Fw: Quantum accident survivor

7 views
Skip to first unread message

Eric Cavalcanti

unread,
Nov 7, 2003, 1:48:40 PM11/7/03
to everyth...@eskimo.com
Hi,

----- Original Message -----
From: "David Kwinter" <da...@kwinter.ca>


> > I mean the absolutely exact same David Kwinter or Eric Cavalcanti as
> was the moment before.

I agree that a moment from now there will be a number of exactly
equal copies. Nevertheless, I am sure I will only experience being
one of them, so this is what I mean by ' me ' - the actual experiences
I will have. Maybe some copy of me will win the lottery every time
I play, but that does not give me reason to spend my money on it. I
still believe that the probability that 'I' win is 1/10^6, even if on a
multiverse sense, the probability that at least one copy of me wins is 1.
The same should be the case with death if we assume a materialistic
position.


> > What do you mean by *entirely equal*?
> >
> > ----- Original Message -----
> > From: "David Kwinter" <da...@kwinter.ca>
> > To: >
> > Sent: Thursday, November 06, 2003 5:19 AM
> > Subject: Re: Quantum accident survivor
> >
> >
> >> On Tuesday, November 4, 2003, at 10:47 AM, Eric Cavalcanti wrote:
> >>>
> >>> Let me stress this point: *I am, for all practical purposes,
> >>> one and only one specific configuration of atoms in a
> >>> specific universe. I could never say that ' I ' is ALL the
> >>> copies, since I NEVER experience what the other copies
> >>> experience. The other copies are just similar
> >>> configurations of atoms in other universes, which shared
> >>> the same history, prior to a given point in time.*
>>
>>
> >> I would consider these other copies entirely equal to myself IF AND
> >> ONLY IF they are succeeding RSSA observer-moments.
> >>
> >>
> >>
> >> Glossary references : )
> >>
> >> RSSA - The Relative Self-Sampling Assumption, which says that you
> >> should consider your next observer-moment to be randomly sampled
> >> from among all
> >> observer-moments which come immediately after your current
> >> observer-moment
> >> and belong to the same observer.
> >>
> > In a materialistic framework, ' I ' am a bunch of atoms. These atoms
> > happen to constitute a system that has self-referential qualities that
> > we call consciousness. If it happened that these atoms temporarily
> > (like in a coma or anesthesy) or permanently (death) lose this quality,
> > so will ' I '.
>
> I respectfully disagree - parallel universes are equally REAL- you will
> still be you! Quantum branches stem from the same exact atoms in the
> versions of us that die in tons of possible accidents everyday.

I believe that they do in fact exist, and that they do stem from the same
atoms. But they are not 'me', in the sense that I don't see through their
eyes. That's what matters when talking about Immortality. We want to
know if WE are immortal - i.e., if our first-person experience is eternal
- not if SOME copy of us will survive.
What QTI assumes is that ' I ' cannot be one of the dead copies - i.e.,
that the dead copies should be excluded from the sampling pool. But
that is a too strong assumption, which I haven't seen any justification for.
Surely my next observer-moment should be alive or it would not be an
observer. But what makes us believe that 'we' - our first-person
individuality - must necessarily have a next observer-moment in the first
place? That is the assumption that does not seem well-based.

If non-observing states are prohibited, then we should never expect to
be in a coma, or anesthesized, for instance. Whenever you would be
submitted to a surgery, you would see that the doctor somehow failed
to apply the anesthesy correctly, and you would have a *very* conscious
experience.

-Eric.

David Kwinter

unread,
Nov 7, 2003, 2:30:45 PM11/7/03
to Eric Cavalcanti, everyth...@eskimo.com

I still think that's you, especially if you just died and they lived
on.. but now we're just beating a dead horse.

> That's what matters when talking about Immortality. We want to
> know if WE are immortal - i.e., if our first-person experience is
> eternal
> - not if SOME copy of us will survive.
> What QTI assumes is that ' I ' cannot be one of the dead copies - i.e.,
> that the dead copies should be excluded from the sampling pool. But
> that is a too strong assumption, which I haven't seen any
> justification for.
> Surely my next observer-moment should be alive or it would not be an
> observer. But what makes us believe that 'we' - our first-person
> individuality - must necessarily have a next observer-moment in the
> first
> place? That is the assumption that does not seem well-based.
>
> If non-observing states are prohibited, then we should never expect to
> be in a coma, or anesthesized, for instance. Whenever you would be
> submitted to a surgery, you would see that the doctor somehow failed
> to apply the anesthesy correctly, and you would have a *very* conscious
> experience.
>
> -Eric.
>


I think that in the case of anesthesia or any other unconscious state
the true or false outcome of whether we regain consciousness with the
passage of time dictates the sampling pool. The collective fates of the
parallel copies of me under anesthesia aren't stricken from the sample
because we must "necessarily have a next observer-moment" - however
this is a concept which I am uncertain about.

Jesse Mazer

unread,
Nov 7, 2003, 2:36:17 PM11/7/03
to everyth...@eskimo.com
Eric Cavalcanti wrote:

> From: "David Kwinter" <da...@kwinter.ca>
>
>
> > > I mean the absolutely exact same David Kwinter or Eric Cavalcanti as
> > was the moment before.
>
> I agree that a moment from now there will be a number of exactly
>equal copies. Nevertheless, I am sure I will only experience being
>one of them, so this is what I mean by ' me ' - the actual experiences
>I will have. Maybe some copy of me will win the lottery every time
>I play, but that does not give me reason to spend my money on it. I
>still believe that the probability that 'I' win is 1/10^6, even if on a
> multiverse sense, the probability that at least one copy of me wins is 1.
>The same should be the case with death if we assume a materialistic
>position.

But you should no more expect to end up in a branch where you died than in a
branch where you were never born in the first place. Consider, instead of a
branching multiverse, a Star-Trek-style transporter/duplicator in a single
universe, which can deconstruct you and reconstruct exact copies
atom-by-atom in distant locations (assuming the error introduced by the
uncertainty principle is too small to make a difference--if you don't want
to grant that, you could also assume this is all happening within a
deterministic computer simulation and that you are really an A.I.). To use
Bruno Marchal's example, suppose this duplicator recreates two identical
copies of you, one in Washington and one in Moscow. As you step into the
chamber, if you believe continuity of consciousness is "real" in some sense
and that it's meaningful to talk about the probabilities of different
possible next experiences, it would probably make sense to predict from a
first-person-point of view that you have about a 50% chance of finding
yourself in Moscow and a 50% chance of finding yourself in Washington.

On the other hand, suppose only a single reconstruction will be performed in
Washington--then by the same logic, you would probably predict the
probability of finding yourself in Washington is close to 100%, barring a
freak accident. OK, so now go back to the scenario where you're supposed to
be recreated in both Washington and Moscow, except assume that at the last
moment there's a power failure in Moscow and the recreator machine fails to
activate. Surely this is no different from the scenario where you were only
supposed to be recreated in Washington--the fact that they *intended* to
duplicate you in Moscow shouldn't make any difference, all that matters is
that they didn't. But now look at another variation on the scenario, where
the Moscow machine malfunctions and recreates your body missing the head. I
don't think it makes sense to say you have a 50% chance of being "killed" in
this scenario--your brain is where your consciousness comes from, and since
it wasn't duplicated this is really no different from the scenario where the
Moscow machine failed to activate entirely. In fact, any malfunction in the
Moscow machine which leads to a duplicate that permanently lacks
consciousness should be treated the same way as a scenario where I was only
supposed to be recreated in Washington, in terms of the subjective
probabilities. Extending this to the idea of natural duplication due to
different branches of a splitting multiverse, the probability should always
be 100% that my next experience is one of a universe where I have not been
killed.

