Superdeterminism And Sabine Hossenfelder

30 views
Skip to first unread message

John Clark

unread,
Dec 19, 2021, 8:25:58 AM12/19/21
to 'Brent Meeker' via Everything List
Sabine Hossenfelder recently posted this video on Youtube, this is my comment: 


I strongly agree with Sabine Hossenfelder that "free will" is incoherent nonsense, but I strongly disagree with her advocacy of superdeterminism. Even if the laws of physics were as deterministic as Newton thought they were and you knew all of them you still couldn't make a prediction unless you knew the initial conditions, that's why I think "superdeterminism" is a pretty good name. When scientists talk about plain old vanilla style Newtonian "determinism" they're only talking about the laws of physics, but superdeterminism means more than that, it's also talking about initial conditions. Occam's Razor says that if 2 theories agree with observations equally well then the theory with the fewest assumptions (NOT the fewest outcomes) is the one to be preferred. It would be absolutely impossible for superdeterminism to contain more assumptions than it does, depending on if the universe is infinite or not and if space and time are quantized or continuous, superdeterminism demands either an astronomical number to an astronomical power of independent assumptions, or more likely an infinite number of such assumptions.  You can get more out of a good theory then you put into it, in fact that's what a "good theory" means. but that would be impossible with superdeterminism because it requires an infinite input.  

Superdeterminism violates Occam's Razor just as badly as the God hypothesis does because they both need to invoke infinity in their assumptions. Superdeterminism assumes that out of the (probably) infinite number of states the universe could've been in at the time of the Big Bang it was actually in the one and only one specific state that would prevent experimenters on the planet Earth 13.8 billion years later from ever performing a simple experiment that would unequivocally show that the world is indeed deterministic, the God hypothesis assumes the existence of an infinitely powerful infinitely intelligent being. By contrast the Many Worlds Theory only makes one assumption, Schrodinger's Equation means what it says. So Many Worlds wins.

 
John K Clark    See what's on my new list at  Extropolis

Jesse Mazer

unread,
Dec 19, 2021, 12:09:02 PM12/19/21
to everyth...@googlegroups.com
Yes, it's misleading for her to suggest the objections to superdeterminism are mainly about "killing free will", rather they're about the way the theory would need a strange "conspiracy" in the initial conditions of the universe to work. The idea is that if two entangled particles are sent out from an emitter to two experimenters in opposite directions, the particles carry hidden variables that predetermine what response they will give to the measurements, and that the hidden variables they are assigned are somehow correlated what variables the two experimenters are going to choose to measure in the future. But even if the experimenters' choice involves no free will, it may depend in a complicated way on events throughout the past light cone of their decisions (as with the butterfly effect in chaos theory), including events right after the Big Bang that are outside the past light cone of the two particles being 'assigned' their hidden variables (most likely when the two particles were generated by the emitter and sent on their way to the experimenters). So you need a conspiracy in the initial state of the universe at the time of the Big Bang to ensure that billions of years later when intelligent beings evolve and do these sorts of experiments, making their choices of measurement variables in whatever arbitrary way they select, the initial conditions in their past light cones that determine the outcome of their selection would always be correlated with the initial conditions in the past light cone of the emission event in a way that ensured the hidden variables and the selected measurements were always statistically correlated in just the right way.

Hossenfelder and a co-author address the conspiracy objection in section 4.2 of their paper at https://www.frontiersin.org/articles/10.3389/fphy.2020.00139/full but the answer seems pretty weak, they basically say "yes it might seem like you'd need some very complicated and arbitrary constraint on the initial conditions, but you can't be *sure* the constraint doesn't actually have high algorithmic compressibility." Would be a lot more convincing if someone could come up with an actual toy model of superdeterminism where the experimenters are treated as complex classical computational systems whose "choices" on each experiment involve chaos theory style sensitive dependence on initial conditions.

Jesse

--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/everything-list/CAJPayv1YBWRDtagjez97hQGcFgmboKSqZ2PyX_1JvxQP64R9xg%40mail.gmail.com.

Lawrence Crowell

unread,
Dec 19, 2021, 3:28:01 PM12/19/21
to Everything List
This invokes some ideas of p-adic numbers. If Sabine is right and superdeterminism is a global law it would require a single consistent algorithm for solving p-adic problems, which are equivalent to Diophantine equations.  Matiyasevich proved a variant of the Godel theorem which disproved this, which is also HIlbert's 10th problem. As a result only local solutions are possible. If Sabine wants superdeterminism, then fine, but it means different observers will interpret reality in fundamentally different, even contradictory, ways. We are seeing something like this with recent demonstrations of Wigner's friend thought experiment. The Frauchiger-Renner work showed how Wigner's friend thought experiment precluded reality if one assumed local hidden variables. This occurs with superdeterminism as well. 

LC

Brent Meeker

unread,
Dec 19, 2021, 7:59:22 PM12/19/21
to everyth...@googlegroups.com


On 12/19/2021 5:25 AM, John Clark wrote:
By contrast the Many Worlds Theory only makes one assumption, Schrodinger's Equation means what it says. So Many Worlds wins.

It also makes the assumption that the eigenvalues of a measurement are realized probabilistically.

