstract pattern be rendered
concrete for it to exist "for itself"? If you are a serious Platonist
of a certain sort, you can argue that *all* lists of bits "exist" in
some mathematical sense, so that anything encoded by them, all
universes, "exist" in some sense. The word "exist" is pretty vague in
this context; this is why Aristotle drew the distinction between the
"potential" and the "actual". All patterns exist, potentially. But
what is it that makes a pattern "actual"?
Presumably it has *something* to do with following the laws of physics
in this universe. This is, after all, the "actual" universe. Or at
least it's one of them! In any event, we don't have especially good
reason to believe that there are lots of radically different ones, where
big things happen all the time for "no reason". So maybe there is
something about the laws of our universe which have a whole lot to do
with why it's "actual". In other words, maybe if we keep trying to
understand the laws of physics better and better, and keep trying to
understand what "actual" means, better and better, we'll eventually see
that the ultimate laws of physics are simply the definition of "actual".
Of course, most people would prefer to be more "pragmatic" about this.
They would say: something is actual if I can trip over it and stub my
toe. In other words, what's actual is what I perceive.
Or did perceive, anyway.
Or will perceive?
Or could perceive if I tried. Or could at least perceive approximately
if we tried. In principle.
Certainly we tend to want to think of the details of a rock on a planet
in another galaxy as being actual, even if we never get around to seeing
it! Of course, this might be a mistake on our part... especially if we
go back and take quantum theory into account, which limits our ability
to measure and record "everything" about the universe, even in
principle. Perhaps quantum theory is crucial when it comes to
understanding these questions in a really deep way. I suspect it is.
Well, there is more to say about this, but I'll quit here.
>I can't help but think (or perhaps hope) that Egan had a more
>consistent idea of what caused the collapse of Elysium (while
>not affecting the fountain inhabitants) than the justification
>that "anything is possible with dust".
Oh, he definitely did. He explained what caused the collapse of Elysium
in the last chapter. But I don't buy it.
Adam Lou Stephanides
>In article <49le5u$e...@nntp4.u.washington.edu>,
>Mary K. Kuhner <mkku...@phylo.genetics.washington.edu> wrote:
>)Someone in my department is researching the genes that produce red and
>)green sensitive visual pigments, which due to a quirk of their structure
>)are very prone to produce hybrid red-green genes (this is a major force
>)behind red/green color blindness). It turns out that a substantial
>)proportion of "normal" males do not see the same set of colors as the
>)rest of us, though they have enough red/green discrimination to pass
>)color blindness screens and generally have no idea that they are
>)different from anyone else. Their green-sensitive pigment may pick
>)up much closer to the red than it should, for example. I wish I knew
>)what that made things look like!
>Very interesting!
>Offhand, I'd say the difference in appearance would show up as
>a different map of color-space. I.e. if you present such a person
>with a monochromatic "rainbow" color and ask him to match it with
>red, green, and blue light, he'd choose different amounts of RGB
>than you would to match the color.
I read an article along these lines in _Nature_ a while back, and
this is precisely the test they used to distinguish the "abnormal"
males (I don't recall if they used lights or pigments).
What really struck me about this article was that the last paragraph
suggested (as a theoretical possibility) that women with one "normal"
gene and one "abnormal" gene for these receptors might be able to
make the color distinctions that both "normal" and "abnormal" males
could, so they would see a much more complex palette of colors than
anyone else. I wonder if any evidence for this has turned up. (You'd
think that something like this would have been noticed long ago if
it existed, but considering how red-green color blindness can go un-
noticed, that may not be the case.)
--Adam
Jim Walters
> If one had the appropriate personality, one could simply go ahead
> and kill oneself, if one knew that a pattern isomorphic to oneself existed
> somewhere. But in practice our personalities have evolved in such a way
> we don't do this; we want ourSELVES to exist, not some isomorphic copy
> encoded somewhere in the universe to exist. To be content with the
> abstract existence of some copy encoded into the universe, completely
> inaccessible, is evolutionarily unfit, i.e. "crazy".
Reminds me of the Star-Trek transporter problem where the assumption
is that transporters work by encoding the configuration of an object
and then assembling a new object from the information. You would
play hell getting me into the device the first time, but I'ld be much
more willing the next time.
Actually, Orson Scott Card wrote a short story on this theme a while
ago.
