How does MWI resolve low probability multiverse events?

[tomkalafut]

Is there some term for what I'm about to describe? I don't know if it's a philosophical or probability discussion or what. Take sports leagues, for example. So am I to believe that every permutation of MLB team records occur in that many number of parallel universes? I mean, what are the chances that half the league goes 162-0 and the other half goes 0-162? Yet, if MWI is fact, this supposedly happens in some parallel universe somewhere every year. I'm not sure what I'm asking. Just wondering how MWI addresses low probability multiverse events.

Not a quantum physicist. Just someone who likes to watch cosmos related shows on THC, Discovery, etc. Sorry if this doesn't belong here. If not, please let me know where I should've posted it, like a "Physics for Dummies" forum. Thanks.

From wiki: The many-worlds interpretation is an interpretation of quantum mechanics.

It is also known as MWI, the relative state formulation, theory of the universal wavefunction, parallel universes, many-universes interpretation or just many worlds.

Many-worlds asserts the objective reality of the wavefunction, but denies the reality of wavefunction collapse. The subjective appearance of wavefunction collapse is explained by the mechanism of quantum decoherence. By this means many-worlds claims to resolve all of the correlation paradoxes of quantum theory, such as the EPR paradox[1][2] and Schrödinger's cat[3], since every possible outcome to every event defines or exists in its own "history" or "world". In layman's terms, there is a very large — perhaps infinite[4] — number of universes, and everything that could possibly have happened in our past (but didn't) has occurred in the past of some other universe or universes.

 

[Dmitry67]

Yes, no matter how low the probability of a 'branch' is, those who are in that branch feel it 'real'. For example, generate a 100-digit number with a true random generator (which uses some qunatum effects, noise etc). Look at that number. So you are on one of 10^100 branches, and your current branch has a probability of 10^-100 (*)

There is nothing strange with it, but the really hard questions in MWI are 'how Born rule is derived in MWI' and even deeper one 'what is a meaning of the probability in MWI'? I don't know the answers. (there are some claims that born rule can be derived in MWI from pure QM, but some people on this forum don;t argee with it, and also it is not clear what is a probability at all in MWI. MWI is deterministic from birds view, so how the 'probability' can be applicable to it?)

(*) The same, BTW, is true in non-MWI (single-histories) theories if Universe is true infinite.

 

[vanesch]

Yes, what is said here, is correct. The difficulty of MWI (of which I'm, btw, a proponent) is to "derive" the probabilities. I personally think it can't be done without some extra postulates or structures and I wrote a paper about that, but I couldn't get it published, not because it was "wrong" but apparently because it was "already known" (?).

The idea I used was that you can define other rules than the Born rule to give you a probability rule, while keeping the unitary part of quantum theory. That is, you could define a toy theory, a la Copenhagen, which is the same unitary quantum theory as we now have, but with a different Born rule (so not |psi|^2, but something else), and this probability rule is in no contradiction with all the other measurement postulates. It is just numerically different.
Of course it is experimentally wrong, no doubt about that. But it is logically compatible with the unitary part of QM. Now, as MWI only takes over the unitary part of QM, my point was that you can't hope to derive the Born rule that way, without "something else". That something else can be very small, such as the assumption of non-contextuality. But you need something that is not purely included in the strictly unitary part of QM...

 

[Fra]

I'm no MWI proponent but I have a different radical view. In my thinking, "probability measures" are subjective and related to "actions" of the observer in the sense of a rationa player that bases his actions upon the available information.

So in my view, "probability zero", doesn't mean represent what can not happen, it only represents from the subjective point of view, what is not EXPECTED to happen. This really makes also the event space subjective.

Since the probability is subjective, the only detecable "meaning" of zero probability is that the system (considered to be the observer) has actions AS IF he was convinced that the zero probability event couldn't happen.

So I think of the construction of probability measures as going hand in hand with construction of action measures.

Of course, one might call this an illusion, but this illusion can be stable or unstable, and the destabilisation process is a physical process. And ideally the lifetime of such an illusion could be predictable if you consider the revision process like a physical reuilibration process.

I don't it makes sense to consider infinite event spaces, because no physical system/observer can I think encode that. Here the limiting information capacity does enter.

If limited information capacity systems are interacting, then for sure one would expect interactions, that are long these lines. Each systems can "evaluate" a probability which effectively encodes also it's actions. That two system disagree on a probability, would not be a contradiction, it would rather imply a prediction of an interaction, and the interaction should be partly predictable from the form of the "inconsistency" and some extra assumptions of how a system handles new information.

All these are currently speculations but it would for sure resolve the question of meaning of low probability, or even zero probability - subjectivity or (intrinsic relativity) of the probabilities and the probabilities spaces, means that two people can disagree of probabilities without problems. It is not a problem, instead it predicts a specific interaction. If that interaction proves to be anything like standard model, then that would be a success. If not, then it's wrong.

/Fredrik

 

[Dmitry67]

So in my view, "probability zero", doesn't mean represent what can not happen, it only represents from the subjective point of view, what is not EXPECTED to happen. This really makes also the event space subjective.
/Fredrik

Yes, but what's about what has already happened?

