[KevinTryon: Jacques Mallah]

[Bruno Marchal]

On 11 Feb 2009, at 18:30, Saibal Mitra wrote:

>
> Welcome back Jack Mallah!
>
> I have a different argument against QTI.
>
> I had a nice dream last night, but unfortunately it suddenly ended.
> Now, this is empirical evidence against QTI because, according to the
> QTI, the life expectancy of the version of me simulated in that dream
> should have been be infinite.




The notion of "ending" makes sense only relatively to something  
"ending latter or not ending".

With most definition of first person (including both UDA and AUDA  
definitions) "first person ending" just makes no sense.

Topologically, life is an open set.

Bruno



http://iridia.ulb.ac.be/~marchal/




--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Bruno Marchal]

On 11 Feb 2009, at 20:47, Brent Meeker wrote:

>
> Bruno Marchal wrote:
>>
>> On 11 Feb 2009, at 00:38, Günther Greindl wrote:
>>
>>> I'm with Mike and Brent.
>>>
>>> Bruno, giving A1 and A2 mirrors which would show different stuff
>>> violates Stathis' assumption of running the _same_ computation - you
>>> can't go out of the system.
>>
>> See my answer to Brent. Once A1 looks at itself in the mirror (and
>> thus A2 too, given the protocol). A1 sees MA1 and A2 sees MA2, and  
>> the
>> computation differs.
>
> If A1 sees MA1 and A2 sees MA2 and they see something different,  
> i.e. MA1 and
> MA2 are distinguishable, then you've violated the hypothesis that the
> computations are identical.

Right? I did change the protocol to make my point, which concerns only  
the probability of finding myself by MA1 or by MA2, but not both.

Bruno





http://iridia.ulb.ac.be/~marchal/




--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Bruno Marchal]

On 11 Feb 2009, at 21:51, Brent Meeker wrote:

>
> Saibal Mitra wrote:
>> Welcome back Jack Mallah!
>>
>> I have a different argument against QTI.
>>
>> I had a nice dream last night, but unfortunately it suddenly ended.
>> Now, this is empirical evidence against QTI because, according to the
>> QTI, the life expectancy of the version of me simulated in that dream
>> should have been be infinite.
>
> Of course maybe in some other branch of the multiverse your dream is  
> continuing.
> That's what makes everything-theories difficult to test.
>
> But you raise an interesting point. Everything-theories that suppose
> consciousness is constituted by the closest continuations need to  
> solve the
> "white rabbit" problem.  But that solution, whatever it is, would  
> equally apply
> in dreams.  So why don't dreams have the same physics as waking life?

Ah! That is a good question. It is equivalent to the first person  
white rabbit problem (which was the point of the original white rabbit  
problem in "conscience and mécanisme".

The answer is that dreams are really stabilized by their relative  
apparition with respect to deep computations with high measure. For  
such relativity we need a mechanist notion of first person PLURAL, and  
then I hope the arithmetical hypostases will confirmed this the  
working of such notion.

Bruno


http://iridia.ulb.ac.be/~marchal/




--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Bruno Marchal]

On 11 Feb 2009, at 22:19, Brent Meeker wrote:


>
> This idea seems inconsistent with MWI.  In QM the  split is uncaused  
> so it's
> hard to see why its influence extends into the past and increases  
> the measure of
> computations that were identical before the split.

I got the inspiration from the MWI, and even from David Deutsch  
convincing point that conceptually differentiation-talk is less wrong  
than bifurcation-talk. But is is not simply, in QM, a consequence of  
the linearity of the tensor product?,  i.e. the fact that the state  
A*(B+C) is equivalent with (A*C)+(A*C), where A, B, C represents kets  
and * represents the tensor product.
Of course the price to pay, as Everett first noticed, is that the  
states become a relative notion, and the probabilities too, making  
RSSA obligatory in QM. With comp it is more subtle (but then Everett  
uses comp and missed or abstracted himself from this subtlety).

>> Of course we still lack a definite criteria of identity for
>> computation. But we can already derive what can count as different
>> computations if we want those measure question making sense.
>
> As I understand it your theory of personal identity depends on  
> computations
> "going through" a particular state.  Intuitively this implies a  
> state at a
> particular moment, but a Y=II representation implies that we are  
> taking into
> account not just the present state but some period of history -  
> which would
> correspond with the usual idea of a person - something with a  
> history, not just
> a state.

Absolutely so. It is the Darwinistic aspect of comp. A species with a  
lot of offsprings makes higher the "time life" of old gene.
Perhaps thats why it is said we should grow and multiply  :)

Bruno


http://iridia.ulb.ac.be/~marchal/




--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Jack Mallah]

--- On Thu, 2/12/09, George Levy <glevy@...> wrote:
> I have also been overwhelmed by the volume on this list.
> The idea is not to take more than you can chew.

Indeed.