Jim Walters
try...@halcyon.com
Stephen Harris
>Isn't not being recordable a fundamental feature of the universe?
You did such a good job of posting my position that I won't repeat it
because of snipping and my mail reader. The idea I think is that
the universe will always grow a step ahead of a mirror universe.
Because the mirror universe was created at a different time.
I think this has to do with the partition problem: Given a finite set
of not necessarily pairwise distinct, positive integers, decide whether
there is a subset whose sum is exactly half the total sum.
Meaning I think that two of these subsets could be added to produce the total.
I have just read a paper which claims to prove the Gauss conjecture that
there are infinitely many real quadratic (rather than finite sh) fields of
class #1. I am an amateur bear in mind, perhaps these do not connect.
Also, I think due to the information involved, there needs to be not only
memory storage, but the information of the address of the storage to be
an identical universe. This ties into my posting that a point which is
dimensionaless without history, differs from from a dimensionaless point
which has a history of compactification of volume and aread (blackhole).
This is also why I think not only the result of equation is relevant to
identity, but also the equational form which produced it, as in transformation.
Sincerely,
Stephen
Greg Egan
> BLURBS WE DOUBT WE'LL SEE ON A DUST JACKET ANY TIME SOON
>>But
>>if it's a pageturner you're looking for, Egan's Permutation City beats
>>Connes' Noncommutative Geometry hands down.
> --ba...@guitar.ucr.edu (john baez)
>Lee Rudolph
I plan to get down on my knees and beg my publishers to put this on the
next edition. Wish me luck.
john baez
>>Okay. So, we inhabit a particular universe. But imagine
>>describing the entire history of this universe using binary digits,
>>following a known pre-specified scheme. Now we have this big list of bits
>>which contains all the information about the universe. It might be an
>>infinite list, by the way... that doesn't bother me.
>What do you mean by a history of the universe, anyway? What are you
>recording? Quantum probabilities?
No, in my thought experiment I explicitly stated that I was assuming the
universe was classical, to keep things simple. The discussion would
need to be drastically changed to take quantum theory into account,
because the notion of "the history of the universe" doesn't make sense
in the same way.
I also wasn't worrying about how the recording was done, because I was
more concerned with the question "suppose there WERE an isomorphic copy
of our universe", rather than the practical question of how one would
get ahold of one. Certainly one would have immense troubles fitting
that copy into our universe... much less getting it by "recording what
happened". That's why I introduced a comic-book version of "God", some
big guy with a gray beard who lives outside our universe and has a squad
of angels writing down everything that takes place. I could have simply
said "assume there were a universe containing an isomorphic copy of our
universe, but I figured the denizens of rec.arts.sf.written would
appreciate a more lively formulation.
As one makes the thought experiment more realistic it gets much more
complicated and interesting!
>Isn't not being recordable a fundamental feature of the universe?
Probably, if you mean recording our universe from within our universe.
Jon Leech
>seconds; after that it's being run on patterns hither and thither in the
>universe. It doesn't even need to run forwards in time; as long as some
>version of its patterns exist in the universe, it is in some sense
>"there". Of course, if we can't interact with it, that doesn't do *us*
>any good. Usually when we simulate something we want to interact with
>it, and in that case the kind of simulation afforded by the "dust
>hypothesis" is completely useless. That's why the whole idea of
>Permutation City is a big, elegant, beautiful, mathematically
>fascinating joke.
Kinda reminds me of an _Analog_ article some years back, discussing the
possibility of life living in a Fourier transform space of the Universe.
Jon
__@/
john baez
>classical, that it is recordable. What I'm saying is that those
>assumptions are so huge and fundamental that a thought experiment
>based on them may have no meaning at all.
My puzzle certainly has *meaning*, because it concerns a logically
possible situation. I agree, though, that its assumptions happen to be
false, and any conclusions based on them have limited relevance to our
particular little world. It's like one of those math problems where you
calculate how many poppy seeds it would take to fill the entire solar
system. In reality, of course, there aren't enough poppy seeds to fill
the solar system. But it can still be amusing to calculate the number.
And sometimes its even useful, in ways that are hard to predict...
Archimedes spent some of his time on such games.
john baez
> Kinda reminds me of an _Analog_ article some years back, discussing the
>possibility of life living in a Fourier transform space of the Universe.
This reminds me of the "goblin worlds" discussed by the physicists
Gell-Man and Hartle.