We know that lifetime of neutron is about 1000s because we made experiments and all these neutrons showed that decay rate

There is another, highly unprobable branch of the universe, where all neutrons (participated in all experiments) were very long-lived: 10000s. No matter how unprobable it is, the scientists in that universe feel themselfs as 'real' as we do (but puzzled by the incosnistency of nature)

So yes, usually we witness the most probable outcomes. May be it is just an inllusion, because all outcomes are being witnesssed, but in that case is there any predictive value in such science? If I throw a dice 6'000'000 times, how many '6' I get? Following the logic above, we can't say, because all othcomes (0,1, ... 6000000) are real, all are observed, and all observers feel it real.

But what should we do with the obvious fact that usually we see the most probable outcomes?

 

[NiH]

Hello everyone!

Many-worlds asserts the objective reality of the wavefunction, but denies the reality of wavefunction collapse. The subjective appearance of wavefunction collapse is explained by the mechanism of quantum decoherence. By this means many-worlds claims to resolve all of the correlation paradoxes of quantum theory, such as the EPR paradox[1][2] and Schrödinger's cat[3], since every possible outcome to every event defines or exists in its own "history" or "world". In layman's terms, there is a very large — perhaps infinite[4] — number of universes, and everything that could possibly have happened in our past (but didn't) has occurred in the past of some other universe or universes.[/i]

If you are interested in stuff like this, take a look at publications by Seth Lloyd. Here is a preprint about the computational capacity of the universe: http://arxiv.org/abs/quant-ph/0110141
He draws the conclusion, that the universe cannot have performed more than 10^120 operations on 10^90 bits.
So, if one bit operation causes two new universes, you would have the incredible high amount of 2^(10^120) universes.

It is very speculative, but nevertheless quite interesting.

 

[Naty1]

Everything that is not prohibited is required.

 

[Fra]

Yes, but what's about what has already happened?

I suspect your point is that in a very old universe, everything sort of has happened, then how do I justify probability zero?

I have response to that in the spirit of my view, but I'll respond more in detail later.

The view I'm advocating is more like a "many observers", rather than "many universes".

/Fredrik

 

[Dmitry67]

Wait, I was talking about very low probability, not probability=0
Again, if there are 2 outcomes:

Dead cat: dead if neutron decays in 1 second (0,1%)
Alive cat: otherwise (99.9%)

As both cats exist and there are both versions of you, one sad and one happy, then why adjustment of the treshold (say, if we inverse it making 99.9% probability of cat gets killed) makes the difference?

 

[Fra]

Wait, I was talking about very low probability, not probability=0
Again, if there are 2 outcomes:

Dead cat: dead if neutron decays in 1 second (0,1%)
Alive cat: otherwise (99.9%)

As both cats exist and there are both versions of you, one sad and one happy, then why adjustment of the treshold (say, if we inverse it making 99.9% probability of cat gets killed) makes the difference?

I'm no longer sure I got your main objection.

I was going to respond that for a given observer, of limiting information capacity, a very low probability is indistinguishable from zero.

Each observer sees a "different history" of the world, and a bounded observer can not retain a complete hypothetical time-history. This effectively means (in my view that is) that history is eventually erased from the obserers memory. This justifies zero probability even for an very old universe, simply because the probability is formulated relative to a bounded inside observer.

Thus, the retained history of the universe, is not unique, because each observer retains a biased and truncated "memory record".

Other than that I'm not sure I understood your objection. Suppse the probability of an even is low, very low, and it still happens. So what? There is no problem with that is it?

I think you pondered how come we usually do observe the most probable events? In my view that is a result of evolution. If we consistently keep observing improbable events, our information measures are updating slowly so that eventually our expectations are in tune with our environment.

??

/Fredrik

 

[Dmitry67]

I think the issue does not appear difficult if we talk about a single fork-even with about 50/50% probability, especially in the future - when we discuss expectations.

But it becomes difficult to explain if we look at series of past events with near-100%/near 0% ratio.

So let me reformulate my question. Why in the history we don't see many tracks of extremely rare events? For example, any atom can decay with a very very low probability. Why we don't see big amounts of matter disintegrating?

Or say I look at 1'000'000 neutrons. How many neutrons do you expect to see in 1000 seconds? (We know the answer, but WHY?)

 

[tomkalafut]

Everything that is not prohibited is required.

So, like isomorphisms, I just have to accept it, even though I can't wrap my pea brain around it. :shy:

 

[Fra]

So let me reformulate my question. Why in the history we don't see many tracks of extremely rare events? For example, any atom can decay with a very very low probability. Why we don't see big amounts of matter disintegrating?

Here is a question with the purposes of clarifying, I'm still not sure I understand but I think I'm getting closer.

Are you asking, why we don't see frequent events that have low probability?

How about that if we did, then we would promptly revise our probability distribution? To keep consume data contradicting current measures, without revising the latter would simply not be rational behaviour.

So I think, "frequent observation of unlikely events" are possible, but it would be transient. It is the observer that keeps observing frequent improbable events that would disintegrate.