The big assumption here, as I said earlier, is that there is some sort of
"objective" truth about continuity of consciousness and subjective
probabilities, that it's not just a bunch of isolated observer-moments who
just have an illusion of a consciousness which changes over time due to
memories and expectations. See my thread on "3 possible views of
'consciousness'" here:

http://www.escribe.com/science/theory/m2358.html

>If non-observing states are prohibited, then we should never expect to
>be in a coma, or anesthesized, for instance. Whenever you would be
>submitted to a surgery, you would see that the doctor somehow failed
>to apply the anesthesy correctly, and you would have a *very* conscious
>experience.
>

I don't see any justification for that. Why can't your "next"
observer-moment after the anesthesia begins to take effect be of waking up
hours later? That's a lot what waking up from dreamless unconsciousness
feels like, subjectively.

Jesse

_________________________________________________________________
Crave some Miles Davis or Grateful Dead? Your old favorites are always
playing on MSN Radio Plus. Trial month free!
http://join.msn.com/?page=offers/premiumradio

Hal Finney

unread,
Nov 7, 2003, 3:22:25 PM11/7/03
to everyth...@eskimo.com
Jesse Mazer writes:
> OK, so now go back to the scenario where you're supposed to
> be recreated in both Washington and Moscow, except assume that at the last
> moment there's a power failure in Moscow and the recreator machine fails to
> activate. Surely this is no different from the scenario where you were only
> supposed to be recreated in Washington--the fact that they *intended* to
> duplicate you in Moscow shouldn't make any difference, all that matters is
> that they didn't....

> Extending this to the idea of natural duplication due to
> different branches of a splitting multiverse, the probability should always
> be 100% that my next experience is one of a universe where I have not been
> killed.

I question this analogy. There is an important numerical distinction
between duplication by matter recreation and by quantum splitting. The
former increases your measure, while the latter does not.

In the case of successful duplication, your measure doubles. If the
duplication fails and you end up with only one copy, your measure stays
the same. But if you flip a quantum coin and end up in two branches,
your measure is constant. If you die in one of the branches, your
measure is halved.

Therefore I don't think you can take conclusions from the one case and
apply them to the other. You wouldn't say that failing to double your
money is the same as halving it.

Measure is important. It is what guides our life every day.
We constantly make decisions so as to maximize the measure of good
outcomes, as nearly as we can judge. I don't think we can neglect it
in these thought experiments.

Hal

Jesse Mazer

unread,
Nov 7, 2003, 3:51:33 PM11/7/03
to everyth...@eskimo.com
Hal Finney wrote:

What type of "measure" are you talking about? I had gotten the impression
reading this list that the measure on "everything", however it's
defined--all possible computations, for example--was an open question, and
that different TOEs might disagree. Are you talking about a type of measure
specific to the MWI of quantum mechanics? I thought there was supposed to be
a problem with this due to the "no preferred basis" problem.

In any case, if there is some sort of theory that would give objective
truths about first-person probabilities in splitting experiments (and I'm
not sure if you believe in continuity of consciousness or that such a theory
is out there waiting to be found), then if first-person probabilities
disagree with "measure", however it's defined, I think most people would
care more about maximizing the first-person probabilities of good outcomes
as opposed to measure. The main reason to care about measure would be for
altruistic reasons, that you don't want friends and families to have a high
probability of suffering because they see you die, but even this could be
stated in terms of maximizing the subjective probability of happy outcomes
for other people.

Jesse Mazer

_________________________________________________________________
Is your computer infected with a virus? Find out with a FREE computer virus
scan from McAfee. Take the FreeScan now!
http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

Hal Finney

unread,
Nov 7, 2003, 5:38:58 PM11/7/03
to everyth...@eskimo.com
Jesse Mazer wrote:

> Hal Finney wrote:
> >Measure is important. It is what guides our life every day.
> >We constantly make decisions so as to maximize the measure of good
> >outcomes, as nearly as we can judge. I don't think we can neglect it
> >in these thought experiments.
>
> What type of "measure" are you talking about? I had gotten the impression
> reading this list that the measure on "everything", however it's
> defined--all possible computations, for example--was an open question, and
> that different TOEs might disagree.

That's true, but the important point is to consider why we are searching
for a measure, or why we even think there might be a measure that is
relevant to our experience.

The reason is because our own existence is not chaotic, but ordered.
Presumably there are observers who see universes that are much more
chaotic than ours, universes where there are no natural laws but the
observers just manage to hang together somehow. Why do we see a lawful
universe?

And in our own universe, why do more probable things happen more often
than less probable ones? It's not tautological! Remember our discussion
of the magical universe where dice always come up "6" but everything
else works OK. Why don't we live in one of those universes?

The same thing happens in the MWI. If you send almost-vertically-
polarized photons through a vertical polarizer then 99 times out of 100
they go through. Each time, the universe splits into two branches.
After 100 photons, only one universe in 2^100 of them will see the
right statistics. By sheer numbers of universes, almost all of them
will see about 50% go through. Why aren't we in one of those universes?

The answer to all of these puzzles must be that fundamentally, some
universes are more likely to be experienced than others. This is the
concept which we refer to as measure. It is a weighting factor that
somehow must make some universes more important in the grand scheme
of things.

You are right that there are many different ideas about how measure works
and how it could apply, in both the MWI and in the larger multiverse.
But this uncertainty doesn't mean that we can reject or ignore the concept
of measure. Its reality is forced upon us by every observation we make.

> Are you talking about a type of measure
> specific to the MWI of quantum mechanics? I thought there was supposed to be
> a problem with this due to the "no preferred basis" problem.

The proper manner for incorporating measure into the MWI is indeed an
open question at this point. The simplest is to just introduce it ad
hoc and define the measure of a branch as the square of its amplitude.
Others claim that they can derive this from more elementary and/or
obvious assumptions. But it's got to be there.


> In any case, if there is some sort of theory that would give objective
> truths about first-person probabilities in splitting experiments (and I'm
> not sure if you believe in continuity of consciousness or that such a theory
> is out there waiting to be found),

Well, I do believe in continuity of consciousness, modulo the issues
of measure. That is, I think some continuations would be more likely to
be experienced than others. For example, if you started up 9 computers
each running one copy of me (all running the same program so they stay
in sync), and one computer running a different copy of me, my current
theory is that I would expect to experience the first version with 90%
probability.

However I don't see any way at this point to test this model.

> then if first-person probabilities
> disagree with "measure", however it's defined, I think most people would
> care more about maximizing the first-person probabilities of good outcomes
> as opposed to measure.

Our experiences every day prove that first person probabilities do
correspond to measure, but that is because we define measure to correspond
to what we experience. That is where the amplitude-squared formula for
probability came from in QM; it is there to make theory match experience.


> The main reason to care about measure would be for
> altruistic reasons, that you don't want friends and families to have a high
> probability of suffering because they see you die, but even this could be
> stated in terms of maximizing the subjective probability of happy outcomes
> for other people.