David Goldfarb
)I don't quite see how the dust idea is compatible with everyday
)notions of probability, and of bricks hurting when I drop them on
)my toe(s). Perhaps there is, but it seems to me that the
)number of incosistent "states" in the dust which are basically
)chaotic should be "greater" than the consistent states in which
)dropping bricks on one's toes always produces pain.
Seems to me that all you need to resolve this is the
Weak Anthropic Principle. All possible versions of you exist
in the dust, but the only ones that would post the above paragraph
are those that experience a consistent set of natural laws.
David Goldfarb <*>|"Dust in the wind...
gold...@ocf.berkeley.edu | All we are is dust in the wind.
gold...@UCBOCF.BITNET | Everything is dust in the wind."
gold...@csua.berkeley.edu |
Jonathan Burns
> 3) That the Autoverse begins to affect Permutation City, which
> is supposed to be running the Autoverse on its processors,
> is silly. Either (a) Permutation City doesn't obey any
> rules because it's just a dust-fantasy anyway, or (b) Egan
> places some magical significance on consciousness. If
> consciousness inside a computer can magically affect the
> universe outside the computer, why don't the original
> Copies start displaying such abilities? Egan certainly
> could have written *that* story if that was the point he
> wanted to make.
I think Egan uses Occam's Razor as a literary device.
Knowing from Part 1 the whole story behind Durham's leap of faith,
the reader could have no complaint if Part 2 began: "Durham, there
are an infinite number of monkeys outside, and they all want royalties
for the programming", for example. Sheckley might have done that. In
an infinite universe, that's just one of the explanations in which
a continuation of the Elysium program could be embedded.
But a causal chain with oo monkeys would entail a complex explanation
for their being there. By comparison, Durham isn't asking the
universe for that much (:-), just a substrate for a digital
automaton to run on, and a whopping great random number (with him
in it) to seed it with.
He's managed to persuade himself that the universe is big enough
to contain an explanation for his continuance, no matter how bad
things look for our hero this time. And so, this time he's setting
things up so that the most complex thing the universe need contain
(to be consistent with his experience of the Elysium seed program)
is another substrate for it to run on.
Lo and behold, there he is in, say, substrate +1, Elysium, embedded
in substrate +0, the automaton. I'd need to check my copy (presently
nested two loans down), but my recollection was that Lambert was
running as +2 on Elysium's computers, and that it was a bit of a
shock to find that the Lambertians were actually on +0. There was
a simpler explanation for their existence than for his!
I could read Permutation City as a riff on a speculation by the
physicist John Wheeler, the "twenty questions" interpretation of
quantum mechanics. The idea is that the universe has an infinitude
of states, worlds if you like, which are consistent with all the
information we have on record. When we make a new observation, it
cuts the infinitude down by a factor of 10^[...], but there remain
an infinitude of worlds in which we might be embedded.
Egan extends that idea to *theories*, to total world explanations.
If we have survived to be conscious in this moment, he says, than
we are probably in one of the simpler explanations for all that
has gone before.
I concur with Scott, to the extent that the Anthropic Principle
is a once-only grant of privileged status to consciousness.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Jonathan Burns | Next week this show having new name
bu...@latcs1.lat.oz.au| ! RONOMOTO, ATOMIC GUMSHOE !
Computer Science Dept | and no Buddhist scriptwriters either....
La Trobe University | -Firesign Theatre
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Mister Skin
eg...@perth.DIALix.oz.au writes:
What I want to know is, whose idea was
"Ten Million People On A Chip"?
Robert Israel
|> Bronis Vidugiris <b...@areaplg2.corp.mot.com> wrote:
|> )notions of probability, and of bricks hurting when I drop them on
|> )my toe(s). Perhaps there is, but it seems to me that the
|> )number of incosistent "states" in the dust which are basically
|> )chaotic should be "greater" than the consistent states in which
|> )dropping bricks on one's toes always produces pain.
|> Seems to me that all you need to resolve this is the
|> Weak Anthropic Principle. All possible versions of you exist
|> in the dust, but the only ones that would post the above paragraph
|> are those that experience a consistent set of natural laws.
"The only ones that would post the above paragraph"? That's a very
strong anthropic principle, I think, if you are saying the only allowed
universes are the ones in which that paragraph would be posted.