Brent




John Clark

unread,
Dec 19, 2021, 8:23:00 PM12/19/21
to 'Brent Meeker' via Everything List
What is the eigenvalue of a temperature of 72°F? It doesn't have one. A measurement doesn't have an eigenvalue but a matrix does, such as the one that describes the Schrodinger Wave. And no quantum interpretation needs to assume there is a relationship between the square of the absolute value of that wave and probability because it is observed to be true. If it were not true Schrodinger's Wave would be completely useless and there would be no reason any physicist would bother to calculate it.

John K Clark    See what's on my new list at  Extropolis
eqa



 

Bruce Kellett

unread,
Dec 19, 2021, 9:05:55 PM12/19/21
to Everything List
On Mon, Dec 20, 2021 at 12:23 PM John Clark <johnk...@gmail.com> wrote:
On Sun, Dec 19, 2021 at 7:59 PM Brent Meeker <meeke...@gmail.com> wrote:

On 12/19/2021 5:25 AM, John Clark wrote:
By contrast the Many Worlds Theory only makes one assumption, Schrodinger's Equation means what it says. So Many Worlds wins.

> It also makes the assumption that the eigenvalues of a measurement are realized probabilistically.


What is the eigenvalue of a temperature of 72°F? It doesn't have one. A measurement doesn't have an eigenvalue but a matrix does, such as the one that describes the Schrodinger Wave. And no quantum interpretation needs to assume there is a relationship between the square of the absolute value of that wave and probability because it is observed to be true.

The Born Rule cannot be derived from the Schrodinger equation; it has to be added as a further independent assumption. So it is not true that Many Worlds makes only one assumption. It requires just as many assumptions as collapse theories.

Bruce

John Clark

unread,
Dec 19, 2021, 9:13:38 PM12/19/21
to 'Brent Meeker' via Everything List
On Sun, Dec 19, 2021 at 9:05 PM Bruce Kellett <bhkel...@gmail.com> wrote:


> The Born Rule cannot be derived from the Schrodinger equation; it has to be added as a further independent assumption. So it is not true that Many Worlds makes only one assumption.

No quantum interpretation needs to derive the Schrodinger Equation nor does it need to be assumed because it can be experimentally verified to be true. And no quantum interpretation is inconsistent with observation, at least not so far. 
John K Clark    See what's on my new list at  Extropolis
070
 

Bruce Kellett

unread,
Dec 19, 2021, 10:04:38 PM12/19/21
to Everything List
On Mon, Dec 20, 2021 at 1:13 PM John Clark <johnk...@gmail.com> wrote:
On Sun, Dec 19, 2021 at 9:05 PM Bruce Kellett <bhkel...@gmail.com> wrote:

> The Born Rule cannot be derived from the Schrodinger equation; it has to be added as a further independent assumption. So it is not true that Many Worlds makes only one assumption.

No quantum interpretation needs to derive the Schrodinger Equation nor does it need to be assumed because it can be experimentally verified to be true. And no quantum interpretation is inconsistent with observation, at least not so far. 

Why do we need any theory at all then? We just have to observe the experimental results and they are true. Perhaps science is about understanding the experimental results, not just accepting them as the truth.

Bruce

Brent Meeker

unread,
Dec 19, 2021, 10:38:56 PM12/19/21
to everyth...@googlegroups.com


On 12/19/2021 5:22 PM, John Clark wrote:


On Sun, Dec 19, 2021 at 7:59 PM Brent Meeker <meeke...@gmail.com> wrote:

On 12/19/2021 5:25 AM, John Clark wrote:
By contrast the Many Worlds Theory only makes one assumption, Schrodinger's Equation means what it says. So Many Worlds wins.

> It also makes the assumption that the eigenvalues of a measurement are realized probabilistically.


What is the eigenvalue of a temperature of 72°F? It doesn't have one. A measurement doesn't have an eigenvalue but a matrix does, such as the one that describes the Schrodinger Wave. And no quantum interpretation needs to assume there is a relationship between the square of the absolute value of that wave and probability because it is observed to be true.

A temperature operator, which would be matrix, might very well return 72degF as the eigenvalue of a state eigenvector.  It's Hermitean operators that have eigenvalues and eigenvectors.  Decoherence theory shows that interactions with an environment approximately diagonalize a density matrix, I don't think anyone has shown how the Hamiltonian of the system+instrument+environment does this; they just sort of assume that it does.  Zeh has proposed a Darwinian selection like effect but I don't see how it's worked out as the level of the Schroedinger eqn.

Yes, it's empirically supported; So's the Schroedinger equation.  But it's part of the application of the Schroedinger equation.  It's not in the equation itself. 

Brent

If it were not true Schrodinger's Wave would be completely useless and there would be no reason any physicist would bother to calculate it.

John K Clark    See what's on my new list at  Extropolis
eqa



 
--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.

Brent Meeker

unread,
Dec 19, 2021, 10:49:05 PM12/19/21
to everyth...@googlegroups.com


On 12/19/2021 6:12 PM, John Clark wrote:


On Sun, Dec 19, 2021 at 9:05 PM Bruce Kellett <bhkel...@gmail.com> wrote:


> The Born Rule cannot be derived from the Schrodinger equation; it has to be added as a further independent assumption. So it is not true that Many Worlds makes only one assumption.

No quantum interpretation needs to derive the Schrodinger Equation nor does it need to be assumed because it can be experimentally verified to be true. And no quantum interpretation is inconsistent with observation, at least not so far.

It can't be experimentally verified that the other world branches exist and the Schrodinger equation cannot be verified except statistically by assuming the Born rule.  Without the Born rule the Schroedinger equation just says that a lot of mutually contradictory possibilities have evolved in the state vector.