It seems that for QS to be an attractive option, you have to believe
that measure applies all the time, except when you die. What is the
justification for making an exception, when all the rest of the time
you act as if you believe in measure? You would take a good bet rather
than a bad bet, but if your death is involved you'll stop caring?

Hal

Bruno Marchal

unread,
Nov 8, 2003, 9:51:33 AM11/8/03
to everyth...@eskimo.com
At 14:36 07/11/03 -0800, Hal Finney wrote:

snip


>Well, I do believe in continuity of consciousness, modulo the issues
>of measure. That is, I think some continuations would be more likely to
>be experienced than others. For example, if you started up 9 computers
>each running one copy of me (all running the same program so they stay
>in sync), and one computer running a different copy of me, my current
>theory is that I would expect to experience the first version with 90%
>probability.


Almost OK, but perhaps false if you put *the measure* on the (infinite)
computations going through those states. I mean, if the 9 computers
running one copy of you just stop (in some absolute way I ask you to
conceive for
the benefit of the argument), and if the one computer running the
different copy, instead of stopping, is multiplied eventually into many
self-distinguishable copies of you, then putting the measure on the
histories should
make you expect to experience (and memorized) the second version more probably.

It is the idea I like to summarize in the following diagram:

\ / | |
\ / | |
\/ = | |
| | |
| | |

That is, it is like a "future" bifurcation enhances your present measure.
It is why I think comp confirms Deutsch idea that QM branching is really
QM differentiation. What do you think? I mean, do you conceive that the
measure could be put only on the "maximal" possible computations?

Bruno


Bruno

Bruno Marchal

unread,
Nov 8, 2003, 12:40:55 PM11/8/03
to everyth...@eskimo.com
My message 6/11 to Alberto Gómez seems not to have gone through.
I send it again. Apology for those who did receive it.
                                                                                B.

At 09:24 06/11/03 +0100, Alberto Gómez wrote:

For me there is no bigger step between to wonder about how conscience
arises from a universe made by atoms in a Newtonian universe, particles
in a quantum universe, quarks in a quantum relativistic universe and
finally, superstring/n-branes in a 11 dimensional universe for one side
and, on the other side, to wonder about how SAS in a complex enough
mathematical structure can have a sense of conscience.



BM: I agree. It is a genuine point.


[SNIP]
AG:That must be true either in our "physical"
world or the world of a geometrical figure in a n-dimensional spacetime,
or in a computer simulation defined by a complex enough algorithm (These
three alternative ways of describing universes may be isomorphic, being
the first a particular case or not. The computability of our universe
doesn't matter for this question).


BM:I disagree, because if you take the comp. hyp. seriously enough
the physical should emerge as some precise modality from an
inside view of Arithmetical Truth. See UDA ref in Hal Finney's post.



AG:So the mathematical existence, when SAS are possible inside the
mathematical formulation, implies existence (the expression "physical
existence" may be a redundancy)


BM:Same remark. What you say is not only true, but with comp it is
quasi-constructively true so that you can extract the logic and probability
"physical rules" in computer science (even in computer's computer science).
making the comp. hyp. popper-falsifiable.


AG:But, for these mathematical descriptions to exist, it is necessary the
existence of being with a higher dimensionality and intelligence that
formulate these mathematical descriptions? That is: every mathematical
object does exist outside of any conscience? The issue is not to
question that "mathematical existence (with SAS) implies physical
existence", (according with the above arguments it is equivalent). The
question is the mathematical existence itself.


BM:Now, it is a fact, the failure of logicism, that you cannot define integers
without implicitely postulating them. So Arithmetical existence is a
quasi necessary departure reality. It is big and not unifiable by any
axiomatisable theory (by Godel).
(axiomatizable theory = theory such that you can verify algorithmically
the proofs of the theorems)
I refer often to Arithmetical Realism AR; and it constitutes 1/3 of
the computationalist hypothesis, alias the comp. hyp., alias COMP:

COMP = AR + CT + YD (Yes, more acronyms, sorry!)

AR = Arithmetical Realism (cf also the "Hardy post")
CT = Church Thesis
YD = (I propose) the "Yes Doctor", It is the belief that you can be
decomposed into part such that you don't experience anything when
those parts are substituted by functionnaly equivalent digital parts.
It makes possible to give sense saying yes to a surgeon who propose
you some artificial substitution of your body. With COMP you can justify
why this needs an irreductible act of faith (the consistency of
COMP entails the consistency of the negation of COMP, this is akin
to Godel's second incompleteness theorem.
It has nothing to do with the hypothesis that there is a physical universe
which would be either the running or the output of a computer program.
Hal, with COMP the "identity problem" is tackled by the venerable old
computer science/logic approach to self-reference (with the result by Godel,
Lob, Solovay, build on Kleene, Turing, Post etc...).


Bruno

Eric Cavalcanti

unread,
Nov 8, 2003, 12:52:55 PM11/8/03
to everyth...@eskimo.com

I don't quite agree with that argument, even though I was intrigued in the
first
read. The reason is similar to those exposed by Hal finney in his reply to
this
post. These copies are not copies made by the branching of MWI.

In fact, I believe that I will never experience being one of those copies.
Let me
see if I can support that:
Suppose you don't destroy the original, but merely make the copies (and this
also answers the later post from someone with the address
log...@adinet.com.uy). If a copy of me is made *in my own universe*, I
don't
expect to have the experiences of the copies. Suppose I sit on this copy
machine
in New York, and the information of the position and velocities (within
quantum uncertainty) of all particles in my body is copied. Suppose, for the
sake of the
argument, that the mere retrieval of this information should pose no problem
to
me. It should me harmless.
This information then travels by wire from the reader to the reproducer. An
almost
perfect copy of me is made in Paris. Should I, in that moment, expect to
have
the first-person 50% probability of suddenly seeing the eiffel tower? I
don't think
anyone would support that. And in that case, you shouldn't support the
notion that
you could ever be a copy of yourself, since you could always NOT destroy the
original in your example. Whenever you did, the original would have the
first-person experience of dying, i.e., it would never be conscious again.

This example is similar to that of the Schwarzenegger movie where he had a
clone of himself made. Of course the making of the clone has no implication
in the original person's experiences whatsoever. For instance, if the bad
guy in the movie
offered you the opportunity of being cloned and win 1 million dollars, but
killing the original (you), would you accept it? Why or why not? The answer
to this question is the only thing that matters to us when talking about
Immortality.

> The big assumption here, as I said earlier, is that there is some sort of
> "objective" truth about continuity of consciousness and subjective
> probabilities, that it's not just a bunch of isolated observer-moments who
> just have an illusion of a consciousness which changes over time due to
> memories and expectations. See my thread on "3 possible views of
> 'consciousness'" here:
>
> http://www.escribe.com/science/theory/m2358.html
>
> >If non-observing states are prohibited, then we should never expect to
> >be in a coma, or anesthesized, for instance. Whenever you would be
> >submitted to a surgery, you would see that the doctor somehow failed
> >to apply the anesthesy correctly, and you would have a *very* conscious
> >experience.
> >
>
> I don't see any justification for that. Why can't your "next"
> observer-moment after the anesthesia begins to take effect be of waking up
> hours later? That's a lot what waking up from dreamless unconsciousness
> feels like, subjectively.