But even given that the paragraph is posted (and that there is
a Usenet to post it to), I don't see how you can conclude anything
about the natural laws. Why couldn't you have a situation in which
the Earth, complete with Bronis and Usenet and all our memories, suddenly
emerges out of chaos and last just long enough for the paragraph to be
posted and read before being destroyed by chaos again? (OK, if you want,
I'll give you a light-day or two of "normal" space around the earth so we
can't see any influence of the chaos beyond for a while). Of course it
would seem to be very unlikely for such a thing to happen, but if you
compare states in a "normal" universe to states in a big chaotic universe
with a small (temporarily) "normal" region, there will be far fewer of the
former than the latter.
--
Robert Israel isr...@math.ubc.ca
Department of Mathematics (604) 822-3629
University of British Columbia fax 822-6074
Vancouver, BC, Canada V6T 1Y4
Chris Thompson
|> Why couldn't you have a situation in which
|> the Earth, complete with Bronis and Usenet and all our memories, suddenly
|> emerges out of chaos and last just long enough for the paragraph to be
|> posted and read before being destroyed by chaos again?
There is a very good reason for this, which has just occurred to **#%^^&112
asa881763273 ***%$31~~~~:;Lklakslskskj njcnjsd 172366
Puevf Gubzcfba
Rznvy: pr...@pnz.np.hx
David Goldfarb
)In article <4a104a$9...@agate.berkeley.edu>, gold...@ocf.berkeley.edu (David Goldfarb) writes:
)|> Bronis Vidugiris <b...@areaplg2.corp.mot.com> wrote:
)|> )...Perhaps there is, but it seems to me that the
)|> )number of incosistent "states" in the dust which are basically
)|> )chaotic should be "greater" than the consistent states in which
)|> )dropping bricks on one's toes always produces pain.
)|>
)|> Seems to me that all you need to resolve this is the
)|> Weak Anthropic Principle. All possible versions of you exist
)|> in the dust, but the only ones that would post the above paragraph
)|> are those that experience a consistent set of natural laws.
)
)"The only ones that would post the above paragraph"? That's a very
)strong anthropic principle, I think, if you are saying the only allowed
)universes are the ones in which that paragraph would be posted.
No, that's not what I meant to say.
)But even given that the paragraph is posted (and that there is
)a Usenet to post it to), I don't see how you can conclude anything
)about the natural laws. Why couldn't you have a situation in which
[stuff pops in and out chaotically, complete with fictitious memories
of being non-chaotic.]
You are quite right, and I expressed myself unclearly. Let me
take another shot at it. If all possible states exist in the dust,
then there will be some that have consistent histories, and some
that don't. People living in the consistent histories will experience
consistent natural laws, and have no interaction with the chaotic states --
by definition. Nonetheless the chaotic states (as we've postulated) exist.
Mr. Vidugiris, if I'm reading him right, is saying that since our
timeline is consistent, he finds it hard to believe in chaotic ones.
But this is a non sequitur, since the fact that non-chaotic states exist
tells you nothing about whether inconsistent ones do.
I've thought before about similar things in conjunction with the
quantum-mechanical Many Worlds idea. If all possible worlds exist, then
there must be some where magic seems to work, even though strictly speaking
the effect of each "spell" is just staggering coincidence. Then there
are the worlds where magic used to work and then stopped. (And started
again...[and stopped again...])
David Goldfarb <*>|"...with very few exceptions, nothing lasts
gold...@ocf.berkeley.edu | forever; and among those exceptions no thought
gold...@UCBOCF.BITNET | or work of man is numbered." -- Iain M. Banks
gold...@csua.berkeley.edu |
Bronis Vidugiris
David Goldfarb <gold...@ocf.berkeley.edu> wrote:
)Bronis Vidugiris <b...@areaplg2.corp.mot.com> wrote:
))I don't quite see how the dust idea is compatible with everyday
))notions of probability, and of bricks hurting when I drop them on
))my toe(s). Perhaps there is, but it seems to me that the
))number of incosistent "states" in the dust which are basically
))chaotic should be "greater" than the consistent states in which
))dropping bricks on one's toes always produces pain.
) Seems to me that all you need to resolve this is the
)in the dust, but the only ones that would post the above paragraph
)are those that experience a consistent set of natural laws.
Well, actually there's a larger set than that that would post
stuff like this.
The universe doesn't have to obey consistent physical laws, really,
it's just my memory (state) that has to think that it does.