Brent


John K Clark    See what's on my new list at  Extropolis
070
 
--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.

spudb...@aol.com

unread,
Dec 20, 2021, 3:29:43 AM12/20/21
to meeke...@gmail.com, everyth...@googlegroups.com
Without invoking MWI which I adore, let us focus upon the less grandiose and ask can one entangle a tardigrade or can't one?

smitra

unread,
Dec 20, 2021, 4:03:40 AM12/20/21
to everyth...@googlegroups.com
On 20-12-2021 03:05, Bruce Kellett wrote:
> On Mon, Dec 20, 2021 at 12:23 PM John Clark <johnk...@gmail.com>
> wrote:
>
>> On Sun, Dec 19, 2021 at 7:59 PM Brent Meeker <meeke...@gmail.com>
>> wrote:
>>
>> On 12/19/2021 5:25 AM, John Clark wrote:
>> By contrast the Many Worlds Theory only makes one assumption,
>> Schrodinger's Equation means what it says. So Many Worlds wins.
>>
>> _> It also makes the assumption that the eigenvalues of a
>> measurement are realized probabilistically._
>
> What is the eigenvalue of a temperature of 72°F? It doesn't have one.
> A measurement doesn't have an eigenvalue but a matrix does, such as
> the one that describes the Schrodinger Wave. And no quantum
> interpretation needs to assume there is a relationship between the
> square of the absolute value of that wave and probability because it
> is observed to be true.
>
> The Born Rule cannot be derived from the Schrodinger equation; it has
> to be added as a further independent assumption. So it is not true
> that Many Worlds makes only one assumption. It requires just as many
> assumptions as collapse theories.
>
> Bruce

Yes, but with those assumptions it yields an unambiguous framework for a
fundamental theory. In case of collapse theories, you're stuck with a
phenomenological theory that cannot be improved, because you are not
allowed to describe observers and observations within the collapse
frameworks. It's a bit like the difference between statistical mechanics
and thermodynamics, if in the latter case textbooks were to insist that
you are only allowed to consider certain types of heat engines that
operate in the quasistatic limit.

Saibal

>
>> If it were not true Schrodinger's Wave would be completely useless
>> and there would be no reason any physicist would bother to calculate
>> it.
>
> --
> You received this message because you are subscribed to the Google
> Groups "Everything List" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to everything-li...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/everything-list/CAFxXSLTFdzA%3Dg9GTTiC2aLdJZ76tHYA3Bvxo2WrrvdnAXY-QQg%40mail.gmail.com
> [1].
>
>
> Links:
> ------
> [1]
> https://groups.google.com/d/msgid/everything-list/CAFxXSLTFdzA%3Dg9GTTiC2aLdJZ76tHYA3Bvxo2WrrvdnAXY-QQg%40mail.gmail.com?utm_medium=email&utm_source=footer

Bruce Kellett

unread,
Dec 20, 2021, 5:28:15 AM12/20/21
to Everything List
On Mon, Dec 20, 2021 at 8:03 PM smitra <smi...@zonnet.nl> wrote:
On 20-12-2021 03:05, Bruce Kellett wrote:

> The Born Rule cannot be derived from the Schrodinger equation; it has
> to be added as a further independent assumption. So it is not true
> that Many Worlds makes only one assumption. It requires just as many
> assumptions as collapse theories.
>
> Bruce

Yes, but with those assumptions it yields an unambiguous framework for a
fundamental theory. In case of collapse theories, you're stuck with a
phenomenological theory that cannot be improved, because you are not
allowed to describe observers and observations within the collapse
frameworks.


I think you are stuck on a very old-fashioned view of collapse theories -- perhaps you are thinking only in terms of theories dominated by Bohr's idea of a separation between the quantum and the classical -- with the classical world necessary to give quantum results meaning. In other words, a fundamental separation between the observer and the observed. This, of course, is problematic in that you cannot describe the observer in quantum terms.

But modern collapse theories, such as Flash-GRW, do not have this limitation. There is no observer/observed distinction in such theories, and they can easily accommodate the idea that everything, including the observer, is quantum.

Besides, MWI is far from unambiguous. For instance, the notion of probability is decidedly problematic in Everettian theory.

Bruce

Bruce Kellett

unread,
Dec 20, 2021, 5:32:14 AM12/20/21
to Everything List
On Mon, Dec 20, 2021 at 7:29 PM spudboy100 via Everything List <everyth...@googlegroups.com> wrote:
Without invoking MWI which I adore, let us focus upon the less grandiose and ask can one entangle a tardigrade or can't one?


John Clark

unread,
Dec 20, 2021, 6:54:56 AM12/20/21
to 'Brent Meeker' via Everything List
On Sun, Dec 19, 2021 at 10:04 PM Bruce Kellett <bhkel...@gmail.com> wrote:

>>> The Born Rule cannot be derived from the Schrodinger equation; it has to be added as a further independent assumption. So it is not true that Many Worlds makes only one assumption.

>> No quantum interpretation needs to derive the Schrodinger Equation nor does it need to be assumed because it can be experimentally verified to be true. And no quantum interpretation is inconsistent with observation, at least not so far. 

> Why do we need any theory at all then? We just have to observe the experimental results and they are true. Perhaps science is about understanding the experimental results, not just accepting them as the truth.