What I was saying was that the usual QI argument seems to imply that there
is
an external consciousness which cannot be definitely unconscious. But if
unconsciousness was prohibited somehow, the most probable thing was that you
would never be unconscious, since there are always universes where you never
are. But this is a very loose argument, and does not matter to the main
discussion.

-Eric.


Eric Hawthorne

unread,
Nov 8, 2003, 1:27:50 PM11/8/03
to everyth...@eskimo.com, everyth...@eskimo.com
Readers of this list interested in issues of personal identity in the
face of replication
might enjoy the Sci-Fi novel "Kiln People" by David Brin.

In the novel, a technology
has been discovered that allows a person's "soul standing wave" (sic) to
be copied into
a kind of bio-engineered clay substance (molded into a shape like you
and animated
by some kind of enzyme-battery energy store that gives it about a day or
two of "life"
before expiry. ) These "ditto people" come in different qualities (more
expensive to
get a super-smart, super-sensitive version of yourself, cheap to get a
worker-droid
rough copy with fuzzy thinking capabilities and dulled senses.) The
novel, apart from
being a hard-boiled detective yarn in this world, explores issues of
identity,
and how social conventions and rights and responsibilities change with
the presence
of replication of personalities.

Brin's one of the "good writer" sci-fi writers.

Hal Finney

unread,
Nov 9, 2003, 2:22:12 PM11/9/03
to everyth...@eskimo.com
Eric Cavalcanti, <er...@fis.puc-rio.br>, writes:
> Suppose I sit on this copy machine in New York, and the information of the
> position and velocities (within quantum uncertainty) of all particles in
> my body is copied. Suppose, for the sake of the argument, that the mere
> retrieval of this information should pose no problem to me. It should
> me harmless. This information then travels by wire from the reader to
> the reproducer. An almost perfect copy of me is made in Paris. Should
> I, in that moment, expect to have the first-person 50% probability of
> suddenly seeing the eiffel tower? I don't think anyone would support that.

I think your argument is valid, that this experiment is indeed the same
as stepping into a destructive duplication machine and having copies
made in two places.

The only place I think you're wrong is in the last sentence. In fact,
I think many people here would in fact "support that", i.e. they would
expect to face a 50% chance of being in the two places.

I have some subtle issues with this expectation which I will explain at
another time, but broadly speaking I would expect that if a copy were
made of me, and that copy were started up, I would in fact experience
a branching of my experience. If I were about to be copied and I knew
that the copy was going to be started up in Paris, I would expect to
experience the two futures equally.

Others who accept the destructive-double-copy experiment would presumably
agree with this basic analysis.

And for the record, my reservation is that it might be psychologically
different to have two different futures for certain than to have two
futures in two different branches of the multiverse. It seems to me that
this follows from the ASSA, which I provisionally accept at present.
It's hard to say what the perceptual difference will be, but it seems
like there ought to be one.

Hal

David Barrett-Lennard

unread,
Nov 10, 2003, 2:41:41 AM11/10/03
to everyth...@eskimo.com
I'm trying to define "identity"...

Let's write x~y if SAS's x and y (possibly in different universes) have
the same identity. I propose that this relation must be reflexive,
symmetric and transitive. This neatly partitions all SAS's into
equivalence classes, and we have no ambiguity working out whether any
two SAS's across the multi-verse have the same identity.

Consider an SAS x that splits into x1, x2 (in child universes under
MWI). We assume x~x1 and x~x2. By symmetry and transitivity we deduce
x1~x2. So this definition of identity is maintained across independent
child universes.

This is at odds with the following concept of identity...

> I am, for all practical purposes, one
> and only one specific configuration of atoms in a specific
> universe. I could never say that ' I ' is ALL the copies, since I
> NEVER experience what the other copies experience

It seems necessary to distinguish between a definition of identity and
the set of memories within an SAS at a given moment.

Is it possible that over long periods of time, the environment can
affect an SAS to such an extent that SAS's in different universe that
are suppose to have the same identity actually have very little in
common?

What happens if we "splice" two SAS's (including their memories)?

It seems to me that the concept of identity is not fundamental to
physics. It's useful for classification purposes as long as one doesn't
stretch it too far and expose its lack of precision.

This reminds me of the problem of defining the word "species". Although
a useful concept for zoologists it is not well defined. For example
there are cases where (animals in region) A can mate with B, B can mate
with C, but A can't mate with C.

- David


Eric Cavalcanti

unread,
Nov 10, 2003, 6:06:45 AM11/10/03
to everyth...@eskimo.com
Hi,

----- Original Message -----
From: "Hal Finney" <h...@finney.org>

> > Suppose I sit on this copy machine in New York, and the information of
the
> > position and velocities (within quantum uncertainty) of all particles in
> > my body is copied. Suppose, for the sake of the argument, that the mere
> > retrieval of this information should pose no problem to me. It should
> > me harmless. This information then travels by wire from the reader to
> > the reproducer. An almost perfect copy of me is made in Paris. Should
> > I, in that moment, expect to have the first-person 50% probability of
> > suddenly seeing the eiffel tower? I don't think anyone would support
that.
>
> I think your argument is valid, that this experiment is indeed the same
> as stepping into a destructive duplication machine and having copies
> made in two places.

OK...

> The only place I think you're wrong is in the last sentence. In fact,
> I think many people here would in fact "support that", i.e. they would
> expect to face a 50% chance of being in the two places.

I wouldn't know. That seems too weird an expectation. But let me try to
think about that...

> I have some subtle issues with this expectation which I will explain at
> another time, but broadly speaking I would expect that if a copy were
> made of me, and that copy were started up, I would in fact experience
> a branching of my experience. If I were about to be copied and I knew
> that the copy was going to be started up in Paris, I would expect to
> experience the two futures equally.
>
> Others who accept the destructive-double-copy experiment would
> presumably agree with this basic analysis.
>
> And for the record, my reservation is that it might be psychologically
> different to have two different futures for certain than to have two
> futures in two different branches of the multiverse. It seems to me that
> this follows from the ASSA, which I provisionally accept at present.
> It's hard to say what the perceptual difference will be, but it seems
> like there ought to be one.

I think that would certainly be psychologically different, and that may be
an argument against that position (how should I call it? Is there a name?).

You already agreed that the non-destructive-copy experiment is equivalent
to the destructive-double-copy experiment, so let me argue using the first
one.
In the case of non-destructive-copy experiment, the copy is
made in a distinct place/time from the original. They could as well be done
100,000 years in the future and in the Andromeda galaxy, and you should
as well expect to have the subjective experience of being that copy with
the same probability as being the smooth continuation of yourself on Earth.
But in the multiverse, there are certainly infinite other perfect copies of
yourself which are not smooth continuations. We can imagine thousands of
ways how these copies could be made. In computer simulations, in a distant
"Earth" in the Tegmark plenitude, or elsewhere.

But suppose you just stepped outside the Paris duplicator. Unaware of the
experiment that is being made, your last memory is sitting in front of your
computer, reading this email. Suddenly, you see the Eiffel Tower. That
would surely be a psychological experience that we don't have too often.
And since there are infinite copies of yourself at any given moment, if you
should expect to be any of them at the next moment, you shouldn't expect
to ever feel the continuous experience you do.