> > --- On Wed, 2/11/09, George Levy
> > If that were the case, the Born Rule would fail.
> Perhaps the probability rule would be more like proportionality to norm^2 exp(entropy) instead of just norm^2.  If that was it, then for example unstable nuclei would be observed to decay a lot faster than the Born Rule predicts.

> I do not understand why you say that the Born rule would fail.

High entropy branches would have more probability than low entropy ones compared to the standard Born rule.
   
> Yes I am linking the entropy to MW branching.

But you should realize that the Born rule is self-consistent in the face of branching.  If there is branching to N states, then on average the squared norm of each will be 1/N of the original.  That much is proven by the math.  Linking squared norm to measure is of course a tougher issue.

> You say that the Born Rule would fail if measure *increases*.

Actually, all I said was that it would fail if measure is linked to entropy.  Any significant modification to it would make it fail.

> Using your own argument I could say that the Born rule would fail if measure *decreases *according to function f(t). For example it could be norm^2 f(t) .

That would make it fail but if the modification is only a function of time it would be hard to detect.  Making it a function of a branch-dependent observable like entropy leads to a much easier-to-detect deviation.

> So using your own argument since the Born rule is only norm^2 therefore measure stays constant?

In ordinary experimental situations, total measure stays constant.

In life or death situations there is a correction factor but it is well known: the measure in a given world is proportional to the number of people alive in it as well as to the squared norm.  This is taken into account under the Anthropic principle, and explains why our universe seems fine-tuned for life even though worlds like that presumably have a relatively small total squared norm compared to the sum of the others.




     


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Bruno Marchal]

On 12 Feb 2009, at 02:59, Jack Mallah wrote:

>
> Hi George.  The everything list feels just like old times, no?


I am afraid we are just a bit bactracking 10 years ago.
No problem. After all, concerning theology, I am asking people to  
backtrack 1500 years ago (1480 to be precise).

> Which is nice in a way but has a big drawback - I can only take so  
> much of arguing the same old things, and being outnumbered.  And  
> that limit is approaching fast again.  At least I think your point  
> here is new to the list.
>
> --- On Wed, 2/11/09, George Levy <glevy@...> wrote:
>> One could argue that measure actually increases continuously and  
>> corresponds to the increase in entropy occurring in everyday life.  
>> So even if you are 90 or 100 years old you could still experience  
>> an increase in measure.
>
> I guess you are basing that on some kind of branch-counting idea.
>
> If that were the case, the Born Rule would fail.  Perhaps the  
> probability rule would be more like proportionality to norm^2  
> exp(entropy) instead of just norm^2.  If that was it, then for  
> example unstable nuclei would be observed to decay a lot faster than  
> the Born Rule predicts.
>
> Conventional half life calculations are accurate.  So either entropy  
> would not be a factor, or the MWI is experimentally disproven  
> already.  Well, if it is a weak enough function of entropy then  
> maybe it hasn't been disproven, but inclusion of free parameters  
> like that which can always be made small enough goes against Occam's  
> Razor.  Otherwise there'd be no end of possible correction factors.
>
> At least your idea was testable, with none of the meaningless "first  
> person" sloganeering.  Ideas like that, keep em' coming!

So you stop at step two of the UDA?
What is wrong with the definition of first and third person views  
notion? I gave a complete third person definition of both notions.
(see the SANE 2004 paper). Or look at the arithmetical definition (the  
Theaetetic one);

>
>
>> In any case, measure is measured over a continuum and its value is  
>> infinite to begin with. So whether it increases or decreases may be  
>> a moot point.
>
> It's not moot.  Just take density ratios.  The size of the universe  
> may be infinite, but that didn't stop Hubble from saying it's  
> getting bigger.
>
>> As I said, the increase or decrease in measure is at the crux of  
>> this problem.  Your paper really did not illuminate the issue in a  
>> satisfactory manner.
>
> It could no doubt use some tweaking, which is why I'm on the list  
> now.  I know I'm not always a good communicator.  What should be  
> clarified or added to it?

You say: "no randomness involved" but you seem to accept  
probabilities. Do I just miss something here?
You seem not taking the 1 pov / 3 pov distinction seriously into  
account. What does mean "questioning immortality" then?

Bruno




http://iridia.ulb.ac.be/~marchal/




--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Bruno Marchal]

On 12 Feb 2009, at 05:38, Tom Caylor wrote:

>
> But of course you would worry just as much if the clone were replaced
> by a zombie...  I guess that gets back to the distinction between
> first person and third person.


It seems to me that is the problem indeed. At the same time, it seems  
obvious to me that if you drop that distinction, you are eliminating  
the first person. Then we are not just mortal, we are already dead.