Some productive scientists are satisfied with the Shut Up And Calculate quantum "interpretation" and that's fine, there is no disputing matters of taste, but some who don't dislike philosophy would like a bit more. The point I was trying to make was that nobody "assumes" the Born Rule anymore than somebody assumes that a body at rest or moving at a constant speed in a straight line will remain at rest or keep moving in a straight line at constant speed unless it is acted upon by a force. Neither the Born Rule or Newton's first law of motion were "assumed" to be true, they were OBSERVED to be true , and in science observation always outranks theory. If a theory concludes that a certain observation can't occur but it is observed to occur then the theory is wrong. A theory needs to be confirmed by observation, but an observation doesn't need to be confirmed by a theory.

Unfortunately none of the quantum interpretations conflicts with observation so to decide on a favorite one should pick the one that makes the fewest assumptions (NOT the one that produces the simplest outcome). And that's why I like Many Worlds, it only makes one assumption. And that's why I think superdeterminism is the very worst quantum interpretation possible, it needs, quite literally, an infinite number of assumptions to work. If that was the best anybody could come up with I'd stick with Shut Up And Calculate, but fortunately we can do better.

Of coursecan't deny it would be great if a quantum interpretation could lead us straight to the Born Rule. It may be premature to claim victory but I think Many Worlds has made much more progress towards accomplishing that goal than any other: 


John K Clark    See what's on my new list at  Extropolis
mwc



John Clark

unread,
Dec 20, 2021, 10:17:54 AM12/20/21
to 'Brent Meeker' via Everything List
On Sun, Dec 19, 2021 at 10:38 PM Brent Meeker <meeke...@gmail.com> wrote:
> >> It also makes the assumption that the eigenvalues of a measurement are realized probabilistically.

>> What is the eigenvalue of a temperature of 72°F? It doesn't have one. A measurement doesn't have an eigenvalue but a matrix does, such as the one that describes the Schrodinger Wave. And no quantum interpretation needs to assume there is a relationship between the square of the absolute value of that wave and probability because it is observed to be true.

> A temperature operator, which would be matrix, might very well return 72degF as the eigenvalue of a state eigenvector. 


A temperature measurement taken at a particular time and place is not a temperature operator, and a measurement is not a probability, although the square of the absolute value of a wave function might tell you the probability of you getting that temperature measurement at that time and place.

>  Yes, it's empirically supported; So's the Schroedinger equation.  But it's part of the application of the Schroedinger equation.  It's not in the equation itself. 

I don't know what you mean by that.  

>> No quantum interpretation needs to derive the Schrodinger Equation nor does it need to be assumed because it can be experimentally verified to be true. And no quantum interpretation is inconsistent with observation, at least not so far.

>It can't be experimentally verified that the other world branches exist

But an astronomical number, or even an infinite number, of other world branches is not inconsistent with experiment or observation, and if you want to hypothesize about what's really going on at the deepest level of reality while making the fewest possible assumptions then Many Worlds is your best bet. At least it's the best bet anyone has come up with so far.
 
> and the Schrodinger equation cannot be verified except statistically by assuming the Born rule.  

I must insist yet again that the Born Rule is NOT assumed to be true nor is it required to be derived to be true because we can do far better than either one of those two things. We can observe the Born Rule to be true.

> Without the Born rule the Schroedinger equation [...]
 
Without the Born rule the Schrodinger Wave Equation would be a silly worthless equation of no interest to anyone, but thanks to observation we know for a fact that the Born Rule is true, and that makes Schrodinger's Equation very important indeed.
John K Clark    See what's on my new list at  Extropolis
swq


Jesse Mazer

unread,
Dec 20, 2021, 12:02:00 PM12/20/21
to everyth...@googlegroups.com
When you say the MWI + Born rule "yields an unambiguous framework for a fundamental theory" are you assuming the idea of probability being equal to amplitude squared only applies to "measurements", or that it would somehow apply at all times in the MWI? If the former there would seem to be some ambiguity about what a "measurement" is; if the latter, I believe MWI advocates still don't have an agreed-upon answer to the "preferred basis problem" discussed at https://physics.stackexchange.com/questions/65177/is-the-preferred-basis-problem-solved

Brent Meeker

unread,
Dec 20, 2021, 5:15:33 PM12/20/21
to everyth...@googlegroups.com
Yes, but it is decoherence theory that extends the theory of measurement
beyond just phenomenological projectors.  And it doesn't reach to
explaining the probabilistic nature of QM.  ISTM that the steps in
Everett's account of measurement where instrument variables become
correlated with quantum system variables and cross terms form
superpositions are set to zero are almost has "hand wavy" as the CI
projection operators.   They seem to be just motivated by "This must be
the way the Schroedinger equation works for macroscopic instruments in
order that we get the same answer as the CI projector after we assume
Born's rule."

Brent

Brent Meeker

unread,
Dec 20, 2021, 5:45:51 PM12/20/21
to everyth...@googlegroups.com

The Born rule, understood as probabilities are predicted by state vector amplitudes squared, is not a problem.  Gleason's theorem shows that this is the only mathematically consistent probability measure on a Hilbert space.  The other part of the Born rule, that QM results are probabilistic and depend only on the state vector, does not follow from Schroedinger's equation, although they are natural and well tested hypotheses. 