Therefore, since I do actually have a continuous experience of myself,
then 'I am not my copies'.

-Eric.

Eric Cavalcanti

unread,
Nov 10, 2003, 8:09:34 AM11/10/03
to everyth...@eskimo.com
Hi,

I disagreed with some points in your argumentation...

----- Original Message -----
From: "David Barrett-Lennard" <d...@fractaltechnologies.com>


> I'm trying to define "identity"...
>
> Let's write x~y if SAS's x and y (possibly in different universes) have
> the same identity.

You did not yet 'define' identity. You just proposed a relation between
two entities which is supposed to mean that these two entities have the
same 'identity'.

> I propose that this relation must be reflexive,
> symmetric and transitive.

This is a proposal which seems self-evident, but let us keep in mind that
we have no reason to propose it yet, since we don't even know what
'identity' means. I believe, in fact, that this relation '~' is NOT
transitive!
Let me try to argue why later.

> This neatly partitions all SAS's into
> equivalence classes, and we have no ambiguity working out whether any
> two SAS's across the multi-verse have the same identity.
>
> Consider an SAS x that splits into x1, x2 (in child universes under
> MWI). We assume x~x1 and x~x2. By symmetry and transitivity we deduce
> x1~x2. So this definition of identity is maintained across independent
> child universes.

This is where the '~' relation shows that it cannot be transitive. I don't
know what your definition of identity is, but in other posts I have
argued that I am not the copies of me in other universes. Therefore,
since you have come to the conclusion that I am, it must be the case
that your assumption of the transitivity of '~' is wrong. To support your
definition of '~', you must give a better reason to believe that you are
the copies of yourself in other worlds. Just defining an arbitrary '~'
relation does not do the job.

In fact, I believe we should define another relation of personal identity,
which is NOT symmetric. I shall use the notation '<' meaning
that if x<y, x is a former state of y. 'x' is unambiguously defined by
following "down" the multiverse branching 'tree'. But we cannot define
the '>' relation, i.e., the relation by which y>x would mean that y is a
continuation of x 'uptree'. Since there are multiple choices for the next
state of x, it cannot be told in advance what the next subjective moment
will be.

So you can say that x<x1 and x<x2, but it does not follow that x1<x2 OR
x2<x1. And since, by my definition, personal identity can be determined
only by the relation '<', x1 does not have the same 'identity' of x2.

> This is at odds with the following concept of identity...
>
> > I am, for all practical purposes, one
> > and only one specific configuration of atoms in a specific
> > universe. I could never say that ' I ' is ALL the copies, since I
> > NEVER experience what the other copies experience
>
> It seems necessary to distinguish between a definition of identity and
> the set of memories within an SAS at a given moment.
>
> Is it possible that over long periods of time, the environment can
> affect an SAS to such an extent that SAS's in different universe that
> are suppose to have the same identity actually have very little in
> common?
>
> What happens if we "splice" two SAS's (including their memories)?
>
> It seems to me that the concept of identity is not fundamental to
> physics. It's useful for classification purposes as long as one doesn't
> stretch it too far and expose its lack of precision.

Maybe it is not fundamental to physics, but it surely is fundamental to
us, since that may be the difference between immortality or otherwise.
Even more importantly, it is the basis for all our daily decisions.
It is not merely a classification purpose. When you decide not to spend
your money on the lottery, you don't think that doing so is good,
because you will be increasing the number of 'yous' who are
rich elsewhere. You don't care for the other 'yous' because you truly
believe that the probability that you will just lose yor money is too high.
And if you don't care for the other 'yous' they are not really 'you', they
are other entities.

> This reminds me of the problem of defining the word "species". Although
> a useful concept for zoologists it is not well defined. For example
> there are cases where (animals in region) A can mate with B, B can mate
> with C, but A can't mate with C.

Although you can safely ignore those classifications when relating to
objects,
you cannot deny that defining your identity is too easy. Cut your finger
and you will know who is feeling the pain.

-Eric.

Message has been deleted

Saibal Mitra

unread,
Nov 10, 2003, 10:03:52 AM11/10/03
to David Barrett-Lennard, everyth...@eskimo.com
There are some problems with this as Eric has pointed out.

The best way to define identity, i.m.o., would be to say that a program is a
SAS having an identity. If that SAS experience the outcome of an experiment,
it's program will be changed by the mere fact it has acquired the memory of
the outcome of the experiment. So the identity has changed because the
program has changed. Programs are what some of us call ''observer moments''.


----- Oorspronkelijk bericht -----
Van: "David Barrett-Lennard" <d...@fractaltechnologies.com>
Aan: <everyth...@eskimo.com>
Verzonden: Monday, November 10, 2003 08:39 AM
Onderwerp: RE: Quantum accident survivor

Message has been deleted

Hal Finney

unread,
Nov 10, 2003, 12:23:45 PM11/10/03
to everyth...@eskimo.com
Eric Cavalcanti, <er...@fis.puc-rio.br>, writes:
> In the case of non-destructive-copy experiment, the copy is
> made in a distinct place/time from the original. They could as well be done
> 100,000 years in the future and in the Andromeda galaxy, and you should
> as well expect to have the subjective experience of being that copy with
> the same probability as being the smooth continuation of yourself on Earth.

Yes, that makes sense.

> But in the multiverse, there are certainly infinite other perfect copies of
> yourself which are not smooth continuations. We can imagine thousands of
> ways how these copies could be made. In computer simulations, in a distant
> "Earth" in the Tegmark plenitude, or elsewhere.

Yes, but keep in mind that there are also infinitely other copies which
*are* smooth continuations. And these probably outnumber the ones which
are discontinuous (assuming that terms like "outnumber" can be generalized
to infinite sets, or else that the sets involved are merely large and not
infinite).

> But suppose you just stepped outside the Paris duplicator. Unaware of the
> experiment that is being made, your last memory is sitting in front of your
> computer, reading this email. Suddenly, you see the Eiffel Tower. That
> would surely be a psychological experience that we don't have too often.
> And since there are infinite copies of yourself at any given moment, if you
> should expect to be any of them at the next moment, you shouldn't expect
> to ever feel the continuous experience you do.

Rather, you should expect to feel both, with some probability. And I
think that the multiverse holds a greater proportion of continuations
that are continuous than that are discontinuous. Fundamentally this
is because the conditions that promote consciousness and therefore the
formation of brains like mine will tend to involve continuous chains
of experience. Only in a relatively few universes will I be subject to
unknowing duplications. Therefore I think it is highly unlikely but not
impossible that I will suddenly experience a discontinuity.

Let us suppose, though, that our society evolves to a state where such
duplications are routine. Anyone may have their brain scanned at any
time, without their knowledge, and new copies of them created. Suppose I
am such a copy, in fact, I am a 10th generation copy; that is, 10 times
in my life I have found myself having an experience similar to what you
described, a discontinuity where I was just walking along or sitting
there, and suddenly found myself stepping out of a duplicating machine
because someone copied me.

I think you will agree that my memories are reasonable; that is, that
anyone who has gone through such an experierence as I describe will in
fact remember these discontinuities.

Given my history, wouldn't it be reasonable for me to expect, at any
future moment, to possibly face another such discontinuity? It has
happened many times before, both to me and to other people that I know;
it is an often-discussed phenomenon of the world, in this scenario.
Just like anything else that happens occasionally to everyone, it would
be perfectly reasonable and rational to have an expectation that it
might happen to you.