And you forgot about the universes in which "I" was being sarcastic
and forgot the smiley :-), or otherwise felt impelled to post stuff
that wasn't true according to my memory. And the universes in which
<insert_a_name_here> put a gun to my head and FORCED me to post
such stuff. Not to mention universes where nobody really had a gun
at my head forcing me to post such stuff, but I thought they did.
And the universes where the cat ran over the keyboard
when I forgot to log out, posting this stuff under my name.
That's a particularly good explanation - that damn cat just can't spell!
:-)
Stephen Harris
BTW, A Lambert is a real person. He wrote a paper: Stability of
wavelenghts and spatiotemporal intermittency in coupled map lattices.
Stephen
Roberta and Craig Becker
: What I want to know is, whose idea was
: "Ten Million People On A Chip"?
Yeah...wotta stinker! And, in my copy at least, there's a typo
on the first line of the inside front cover teaser copy.
I dunno...HarperPrism reminds me of that joke in _Bimbos Of The
Death Sun_: "...ought to be in charge of national defense, because
they were so good at keeping secrets."
Craig
--
-- Craig Becker bec...@bga.com http://www.bga.com/~beckers Austin, TX USA --
-- 'Literal immortality? Outliving the universe?' 'That's what the word --
-- means. Not: dying after a very long time. Just: not dying, period.' --
Roberta and Craig Becker
: Kinda reminds me of an _Analog_ article some years back, discussing the
: possibility of life living in a Fourier transform space of the Universe.
Yet more Egan: his short story "Wang's Carpets" deals with this.
Robert Israel
|> You are quite right, and I expressed myself unclearly. Let me
|> take another shot at it. If all possible states exist in the dust,
|> then there will be some that have consistent histories, and some
|> that don't. People living in the consistent histories will experience
|> consistent natural laws, and have no interaction with the chaotic states --
|> by definition. Nonetheless the chaotic states (as we've postulated) exist.
|> Mr. Vidugiris, if I'm reading him right, is saying that since our
|> timeline is consistent, he finds it hard to believe in chaotic ones.
|> But this is a non sequitur, since the fact that non-chaotic states exist
|> tells you nothing about whether inconsistent ones do.
|>
|> I've thought before about similar things in conjunction with the
|> quantum-mechanical Many Worlds idea. If all possible worlds exist, then
|> there must be some where magic seems to work, even though strictly speaking
|> the effect of each "spell" is just staggering coincidence. Then there
|> are the worlds where magic used to work and then stopped. (And started
|> again...[and stopped again...])
The "magical worlds" would presumably be so few in comparison to the non-magical
ones that there's no practical problem. On the other hand, with chaotic states
there is a problem, because there could be so many more chaotic than non-chaotic
ones. Even if you use an Anthropic Principle to throw out all states that
don't have the Earth in substantially its present condition at a certain time.
Unless we know a priori that we are in a non-chaotic state, there would always
be a very strong possibility that things could turn chaotic at any moment.
The experimental fact that this doesn't happen would seem to indicate that
there's something wrong, either with the "dust" hypothesis or the idea that
we are in a random state conditioned only by some version of the Anthropic
Principle.
Vincent TMS Archer
> I've thought before about similar things in conjunction with the
> quantum-mechanical Many Worlds idea. If all possible worlds exist, then
> there must be some where magic seems to work, even though strictly speaking
> the effect of each "spell" is just staggering coincidence. Then there
That's one of the given of the Many Worlds.
Things like the 2nd Law of Thermodynamics, for example, fall victim to
the Many Worlds. On a sufficiently small system, there will be periods
of reversed entropy. As the size of the system increase, or the duration
of observations increase, the probability of reversed entropy diminishes,
but never disappears. The Many Worlds say that there are a (infinite)
number of universes in which entropy runs in reverse.
It's just that they're in a set of measure zero, compared to *ALL* the
universes, so they are reputed *non-existent*.
(There seem to be a quantum indeterminacy around the Net these times. Are
we in rec.arts.sf.written, or sci.physics.quantum.philosophy? :-)
--
Vincent ARCHER -=-=- Herve Schauer Consultants -=-=- arc...@hsc.fr.net
Tel: +33 1 46388990 Fax: +33 1 46380505
Greg Egan
>What I want to know is, whose idea was
>"Ten Million People On A Chip"?