Bruno

>
>
> On Feb 11, 9:05 pm, Tom Caylor <daddycay...@...> wrote:
>> The effects of have clones is interesting, though, regardless of the
>> "sapping strength" notion.  You would have reason to worry about  
>> being
>> killed if there were clones and then a "shell game" was played with
>> you being mixed up with the clones, and then all of the "yous" were
>> killed except one.  All of the "yous" would have reason to worry.
>> This has implications on ethics of cloning and killing clones.
>> As far as measure, it seems that having a clone of you and killing  
>> one
>> of you while you were asleep would be equivalent (w.r.t how much you
>> should worry at least) to not having any clones and someone saying
>> they were going to roll a die and if it came up odd they would kill
>> you.
>>
>> Tom
>>
>> On Feb 11, 8:44 pm, Jack Mallah <jackmal...@...> wrote:
>>
>>
>>
>>> --- On Wed, 2/11/09, Stathis Papaioannou <stath...@...> wrote:
>>
>>>> You agree that if one version of me goes to bed tonight and one  
>>>> version of me wakes up tomorrow, then I should expect to wake up  
>>>> tomorrow. But if extra versions of me are manufactured and run  
>>>> today, then switched off when I go to sleep, then you are saying  
>>>> that I might not wake up tomorrow.
>>
>>> You won't know this evening if you are one of the "extra versions"  
>>> or the original.  So yes, in that situation, you will probably not  
>>> be around tomorrow.  Only the original will.
>>
>>>> The extra copies of me have somehow sapped my life strength.
>>
>>> Not at all.  I guess that is a joke?
>>
>>> Creating more copies, then getting rid of the same number, does  
>>> not result in a net decrease in measure.  That is why the movie  
>>> "The Prestige" bears no resemblance whatsoever to QS despite  
>>> rumors to the contrary.
>>
>>> If you create extra copies and leave them alive, there is a net  
>>> increase in measure.  That is equivalent to new people being born  
>>> even if they have your memories.  This once happenned to Will  
>>> Riker on Star Trek: TNG.- Hide quoted text -
>>
>> - Show quoted text -
> >

http://iridia.ulb.ac.be/~marchal/




--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Jack Mallah]

--- On Wed, 2/11/09, Brent Meeker <meekerdb@...> wrote:
> >> Two copies don't increase the measure of a computation and reducing it's vector in Hilbert space doesn't diminish it.
> >
> > If that is so then how do you explain the Born rule?
>
> The Born rule assumes you start with a normalized vector (i.e. ray), so it calculates predicted probabilities conditional on the state preparation.  After each measurement, the vector is renormalized because the prediction is always conditional on the present state.

How do you explain why it works?  I say it is because people in higher amplitude branches have more measure.

> This is quite different from applying a probability measure to the evolution of a multiverse in which decoherence defines many different orthogonal subspaces, each of which gets a small projection of the state vector of the multiverse.

Then it is not the standard MWI in the Everett tradition.




     


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Brian Tenneson]

Thanks Günther.  I'll look into it. 

I also came across this article and thought some people might want to discuss it.
http://arxiv.org/abs/0810.4339

abstract
We employ the Zermelo-Fraenkel Axioms that characterize sets as mathematical primitives. The Anti-foundation Axiom plays a significant role in our development, since among other of its features, its replacement for the Axiom of Foundation in the Zermelo-Fraenkel Axioms motivates Platonic interpretations. These interpretations also depend on such allied notions for sets as pictures, graphs, decorations, labelings and various mappings that we use. A syntax and semantics of operators acting on sets is developed. Such features enable construction of a theory of non-well-founded sets that we use to frame mathematical foundations of consciousness. To do this we introduce a supplementary axiomatic system that characterizes experience and consciousness as primitives. The new axioms proceed through characterization of so- called consciousness operators. The Russell operator plays a central role and is shown to be one example of a consciousness operator. Neural networks supply striking examples of non-well-founded graphs the decorations of which generate associated sets, each with a Platonic aspect. Employing our foundations, we show how the supervening of consciousness on its neural correlates in the brain enables the framing of a theory of consciousness by applying appropriate consciousness operators to the generated sets in question.

On Sun, Feb 8, 2009 at 7:49 AM, Günther Greindl <guenther.greindl@...> wrote:

Brian,

Tononi's information integration view of consciousness might fit your bill.

Overview:
http://www.spectrum.ieee.org/jun08/6315

Paper (open access):
http://www.biomedcentral.com/1471-2202/5/42

Cheers,
Günther

Brian Tenneson wrote:
> Dear Everything List,
>
> Tegmark mentioned in an article the idea of self-aware structures, SAS.
> He wrote that the search for such structures is ongoing, i.e., he
> postulated the existence of such structures without giving examples.
>
> I'm wondering if consciousness and self-awareness has been
> "mathematized" somewhere, preferably in documents I can download without
> academic affiliation / subscription.
>
> I'm inclined to thing agents might be a pathway to this end, as well as
> what David Wolpert is calling a device.
>
> Do any of the formalizations come close to being reflective of human
> consciousness?  In other words a mathematical model of human consciousness?
>
> Thank you.
>
> >
>

--
Günther Greindl
Department of Philosophy of Science
University of Vienna
guenther.greindl@...

Blog: http://www.complexitystudies.org/
Thesis: http://www.complexitystudies.org/proposal/

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list@...
To unsubscribe from this group, send email to everything-list+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---