Where I have doubts about Everett and many-worlds is (1) the many-worlds are NOT observable and have no empirical content and (2) the diagonalization of the density matrix seems to beg the question of how the Schroedinger equation defines a measurement just as much as the projection postulate.  Nobody writes down the Hamiltonian of the instrument and the interaction explicitly and applies the Schroedinger equation; they just assume the Hamiltonian of the instrument and the interaction are such as to act like a projection operator.  Dieter Zeh has suggested that there is a kind quantum Darwinism that produces this result, but I've not seen an explicit calculation showing it.

Brent

--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.

Brent Meeker

unread,
Dec 20, 2021, 5:56:46 PM12/20/21
to everyth...@googlegroups.com


On 12/20/2021 7:17 AM, John Clark wrote:
On Sun, Dec 19, 2021 at 10:38 PM Brent Meeker <meeke...@gmail.com> wrote:
> >> It also makes the assumption that the eigenvalues of a measurement are realized probabilistically.

>> What is the eigenvalue of a temperature of 72°F? It doesn't have one. A measurement doesn't have an eigenvalue but a matrix does, such as the one that describes the Schrodinger Wave. And no quantum interpretation needs to assume there is a relationship between the square of the absolute value of that wave and probability because it is observed to be true.

> A temperature operator, which would be matrix, might very well return 72degF as the eigenvalue of a state eigenvector. 


A temperature measurement taken at a particular time and place is not a temperature operator, and a measurement is not a probability, although the square of the absolute value of a wave function might tell you the probability of you getting that temperature measurement at that time and place.

>  Yes, it's empirically supported; So's the Schroedinger equation.  But it's part of the application of the Schroedinger equation.  It's not in the equation itself. 

I don't know what you mean by that. 

It's the projection postulate in the Copenhagen interpretation that applies the Born rule.  In MWI it's the Born rule plus some kind of self-locating uncertainty to allow for the probabilistic observations.  So those are things not in the Schroedinger equation.



>> No quantum interpretation needs to derive the Schrodinger Equation nor does it need to be assumed because it can be experimentally verified to be true. And no quantum interpretation is inconsistent with observation, at least not so far.

>It can't be experimentally verified that the other world branches exist

But an astronomical number, or even an infinite number, of other world branches is not inconsistent with experiment or observation, and if you want to hypothesize about what's really going on at the deepest level of reality while making the fewest possible assumptions then Many Worlds is your best bet. At least it's the best bet anyone has come up with so far.
 
> and the Schrodinger equation cannot be verified except statistically by assuming the Born rule.  

I must insist yet again that the Born Rule is NOT assumed to be true nor is it required to be derived to be true because we can do far better than either one of those two things. We can observe the Born Rule to be true.

No, you can't observe the Born rule to be true any more (or less) than you can observe Schroedinger's equation to be true.  They are theories that predict a result in every time and place, past and future.  If they fail, even on a set of measure zero, in this infinitude they are invalidated.  Every theory must go beyond what has been observed to be useful...that's the whole point of having theories instead of just catalogues of observations.

Brent


Stathis Papaioannou

unread,
Dec 20, 2021, 7:19:53 PM12/20/21
to everyth...@googlegroups.com
On Tue, 21 Dec 2021 at 09:56, Brent Meeker <meeke...@gmail.com> wrote:


On 12/20/2021 7:17 AM, John Clark wrote:
On Sun, Dec 19, 2021 at 10:38 PM Brent Meeker <meeke...@gmail.com> wrote:
> >> It also makes the assumption that the eigenvalues of a measurement are realized probabilistically.

>> What is the eigenvalue of a temperature of 72°F? It doesn't have one. A measurement doesn't have an eigenvalue but a matrix does, such as the one that describes the Schrodinger Wave. And no quantum interpretation needs to assume there is a relationship between the square of the absolute value of that wave and probability because it is observed to be true.

> A temperature operator, which would be matrix, might very well return 72degF as the eigenvalue of a state eigenvector. 


A temperature measurement taken at a particular time and place is not a temperature operator, and a measurement is not a probability, although the square of the absolute value of a wave function might tell you the probability of you getting that temperature measurement at that time and place.

>  Yes, it's empirically supported; So's the Schroedinger equation.  But it's part of the application of the Schroedinger equation.  It's not in the equation itself. 

I don't know what you mean by that. 

It's the projection postulate in the Copenhagen interpretation that applies the Born rule.  In MWI it's the Born rule plus some kind of self-locating uncertainty to allow for the probabilistic observations.  So those are things not in the Schroedinger equation.

Self-locating uncertainty is not dependent on any particular theory. It’s the same whether it’s the Many Worlds, the Star Trek teleporter or God that does the duplicating.

>> No quantum interpretation needs to derive the Schrodinger Equation nor does it need to be assumed because it can be experimentally verified to be true. And no quantum interpretation is inconsistent with observation, at least not so far.

>It can't be experimentally verified that the other world branches exist

But an astronomical number, or even an infinite number, of other world branches is not inconsistent with experiment or observation, and if you want to hypothesize about what's really going on at the deepest level of reality while making the fewest possible assumptions then Many Worlds is your best bet. At least it's the best bet anyone has come up with so far.
 
> and the Schrodinger equation cannot be verified except statistically by assuming the Born rule.  

I must insist yet again that the Born Rule is NOT assumed to be true nor is it required to be derived to be true because we can do far better than either one of those two things. We can observe the Born Rule to be true.