Hal

David Barrett-Lennard

unread,
Nov 10, 2003, 9:07:17 PM11/10/03
to everyth...@eskimo.com
Hi Eric,

> In fact, I believe we should define another relation of personal
identity,
> which is NOT symmetric.

I agree that this has greater relevance to QTI, but note that saying
that "identity" is not symmetric is at odds with most people's usage of
the word. Eg you can't say "x,y have the same identity".

You proposed the notation '<' meaning that if x<y, x is a former state
of y. For the purposes of QTI, I think a more relevant definition of
'<' would be based on successions of growing memories. This would make
it clear that a clone and an original have equal right to prior states
of the SAS.

- David


Eric Cavalcanti

unread,
Nov 11, 2003, 7:01:09 AM11/11/03
to everyth...@eskimo.com
Hi,

----- Original Message -----
From: "Hal Finney" <h...@finney.org>

> Eric Cavalcanti, <er...@fis.puc-rio.br>, writes:
> > In the case of non-destructive-copy experiment, the copy is
> > made in a distinct place/time from the original. They could as well be
done
> > 100,000 years in the future and in the Andromeda galaxy, and you should
> > as well expect to have the subjective experience of being that copy with
> > the same probability as being the smooth continuation of yourself on
Earth.
>
> Yes, that makes sense.
>
> > But in the multiverse, there are certainly infinite other perfect copies
of
> > yourself which are not smooth continuations. We can imagine thousands of
> > ways how these copies could be made. In computer simulations, in a
distant
> > "Earth" in the Tegmark plenitude, or elsewhere.
>
> Yes, but keep in mind that there are also infinitely other copies which
> *are* smooth continuations. And these probably outnumber the ones which
> are discontinuous (assuming that terms like "outnumber" can be generalized
> to infinite sets, or else that the sets involved are merely large and not
> infinite).

I don't think so. Suppose you have at least one other perfect copy of
yourself,
such that you could expect that your next experience be one of that copies'
with the same probability as the smooth continuation. A moment dt from now
the original 'you' will have branched into a number N of possible future
states.
Since the copy is perfectly equal, the copy will also evolve to a number of
future
states that is of the same order of magnitude of N. According to your view,
each of these states is a continuation of yourself with equal probability,
so
that you should expect to have about 50% probability of being your copy.
But, if the Plenitude deserves the name, then we should expect to have
at least a Huge number of copies at any moment.
Therefore, either there are no other copies - i.e. the plenitude is not
real, and
there are no simulations of yourself anywhere in the multiverse, etc. - or
you
cannot experience being one of your copies, and QTI is not real. One of
these has to go.

> > But suppose you just stepped outside the Paris duplicator. Unaware of
the
> > experiment that is being made, your last memory is sitting in front of
your
> > computer, reading this email. Suddenly, you see the Eiffel Tower. That
> > would surely be a psychological experience that we don't have too often.
> > And since there are infinite copies of yourself at any given moment, if
you
> > should expect to be any of them at the next moment, you shouldn't expect
> > to ever feel the continuous experience you do.
>
> Rather, you should expect to feel both, with some probability. And I
> think that the multiverse holds a greater proportion of continuations
> that are continuous than that are discontinuous. Fundamentally this
> is because the conditions that promote consciousness and therefore the
> formation of brains like mine will tend to involve continuous chains
> of experience. Only in a relatively few universes will I be subject to
> unknowing duplications. Therefore I think it is highly unlikely but not
> impossible that I will suddenly experience a discontinuity.

I have argued above about the proportion of smooth/discontinuous states.

> Let us suppose, though, that our society evolves to a state where such
> duplications are routine. Anyone may have their brain scanned at any
> time, without their knowledge, and new copies of them created. Suppose I
> am such a copy, in fact, I am a 10th generation copy; that is, 10 times
> in my life I have found myself having an experience similar to what you
> described, a discontinuity where I was just walking along or sitting
> there, and suddenly found myself stepping out of a duplicating machine
> because someone copied me.
>
> I think you will agree that my memories are reasonable; that is, that
> anyone who has gone through such an experierence as I describe will in
> fact remember these discontinuities.
>
> Given my history, wouldn't it be reasonable for me to expect, at any
> future moment, to possibly face another such discontinuity? It has
> happened many times before, both to me and to other people that I know;
> it is an often-discussed phenomenon of the world, in this scenario.
> Just like anything else that happens occasionally to everyone, it would
> be perfectly reasonable and rational to have an expectation that it
> might happen to you.

It would be perfectly normal, in such a society, to expect to BE a clone,
if you have some reason to believe you are, such as a long-forgot
discontinuity of experience.
But one should not expect to ever BECOME a clone, for the reasons
I argued above.

-Eric.

Message has been deleted

Jesse Mazer

unread,
Nov 14, 2003, 4:45:56 AM11/14/03
to everyth...@eskimo.com
Hal Finney wrote:
>
>Jesse Mazer writes:
> > In your definition of the ASSA, why do you define it in terms of your
>next
> > observer moment?
>
>The ASSA and the RSSA were historically defined as competing views.
>I am not 100% sure that I have the ASSA right, in that it doesn't seem
>too different from the SSSA. (BTW I have kept the definitions at the end
>of this email.) (BTW, BTW means By The Way.) But I am pretty sure about
>the RSSA being in terms of the "next" moment, so I defined the ASSA the
>same way, to better illustrate its complementary relationship to the RSSA.
>
>The real difference between these views was not addressed in my
>glossary, which is that the RSSA is supposed to justify the QTI, the
>quantum theory of immortality, while the ASSA is supposed to refute it.
>That is, if you only experience universes where your identity continues,
>as the RSSA implies, then it would seem that you will never die. But if
>your life-moments are ruled by statistics based on physical law as the
>ASSA says, then the chance that you will ever experience being extremely
>old is infinitesimal.
>
>Personally I think the ASSA as I have it is somewhat incoherent, speaking
>of a "next" observer moment in a framework where there really isn't any
>such notion. But as I said it has been considered as the alternative
>to the RSSA. I invite suggestions for improved wording.

I think that proponents of the type of ASSA you’re talking about would say
that the experience of consciousness passing through multiple
observer-moments is simply an illusion, and that I am nothing more than my
current observer-moment. Therefore they would not believe in quantum
immortality, and they also would not define the ASSA in terms of the "next"
observer-moment, only the current observer-moment. I think you’d be
hard-pressed to find any supporters of the ASSA who would define it in the
way you have.

But as I say below, I think it is possible to have a different
interpretation of the ASSA in which consciousness-over-time is not an
illusion, and in which it can be compatible with the RSSA, not opposed to
it.

>
> > Wouldn't it be possible to have a version of the SSA where
> > you consider your *current* observer moment to be randomly sampled from
>the
> > set of all observer-moments, but you use something like the RSSA to
>guess
> > what your next observer moment is likely to be like?
>
>That seems contradictory. You have one distribution for the current
>observer-moment (sampled from all of them), and another distribution for
>the next observer-moment (sampled from those that are continuous with
>the same identity). But the current observer-moment is also a "next"
>observer-moment (relative to the previous observer-moment). So you can't
>use the ASSA for current OM's and the RSSA for next OM's, because every
>next is a current, and vice versa. (By OM I mean observer-moment.)