Not mine, I can promise you. I'm still trying to guess whether whoever
put this on didn't understand (or hadn't read) the book ... or whether they
just didn't care that it bore no relationship to the actual content.
Avram Grumer
Becker) wrote:
> Yeah...wotta stinker! And, in my copy at least, there's a typo
> on the first line of the inside front cover teaser copy.
And the phrase spelled "bail-out" on the back cover is consistently
spelled "bale-out" in the actual text of the book.
+--------------------+---------------------------------------------+
| Avram Grumer | http://www.users.interport.net/~avram |
+--------------------+---------------------------------------------+
| If music be the food of love, then some of it be the twinkies of |
| dysfunctional relationships. |
+------------------------------------------------------------------+
john baez
>Unless we know a priori that we are in a non-chaotic state, there would always
>be a very strong possibility that things could turn chaotic at any moment.
>The experimental fact that this doesn't happen would seem to indicate that
>there's something wrong, either with the "dust" hypothesis or the idea that
>we are in a random state conditioned only by some version of the Anthropic
>Principle.
I'd try to boil it down thus: either we have to believe we are in a very
special world among all the logically possible worlds, or that for some
reason not all the worlds we think are logically possible are
"physically possible". It might not be too important at this point to
much about which is the case, since in either case the practical
ramifications are the same: we focus our attention on models of the
world which have features resembling what we actually observe. In the
long run, since I hope that the universe is understandable, I favor the
second idea.
In other words, I hope that there are some special things about the laws
of physics in our universe, which are conducive to this universe
actually existing! Making sense of this requires understanding what we
actually mean by "existing". For example, "existence" seems to have
something to do with "time", and also with "being a particular way",
which is related to logic, in that P(x) asserts a particular proposition
P of an entity x. One must go much further before one starts seeing
anything like ordinary physics emerge from such considerations, but I
actually think one can begin to catch glimpses of it in the study of
n-categories.
Here's a fun quote from John Wheeler about this subject, from the book
Gravitation:
"Paper in white the floor of the room, and rule it off in one-foot
squares. Down on one's hands and kneews, write in the first square
a set of equations conceived as able to govern the physics of the
universe. Think more overnight. Next day put a better set of
equations in to square two. Invite ones most respected colleagues to
contribute to other squares. At the end of these labors, one has worked
oneself out into the doorway. Stand up, look back on all those
equations, some perhaps more hopeful than others, raise ones finger
commandingly, and give the order "Fly!" Not one of those equations will
put on wings, take off, or fly. Yet the universe "flies".
Some principle uniquely right and uniquely simple must, when one knows
it, be also so compelling that it is clear the universe is built, and
must be built in such and sucb aa way, and that it could not possibly be
otherwise. But how can one discover that principle?"
I don't think the second sentence is necessarily true. But it's
inspiring. We'll only see if it's true or not, by trying.
Jim Walters
> "Paper in white the floor of the room, and rule it off in one-foot
> squares. Down on one's hands and kneews, write in the first square
> a set of equations conceived as able to govern the physics of the
> universe. Think more overnight. Next day put a better set of
> equations in to square two. Invite ones most respected colleagues to
> contribute to other squares. At the end of these labors, one has worked
> oneself out into the doorway. Stand up, look back on all those
> equations, some perhaps more hopeful than others, raise ones finger
> commandingly, and give the order "Fly!" Not one of those equations will
> put on wings, take off, or fly. Yet the universe "flies".
And this seems similar to the situation of the color-blind neuroscientist described
by Chalmers in the recent Scientific American. In both situations there seems
to be a lack of sufficient "oomph" to justify a fact of experience. In the case
of physics, as John suggested in his post and Wheeler elsewhere in his book, it
might be hoped that a variant of logicism might tighten down the equations
far enough to force a single theoretical framework. However, even if that should
succeed, we'll likely be left with the old chestnut: why something rather than
nothing.
Jim Walters
try...@halcyon.com
john baez
>However, even if that should
>succeed, we'll likely be left with the old chestnut: why something rather than
>nothing.
Why does it need to be an either/or situation: probably there is both
something, and nothing. The something just stands out more. :-)
Heidegger's Introduction to Metaphysics makes for fascinating reading on
this question, by the way.
David Goldfarb
)The "magical worlds" would presumably be so few in comparison to the non-magical
)ones that there's no practical problem. On the other hand, with chaotic states
)there is a problem, because there could be so many more chaotic than non-chaotic
)ones.