No, you can't observe the Born rule to be true any more (or less) than you can observe Schroedinger's equation to be true.  They are theories that predict a result in every time and place, past and future.  If they fail, even on a set of measure zero, in this infinitude they are invalidated.  Every theory must go beyond what has been observed to be useful...that's the whole point of having theories instead of just catalogues of observations.

Brent


--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
--
Stathis Papaioannou

Brent Meeker

unread,
Dec 20, 2021, 8:05:00 PM12/20/21
to everyth...@googlegroups.com


On 12/20/2021 4:19 PM, Stathis Papaioannou wrote:


On Tue, 21 Dec 2021 at 09:56, Brent Meeker <meeke...@gmail.com> wrote:


On 12/20/2021 7:17 AM, John Clark wrote:
On Sun, Dec 19, 2021 at 10:38 PM Brent Meeker <meeke...@gmail.com> wrote:
> >> It also makes the assumption that the eigenvalues of a measurement are realized probabilistically.

>> What is the eigenvalue of a temperature of 72°F? It doesn't have one. A measurement doesn't have an eigenvalue but a matrix does, such as the one that describes the Schrodinger Wave. And no quantum interpretation needs to assume there is a relationship between the square of the absolute value of that wave and probability because it is observed to be true.

> A temperature operator, which would be matrix, might very well return 72degF as the eigenvalue of a state eigenvector. 


A temperature measurement taken at a particular time and place is not a temperature operator, and a measurement is not a probability, although the square of the absolute value of a wave function might tell you the probability of you getting that temperature measurement at that time and place.

>  Yes, it's empirically supported; So's the Schroedinger equation.  But it's part of the application of the Schroedinger equation.  It's not in the equation itself. 

I don't know what you mean by that. 

It's the projection postulate in the Copenhagen interpretation that applies the Born rule.  In MWI it's the Born rule plus some kind of self-locating uncertainty to allow for the probabilistic observations.  So those are things not in the Schroedinger equation.

Self-locating uncertainty is not dependent on any particular theory. It’s the same whether it’s the Many Worlds, the Star Trek teleporter or God that does the duplicating.

Not exactly.  In those other theories you mention you can put all the duplicates side by side and there's no sense to the question, which one happened?  They all did and there's no probability to assign to them; because probability only makes sense when something happens and other things don't.

Brent

Stathis Papaioannou

unread,
Dec 20, 2021, 9:13:59 PM12/20/21
to everyth...@googlegroups.com
The probabilities come from the fact that observers consider themselves unique individuals persisting through time.
--
Stathis Papaioannou

Brent Meeker

unread,
Dec 20, 2021, 11:55:42 PM12/20/21
to everyth...@googlegroups.com
But that doesn't imply any kind of probability unless they regard themselves as the one member of an ensemble that is unique, e.g. the one that really exists or the one that's really me.  Otherwise they are just like the duplicate Captain Kirks.

Brent

Stathis Papaioannou

unread,
Dec 21, 2021, 1:50:20 AM12/21/21
to everyth...@googlegroups.com
Each copy does indeed feel as if they are the one true continuation of the original even though they know that they are not, because that is the nature of first person experience.
--
Stathis Papaioannou

Bruce Kellett

unread,
Dec 21, 2021, 2:12:12 AM12/21/21
to Everything List
You still need to introduce an independent notion of probability because each member must consider himself to be a random selection from the ensemble. The notion of a random selection cannot be defined without reference to some prior notion of probability.

Bruce

Stathis Papaioannou

unread,
Dec 21, 2021, 2:51:44 AM12/21/21
to everyth...@googlegroups.com
Yes, but you don't need any specific theory about how your identity moves from one body into the next.


--
Stathis Papaioannou

Bruce Kellett

unread,
Dec 21, 2021, 3:35:57 AM12/21/21
to Everything List
You just need some credible evidence that such a notion even begins to make sense.

Bruce

Stathis Papaioannou

unread,
Dec 21, 2021, 3:50:05 AM12/21/21
to everyth...@googlegroups.com
It makes sense that I feel myself to be a unique individual persisting through time, because everyone understands what it means. Some people try to come up with theories based on this feeling, such as the existence of an immaterial soul, but that doesn’t follow. My feeling that I am a unique individual persisting through time stands independently of whatever entity or gives rise to this feeling.
--
Stathis Papaioannou

Bruce Kellett

unread,
Dec 21, 2021, 4:29:22 AM12/21/21
to Everything List
I don't know where you think you are going with this. Continuation of  personal identity through time was not what we were talking about. Persistence through time does not involve self-locating uncertainty from an ensemble at a point in time.

Bruce

Stathis Papaioannou

unread,
Dec 21, 2021, 5:31:08 AM12/21/21
to everyth...@googlegroups.com
If one version of me will see the atom decay and the other version of me will not see the atom decay, there is a 1/2 chance that I will see the atom decay, because of the symmetry of the situation and because I feel myself to be a unique individual persisting through time, even though I might know the objective details of what is occurring.


--
Stathis Papaioannou

Bruce Kellett

unread,
Dec 21, 2021, 5:51:14 AM12/21/21
to Everything List
I don't see how persistence through time has any bearing on the probability. If there is a split, then the probability that you will see one or the other result depends on the magnitudes of the wave function for the branches. That is the Born rule, and it is an independent assumption, as is the fact that there is a probability interpretation at all. Self-locating uncertainty only gives you a measure of the probability if the number of branches with each outcome matches the Born probabilities.