Well, any theory involving splitting/merging consciousness is naturally
going to privilege the current observer-moment, because it’s the only thing
you can be really sure of a la "I think therefore I am"…when talking about
the past or the future, there will be multiple pasts and multiple futures
compatible with your present OM, so you can only talk about a sort of
probabilistic spread.

That said, although some might argue there’s a sort of philosophical
contradiction there, I think it is possible to conceive of a mathematical
theory of consciousness which incorporates both the ASSA and the RSSA
without leading to any formal/mathematical contradictions. There could even
be a sort of "complementarity" between the two aspects of the theory, so
that OM’s with the highest absolute probability-of-being would also be the
ones that have the most other high-absolute-probability OM’s that see them
as a likely "successor" in terms of relative probability-of-becoming. In
fact, an elegant solution for determining a given OM’s absolute
probability-of-being might be to simply do a sum over the probability of
becoming that OM relative to all the other OM’s in the multiverse, weighted
by their own probability-of-being.

Here’s a simple model for how this could work. Say you have some large set
of all the OM’s in the multiverse, possibly finite if there is some upper
limit on the complexity of an OM’s, but probably infinite. You have some
theory of consciousness that quantifies the "similarity" S between any two
given OM’s, which deals with how well they fit as the same mind at different
moments, how many of the same memories they share in common, how similar are
their causal patterns, and so on. You also have some absolute measure on all
the OM’s, a "probability-of-being" B assigned to each one—this is basically
just my idea that the self-sampling assumption could be weighted somehow, so
that the ideal way to use the ASSA is to assume that your current OM is
randomly sampled from the set of all possible observer-moments, weighted by
their own probability-of-being B.

Then, to determine the relative probability-of-becoming various possible
OM’s, I could just multiply their similarity S to my own current OM by their
absolute measure B representing each one’s probability-of-being. This would
insure that even though a version of me observing a dragon popping out of my
computer screen may be have just as much similarity S to my current mental
state, in terms of memories and the like, as a version of me who’s watching
the computer screen behaving normally, if one OM is objectively less
probable (lower B) due to the laws of nature, I will have a higher relative
probability of becoming the OM who sees business-as-usual. This would also
insure that if I step into a teleportation machine and the machine
reconstructs two people, one whose brain is close to identical to mine and
one who has a very different personality and memories, then even if the OM’s
of both these people have about the same absolute probability-of-being B, I
am far more likely to become the one who’s more similar to me because his
similarity S to my current OM would be much higher.

And as I suggested earlier, it would be neat if the probability-of-being B
could itself be derived by something like a sum over the S’s between me and
all the other other OM’s, each one weighted by their own B-rating. This idea
could be summed up by the slogan "the most probable present experiences are
the ones that are high-probability successors to other experiences that are
themselves highly probable present experiences". In this way it might even
be possible to bootstrap a unique B-rating for all OM’s, starting with only
a knowledge of the similarity ratings between them. Consider the following
simple universe with only three observers X, Y, and Z, and a known matrix of
similarity ratings S between each pair:

X Y Z
X1.00 0.60 0.35
Y 0.60 1.00 0.25
Z 0.35 0.25 1.00

In this case, if the B-ratings for each one were determined by a sum over
the S-ratings for the others weighted by their own B-ratings, and you
represent X’s B-rating by the variable x, Y’s B-rating by the variable y,
etc., then you’d have some simultaneous equations that’d actually allow you
to find a unique self-consistent solution for x, y, and z:

x = (0.60)y + (0.35)z
y = (0.60)x + (0.25)z
z = (0.35)x + (0.25)y

I haven’t actually planned these numbers out, so the solution probably leads
to some variables being negative or greater than one, which doesn’t really
make sense if the B’s are supposed to be probabilities, but the basic idea
here is that you can bootstrap the B’s just by knowing the S’s.

Now keep in mind, this is all a very cartoonish sketch, I don’t really think
whatever theory of consciousness is used to determine relative probabilities
would be as simple as multiplying a "similarity rating" by an absolute
probability; among other things, "similarity" fails to capture the crucial
issue of the directionality of subjective time, my current OM might be just
as similar to an OM 2 seconds ago as it is to one 2 second from now, but I
expect a higher probability I’ll become the one 2 seconds in the future.
Also, I suggested earlier that the complexity of an OM’s consciousness might
play a part in both the absolute probability (so my present experience is
more likely to be that of a human than an insect) and relative probability
(so I am more likely to experience becoming a copy with an intact brain than
one with brain damage), but the model I presented doesn’t take that into
account. Still, it’s sort of a pet theory of mine that the real TOE will
turn out to be analogous to this model in the following ways:

1. It will include a theory of consciousness that can take my present OM
along with various possible future OM’s, and determine the relative
probability of my experiencing each one in my future based on a combination
of features that are inherent to each OM (analogous to the ‘similarity’
rating in my model) and an external measure which assigns each one an
absolute probability. The relative probability on different future
observer-moments would be used as weights in the RSSA, and the absolute
probability of different present observer-moments would be used in a
weighted ASSA.

2. Even if you don’t know the correct absolute probability of any of the
OM’s to start with, there will turn out to be a unique self-consistent
solution to what this absolute measure on OM’s has to look like, given only
the theory of consciousness and the assumption that all possible OM’s exist
(the ‘everything’ part of the theory). This would be analogous to the unique
solution to the simultaneous equations in the cartoon model above.

This would be neat because the laws of physics we observe could hopefully be
derived (in principle anyway) from the absolute and relative measures on all
OM’s, so you’d basically be deriving all the laws of the universe from just
a theory of consciousness and platonic assumption that every conscious
pattern that can exist, does exist. The problem with any TOE that
incorporates a "theory of consciousness" is that it runs the risk of being a
dualist theory if any aspect of first-person probabilities derives from
something other than that theory (like an objective measure on universes
rather than OM’s to explain why I don’t experience Harry Potter worlds), but
this idea is nicely monist and simple.

It might seem that a theory centered on consciousness and observer-moments
would suggest that any part of the universe that isn’t observed by a
sentient being doesn’t really exist, but I imagine identifying distinct
"observer-moments" with something like "patterns of causal relationships"
(or finite computations, perhaps), so that all such patterns, even the
random jostling of molecules in a cloud of gas, would qualify as
observer-moments with very low-grade levels of consciousness. That way the
absolute probability of each such pattern, along with the probabilistic
relationships between different patterns, might be used to derive what we
ordinarily think of as the laws of physics, especially if the laws of
physics can ultimately be stated in terms of nothing but relationships
between elementary events, as physicists like Lee Smolin have suggested.
This is similar to the "naturalistic panpsychism" idea I found described on
the same website that hosts the many-worlds FAQ (although I disagree with
them on a few points):

http://www.hedweb.com/lockwood.htm

Apologies for the long post, but I haven’t really outlined my own pet TOE on
this list before, so I wanted to get all the major details in there.

Jesse Mazer

_________________________________________________________________
MSN Messenger with backgrounds, emoticons and more.
http://www.msnmessenger-download.com/tracking/cdp_customize

Bruno Marchal

unread,
Jan 31, 2004, 10:17:09 AM1/31/04
to everyth...@eskimo.com
Here is an interesting post by Jesse. Curiously I have not been able to find it
in the archive, but luckily I find it in my computer memory.