This is true but I don't see why the relative quantities of each
is relevant. As long as you have *some* consistent worlds it doesn't
matter what the relative frequency is of inconsistent ones is.
Even if you use an Anthropic Principle to throw out all states that
)don't have the Earth in substantially its present condition at a certain time.
)Unless we know a priori that we are in a non-chaotic state, there would always
)be a very strong possibility that things could turn chaotic at any moment.
Yes, there is. At the risk of sounding rude, so what?
)The experimental fact that this doesn't happen would seem to indicate that
)there's something wrong, either with the "dust" hypothesis or the idea that
)we are in a random state conditioned only by some version of the Anthropic
)Principle.
Well, the idea is that it *does* happen; it's just not us that sees
it. This idea is quite bulletproof; on the other hand it doesn't make any
practical difference either.
David Goldfarb <*>|"Poor dominoes. Your pretty empire took so long
gold...@ocf.berkeley.edu | to build. Now, with a snap of history's fingers...
gold...@UCBOCF.BITNET | down it goes."
gold...@csua.berkeley.edu | -- Alan Moore, _V for Vendetta_
Stephen Harris
>Even if you use an Anthropic Principle to throw out all states that
>)don't have the Earth in substantially its present condition at a certain time.
I came across an interesting comment pertaining to this is a recreational
math book by Coxeter and Ball. Coxeter said that a distinguished
professor had mentioned that he could imagine the universe in a very
different manner as perceived or inhabited by extraterrestrials.
However, he said, one thing we would have to share are the transcendentals
e and pi. Its hard to know whether the universe has immutable laws.
Stephen
Lee Rudolph
>In article <30CB81...@halcyon.com> Jim Walters <try...@halcyon.com> writes:
>>However, even if that should
>>succeed, we'll likely be left with the old chestnut: why something rather than
>>nothing.
>
>Why does it need to be an either/or situation: probably there is both
>something, and nothing. The something just stands out more. :-)
But now we've got a doubly-binary situation: some something is,
some something isn't, some nothing is, some nothing isn't.
Does the something that isn't stand out more than the nothing
that is?
Reminds me of the other old chestnut, the one that ends,
"I don't know, I've never nothed."
Lee Rudolph
Avram Grumer
Archer) wrote:
> (There seem to be a quantum indeterminacy around the Net these times. Are
> we in rec.arts.sf.written, or sci.physics.quantum.philosophy? :-)
The posts are in an indeterminate condition, potentially existing in both
news groups, until you collapse the probablility bubble by reading them in
one particular place.
Benjamin J. Tilly
gold...@ocf.berkeley.edu (David Goldfarb) writes:
> Robert Israel <isr...@math.ubc.ca> wrote:
> )The "magical worlds" would presumably be so few in comparison to the non-magical
> )ones that there's no practical problem. On the other hand, with chaotic states
> )there is a problem, because there could be so many more chaotic than non-chaotic
> )ones.
>
> This is true but I don't see why the relative quantities of each
> is relevant. As long as you have *some* consistent worlds it doesn't
> matter what the relative frequency is of inconsistent ones is.
>
It is relevant for the following reasons.
1) We are assured that the chance that our world will become one is
very, very, very small.
2) Any world that is one is liable to stop being one very, very, very
quickly.
3) A world which qualifies as one is liable to have very, very, very
low levels of fulfilled spells.
Think of it this way. There is a measureable probability that a
cobblestone will jump in the air about 20' and fall on your head,
killing you. However if you died this way, nobody would probably guess
it. Why?
> Even if you use an Anthropic Principle to throw out all states that
> )don't have the Earth in substantially its present condition at a certain time.
> )Unless we know a priori that we are in a non-chaotic state, there would always
> )be a very strong possibility that things could turn chaotic at any moment.
>
> Yes, there is. At the risk of sounding rude, so what?
>
Look at 2) above.
> )The experimental fact that this doesn't happen would seem to indicate that
> )there's something wrong, either with the "dust" hypothesis or the idea that
> )we are in a random state conditioned only by some version of the Anthropic
> )Principle.
>
> Well, the idea is that it *does* happen; it's just not us that sees
> it. This idea is quite bulletproof; on the other hand it doesn't make any
> practical difference either.
Exactly true.
Ben Tilly