Bruce

Stathis Papaioannou

unread,
Dec 21, 2021, 6:36:07 AM12/21/21
to everyth...@googlegroups.com
Under MWI every outcome happens, so the probability of each outcome is 1. How do you justify calculating probabilities for outcomes that are less than 1?
--
Stathis Papaioannou

smitra

unread,
Dec 21, 2021, 7:34:41 AM12/21/21
to everyth...@googlegroups.com
Yes, I agree, I've read about such theories some time ago. I don't
remember the details, but I think they do predict new physics, so they
can be tested and falsified.

The notion of probability in Everettian theory is indeed problematic,
but you'll have similar problems in a Bruno-type copying experiment
where I make 99 copies of you that will observe one thing and 1 copy
will observe something else. Then all 100 observations will exist, but
you can still say that you'll expect to observe the first outcome with
99% probability.

Saibal

smitra

unread,
Dec 21, 2021, 8:20:45 AM12/21/21
to everyth...@googlegroups.com
One needs to define the observer + measurement apparatus. In
conventional QM one associates to a measurement device made ready to
measure some property of a system a Hermitian operator. But if we pursue
the MWI rigorously, then one should associate a set of commuting
operators to a conscious experience of an observer. And this should then
follow from a description of consciousness as the computational state of
some algorithm.

So, at this very moment my experience while I'm typing these words must
be the result of a particular algorithm that is processing a particular
set of data. I cannot in principle see the difference of that exact
algorithm being implemented in different parts of the multiverse. At
least not at that very moment when that particular set of data is
processed.

An algorithm can be described by a time evolution operator U that acts
on an input state and maps it to the state one computational step later.
The preferred basis then arises from selecting the sectors where the
algorithm representing the observer exists. The algorithm will have some
valid input states |input> and corresponding output states |output(in)>
= U|input>, we can then write:

U = sum over all |input>, |output> of |output><output|U|input><input| =
sum over all |input> of |output(input)><input|



Suppose that we seek the observer represented by the operator U who has
some definite experience. Whatever the observer is experiencing will be
some course grained description of the precise inputs the actual data
that U could be processing. So, we then consider limiting the summation
above to only those input states that correspond to the specified coarse
grained description. So, one can say that we need to narrow down U to
some definite experience, but we then still end up summing over a very
large set of states, because a particular conscious experience does not
correspond to a definite physical state.

A simple example is to consider simulating a spin measurement using a
quantum computer. The spin is then represent by a qbuit and the
"observer" measuring the spin would then be the CNOT operator that takes
the qubit representing the spin as the control qubit while the other
qubit that it acts on is initialized to be |0>. So, one can then say
that there exists a "CNOT observer". This definition is then well
defined w.r.t. changing the basis.

Saibal
> https://groups.google.com/d/msgid/everything-list/CAPCWU3JOTt6Msg%3DXiGTpoEK8TNq3mqqWtSi-s9xgA4OPpQGRjw%40mail.gmail.com
> [1].
>
>
> Links:
> ------
> [1]
> https://groups.google.com/d/msgid/everything-list/CAPCWU3JOTt6Msg%3DXiGTpoEK8TNq3mqqWtSi-s9xgA4OPpQGRjw%40mail.gmail.com?utm_medium=email&utm_source=footer

smitra

unread,
Dec 21, 2021, 8:27:34 AM12/21/21
to everyth...@googlegroups.com
I agree that this is a problem. But as as I explained just now to Jesse
Mazer, one should be able to make progress by including an observer
defined as an algorithm. This should amount to the same thing as is done
by Everett, but it's then motivated by the actual physics.

Saibal

Brent Meeker

unread,
Dec 21, 2021, 3:40:45 PM12/21/21
to everyth...@googlegroups.com
Yes, but symmetric examples are deceptive.  What do you fell when the probability of decay is 1/100 and non-decay 99/100.

Brent

Brent Meeker

unread,
Dec 21, 2021, 4:04:39 PM12/21/21
to everyth...@googlegroups.com
That would be better than just hand waving.  But can a macroscopic
observer really be approximated by a simple algorithm?  One of the
things that makes an "observer" is that it interacts with an environment
and has an effectively infinite degrees of freedom.

Brent

smitra

unread,
Dec 22, 2021, 5:39:23 AM12/22/21
to everyth...@googlegroups.com
What we called an "observer moment" (OM) in this list, should be
considered to be an algorithm that the universe is running using some
fraction of the physical degrees of freedom located in he brain and
perhaps also some other body parts. So, while the algorithm is not going
to be located at exactly one point, it's not infinite in extent either.
The physical state of the brain will, of course, be entangled with the
environment. But there exists a local description of how the brain
works, and there are good reasons foe believing that whatever we
experience (including the notion we have about our personal identity) is
due to running a particular algorithm.

Saibal














John Clark

unread,
Dec 22, 2021, 6:02:08 AM12/22/21
to 'Brent Meeker' via Everything List
On Tue, Dec 21, 2021 at 1:50 AM Stathis Papaioannou <stat...@gmail.com> wrote:

Each copy does indeed feel as if they are the one true continuation of the original even though they know that they are not,

They are a true continuation of the original, they're just not the "one" true continuation of the original.