Is that normal? I will try again later.

Jesse's TOE pet is very similar to the type of TOE compatible with the comp
hyp, I guess everyone can see that.

Jesse, imo, that post deserves to be developed. The way you manage to save
partially the ASSA (Absolute Self-Sampling Assumption) is not very clear to me.

Bruno

Jesse Mazer

unread,
Feb 1, 2004, 12:07:36 AM2/1/04
to everyth...@eskimo.com
>From: Bruno Marchal <mar...@ulb.ac.be>
>To: everyth...@eskimo.com
>Subject: Re: Request for a glossary of acronyms
>Date: Sat, 31 Jan 2004 16:11:39 +0100

>
>Here is an interesting post by Jesse. Curiously I have not been able to
>find it
>in the archive, but luckily I find it in my computer memory.
>
>Is that normal? I will try again later.

Thanks for reviving this post, it's in the archives here:
http://www.escribe.com/science/theory/m4882.html
It was part of this thread:
http://www.escribe.com/science/theory/index.html?by=OneThread&t=Request%20for%20a%20glossary%20of%20acronyms

>
>Jesse's TOE pet is very similar to the type of TOE compatible with the comp
>hyp, I guess everyone can see that.
>
>Jesse, imo, that post deserves to be developed. The way you manage to save
>partially the ASSA (Absolute Self-Sampling Assumption) is not very clear to
>me.
>
>Bruno

Well, the idea I discussed was somewhat vague, I think to develop it I'd
need to have better ideas about what a theory of consciousness should look
like, and I don't know where to begin with that. But as for how the ASSA is
incorporated, I'll try to summarize again and maybe make it a little
clearer. Basically my idea was that there would be two types of measures on
observer-moments: a relative measure, which gives you answers to questions
like "if I am currently experiencing observer-moment A, what is the
probability that my next experience will be of observer-moment B?", and an
absolute measure, which is sort of like the probability that my current
observer-moment will be A in the first place. This idea of absolute measure
might seem meaningless since whatever observer-moment I'm experiencing right
now, from my point of view the probability is 1 that I'm experiencing that
one and not some other, but probably the best way to think of it is in terms
of the self-sampling assumption, where reasoning *as if* I'm randomly
sampled from some group (for example, 'all humans ever born' in the doomsday
argument) can lead to useful conclusions, even if I don't actually believe
that God used a random-number generator to decide which body my preexisting
soul would be placed in.

So, once you have the idea of both a relative measure
('probability-of-becoming') and an absolute measure ('probability-of-being')
on observer-moments, my idea is that the two measures could be interrelated,
like this:

1. My probability-of-becoming some possible future observer-moment is based
both on something like the 'similarity' between that observer-moment and my
current one (so my next experience is unlikely to be that of George W. Bush
sitting in the White House, for example, because his memories and
personality are so different from my current ones) but also on the absolute
probability of that observer-moment (so that I am unlikely to find myself
having the experience of talking to an intelligent white rabbit, because
even if that future observer-moment is fairly similar to my current one in
terms of personality, memories, etc., white-rabbit observer-moments are
objectively improbable). I don't know how to quantify "similarity" though,
or exactly how both similarity and absolute probabilities would be used to
calculate the relative measure between two observer-moments...this is where
some sort of "theory of consciousness" would be needed.

2. Meanwhile, the absolute measure is itself dependent on the relative
measure, in the sense that an observer-moment A will have higher absolute
measure if a lot of other observer-moments that themselves have high
absolute measure see A as a likely "next experience" or a likely "past
experience" (ie there's a high relative measure between them). This idea is
based partly on that thought experiment where two copies of a person are
made, then one copy is itself later copied many more times, the idea being
that the copy that is destined to be copied more in the future has a higher
absolute measure because there are more future observer-moments
"reinforcing" it (see http://www.escribe.com/science/theory/m4841.html for
more on this thought-experiment). I think of this whole idea in analogy to
the way Google's ranking system works: pages are ranked as more popular if
they are linked to by a lot of other pages that are themselves highly
ranked. So, the popularity of a particular page is sort of like the absolute
probability of being a particular observer-moment, while a link from one
page to another is like a high relative probability from one observer-moment
to another (to make the analogy better you'd have to use weighted links, and
you'd have to assume the weight of the link between page A and page B itself
depends partly on B's popularity).

The final part of my pet theory is that by having the two measures
interrelated in this way, you'd end up with a unique self-consistent
solution to what each measure would look like, like what happens when you
have a bunch of simultaneous equations specifying how different variables
relate to one another, and they determine a unique solution. This would
provide a rationale for having a non-arbitrary choice of absolute and
relative measure (see my comments about the 'arbitrariness problem' in my
very first post on this list at
http://www.escribe.com/science/theory/m2606.html ). Also, this type of TOE
would give a precise answer to the 'problem of the reference class' which
Nick Bostrom talks about in his work on the self-sampling assumption, the
answer being that you should reason as if you were randomly sampled from the
set of all observer-moments, weighted by their absolute measure. The final
benefit of this type of theory is that you wouldn't need a two-step
procedure of first coming up with a measure on "universes" and then
afterwards adding anthropic considerations as a second step--I think that
two-step idea depends on a fundamentally dualistic view of the mind/body
problem, as I said in my post at
http://www.escribe.com/science/theory/m5069.html

So, does any of this help explain how I try to save the ASSA? I guess it
depends on what you think the basic problem the ASSA has that requires it to
be "saved", I think Hal Finney was saying the problem was that it could lead
to predictions incompatible with those of the RSSA, while others seem to
have more of a philosophical problem with talking about the "probability"
that my current observer-moment could be anything other than what it
actually is (if 'I' were someone else, 'I' wouldn't be me!) What is your
basic objection to the ASSA, and do you think my pet theory offers at least
one possible way to resolve it?

Jesse Mazer

_________________________________________________________________
Find high-speed ‘net deals — comparison-shop your local providers here.
https://broadband.msn.com

Bruno Marchal

unread,
Feb 3, 2004, 10:18:03 AM2/3/04
to everyth...@eskimo.com
Thank you Jesse for your clear answer. Your comparison
of your use of both ASSA and RSSA with Google ranking system
has been quite useful.
This does not mean I am totally convince because ASSA raises the
problem of the basic frame: I don't think there is any sense to compare
the probability of "being a human" or "being a bacteria" ..., but your
"RSSA use of ASSA" does not *necessarily* give a meaning to such
strong form of absolute Self Sampling Assumption, or does it?
I think also that your view on RSSA is not only compatible with
the sort of approach I have developed, but is coherent with
"Saibal Mitra" backtracking, which, at first I have taken
as wishful thinking. OK you make me feel COMP could be a little less
frightening I'm use to think.
Concerning consciousness theory and its use to isolate a similarity
relation on the computational histories---as seen from some first person
point of view, I will try to answer asap in a common answer to
Stephen and Stathis (and you) who asked very related questions.
Alas I have not really the time now---I would also like to find a way to
explain
the consciousness theory without relying too much on mathematical logic,
but the similarity between 1-histories *has* been derived technically in
the part
of the theory which is the most counter-intuitive ... mmh I will try soon ...

Bruno

Reply all
Reply to author
Forward
0 new messages