John K Clark    See what's on my new list at  Extropolis
889


John Clark

unread,
Dec 22, 2021, 6:33:35 AM12/22/21
to 'Brent Meeker' via Everything List
On Tue, Dec 21, 2021 at 2:12 AM Bruce Kellett <bhkel...@gmail.com> wrote:

> You still need to introduce an independent notion of probability because each member must consider himself to be a random selection from the ensemble. The notion of a random selection cannot be defined without reference to some prior notion of probability.

Even if there are an infinite, and not just an astronomically large, number of other worlds it would still not be difficult to introduce the idea of probability. Each "you" in the Multiverse will live in a world that is different, and thus each "you" will be different, sometimes that difference will be very slight and sometimes it will be very large. The determination of how different "you" can be and still be considered "you" is arbitrary, but as long as consistency is maintained in your choice and the amount of difference allowed is greater than zero then there will always be more versions of "you" near the center of the Bell Curve than at the outer edges. So if I have a lottery ticket I can predict that tomorrow when the drawing is held I am far more likely to find myself in the losers center of the Bell Curve than at the millionaire trailing edge, although a small minority of "yous" will beat the odds and become rich

And I should add that if you demand perfection with zero slop then you'd have to conclude that you become a different person every time you take a sip of coffee, or even water.     

   John K Clark    See what's on my new list at  Extropolis
dpx

 

John Clark

unread,
Dec 22, 2021, 6:47:38 AM12/22/21
to 'Brent Meeker' via Everything List
On Tue, Dec 21, 2021 at 4:04 PM Brent Meeker <meeke...@gmail.com> wrote:

 > But can a macroscopic observer really be approximated by a simple algorithm?
 
Certainly not, if the algorithm was simple we would have found it by now, but a macroscopic observer can be created by a complex algorithm, after all we know that Darwinian Evolution managed to do it, and I have no doubt that in the near future intelligent humans will be able to do it too. After all, if random mutation and natural selection can do something then intelligent design can do it too, and do it faster and better.

John K Clark    See what's on my new list at  Extropolis
cde


Brent Meeker

unread,
Dec 22, 2021, 1:40:29 PM12/22/21
to everyth...@googlegroups.com


On 12/22/2021 3:32 AM, John Clark wrote:
On Tue, Dec 21, 2021 at 2:12 AM Bruce Kellett <bhkel...@gmail.com> wrote:

> You still need to introduce an independent notion of probability because each member must consider himself to be a random selection from the ensemble. The notion of a random selection cannot be defined without reference to some prior notion of probability.

Even if there are an infinite, and not just an astronomically large, number of other worlds it would still not be difficult to introduce the idea of probability.

An actual countable infinity creates problems like Hilbert's Hotel.


Each "you" in the Multiverse will live in a world that is different, and thus each "you" will be different, sometimes that difference will be very slight and sometimes it will be very large. The determination of how different "you" can be and still be considered "you" is arbitrary, but as long as consistency is maintained in your choice and the amount of difference allowed is greater than zero then there will always be more versions of "you" near the center of the Bell Curve than at the outer edges. So if I have a lottery ticket I can predict that tomorrow when the drawing is held I am far more likely to find myself in the losers center of the Bell Curve than at the millionaire trailing edge, although a small minority of "yous" will beat the odds and become rich

And I should add that if you demand perfection with zero slop then you'd have to conclude that you become a different person every time you take a sip of coffee, or even water. 

Hmm.  When I brought this up earlier you thought it was no problem that the same you would be living in innumerably many worlds simply because cosmic rays and radioactive decays were leaving macroscopic records that split the world.  So innumerably many yous shouldn't be a problem either.  But why not look at it the other way around and define "world" with some slop?  There's an essay by Vaidman that takes this approach http://philsci-archive.pitt.edu/19979/1/PreprintWFR.pdf

Brent

Brent

John Clark

unread,
Dec 22, 2021, 3:39:39 PM12/22/21
to 'Brent Meeker' via Everything List
On Wed, Dec 22, 2021 at 1:40 PM Brent Meeker <meeke...@gmail.com> wrote:

>> Even if there are an infinite, and not just an astronomically large, number of other worlds it would still not be difficult to introduce the idea of probability.
 
> An actual countable infinity creates problems like Hilbert's Hotel.

Hilbert's Hotel is strange but it creates no logical paradoxes, and there's nothing even strange about dividing up a continuum into a finite number of finite pieces, you do it every time you cut a piece of string.   

>> Each "you" in the Multiverse will live in a world that is different, and thus each "you" will be different, sometimes that difference will be very slight and sometimes it will be very large. The determination of how different "you" can be and still be considered "you" is arbitrary, but as long as consistency is maintained in your choice and the amount of difference allowed is greater than zero then there will always be more versions of "you" near the center of the Bell Curve than at the outer edges. So if I have a lottery ticket I can predict that tomorrow when the drawing is held I am far more likely to find myself in the losers center of the Bell Curve than at the millionaire trailing edge, although a small minority of "yous" will beat the odds and become richAnd I should add that if you demand perfection with zero slop then you'd have to conclude that you become a different person every time you take a sip of coffee, or even water. 

> Hmm.  When I brought this up earlier you thought it was no problem that the same you would be living in innumerably many worlds simply because cosmic rays and radioactive decays were leaving macroscopic records that split the world.  So innumerably many yous shouldn't be a problem either.  But why not look at it the other way around and define "world" with some slop? 

Call it whatever you like, but I don't see your point.

John K Clark    See what's on my new list at  Extropolis
hhh

 


 
Reply all
Reply to author
Forward
0 new messages