Does that make sense at all, or did I misinterpret your issue?

/Fredrik

 

[Dmitry67]

How about that if we did, then we would promptly revise our probability distribution?

If was consistent, yes
But what if in some universes neutron does not decay on wednesdays?

But let me try to make my question as short as possible

What is a difference between having outcomes
A (99.999%) and B (0.001%)
and A (0.001%) and B (99.999%) ?
as both outcomes are ALWAYS observed and BOTH are real?

 

[Fra]

If was consistent, yes
But what if in some universes neutron does not decay on wednesdays?

Maybe it's something general about how we reason that makes me not get it.

What is inconsistent about revising the probability distribution in the light of new evidence? And inconsistent with what? Note that in my view there are no timelss birds views.

What is a difference between having outcomes
A (99.999%) and B (0.001%)
and A (0.001%) and B (99.999%) ?
as both outcomes are ALWAYS observed and BOTH are real?

IMO, the two cases correspond to different observers first of all.

This fact does not make them mutually inconsistent IMO, it means there expected to be an interaction between them.

OTOH, if you are considering them to be in different universes, never in communication then your entire contradiction is non-physical and would never be physically observed as such by a real observer.

I'm still not sure I understand your issue.

/Fredrk

 

[Dmitry67]

yes, sorry, it is very difficult to explain, because I don't understand it too.

Let me make yet another attempt.
Say there are 2 outcomes for some repeatable experiment: f(requent) and R(are). The probability of these outcomes is established based on different experiments in other laboratories. (*)

So if I repeat this experiment I get something like

blackblackblackblackfffRblackblackblackfffRblackRblackblackblackblackf

So I ask myself: if I make 1000 experiments, I have 2^1000 outcomes, including all R. However, I expect to see most of ‘f’s, and yes, this is true. I am happy. Why does it happen?

I believe in MWI, but if we take its logic literally, then there is no difference between all f’s and all R’s: just 2 different branches. So in this pure form MWI does not have any predictive value, because everything can happen.

So the contradiction I see is that
1. in MWI different branches have different probabilities, or different weight
2. but the perception of reality and of consciousness does not depend on it, no matter how improbable the branch is we feel it real
3. so when we talk about the impressions of an observer, we should ignore probabilities
4. but it is not in agreement with our experience and does not have any predictive power.

 

[Fra]

I'll point out that I don't adhere to MWI - so I don't see as my task to efend it, I will just reflect over your issues from my point of view as part of a general discussion.

(I do not see the predictive value in pondering multiple non-communicating universes either)

Let me make yet another attempt.
Say there are 2 outcomes for some repeatable experiment: f(requent) and R(are). The probability of these outcomes is established based on different experiments in other laboratories. (*)

So if I repeat this experiment I get something like

blackblackblackblackfffRblackblackblackfffRblackRblackblackblackblackf

So I ask myself: if I make 1000 experiments, I have 2^1000 outcomes, including all R. However, I expect to see most of ‘f’s, and yes, this is true. I am happy. Why does it happen?

As I see it, you are questioning one of your own premises. You say it's established that the probability is such and such, and then ask why your future observations fit this?

If you are really asking why the future observations should necessarily fit the expectation acquire from the past (this is a valid question, since this is after all an expectation only, not a certain prediction) - then IMHO, this is not certain! However, a key point in my view is that in despite of this, the only rational choice is to act as per the information at hand.

The predicive value lies in assuming that to the extent possible, each observer/system has an action that is somewhat in line with a "rational players". The point is not that the expectation must match the future, the point is that there is no more rational choice than to act according to the given information, even if it later proves to be wrong.

Now, if you put this idea in an evolutionary context, a general prediction would be as the parts of the universe is equilibrated, their expectations become tuned. So, in an evolutionary context, it is not a conincidence that your observed frequences match the expected one acquired from past experiences.

Given again I'm not MWI advocate, if I were to project my thinking onto the MWI stuff, there are interactions between the branches. But of course, I do not think in terms of branches of the universes, I simply think in terms of differnt observers. The different "universes" are simply the different "inside views" of the one single universe. And the interaction takes place between the "host systems" of the "images of the apparently different universes"

Two such observers, that are in communication must interact in a very special way - the way which is consistent with the rational player assumption.

I'm simplifying my view here, since the "rational player" assumption is mroe complex, the rational player doesn't yield a deterministic action, only a constraint on the possible actions.

/Fredrik

 

[Fra]

Again i'll try to project this onto my own thinking and comment from there, meaning I will interpret the MWI branches are different observers.

So the contradiction I see is that
1. in MWI different branches have different probabilities, or different weight
2. but the perception of reality and of consciousness does not depend on it, no matter how improbable the branch is we feel it real
3. so when we talk about the impressions of an observer, we should ignore probabilities
4. but it is not in agreement with our experience and does not have any predictive power.

My idea of predictive power lies in the connection between probabilities over states, and transitions probabilities (ie actions).

In my view cosnciousness has nothing to do with this. To me, the observer is a physical system and the systems "knowledge about it's environment" is encoded in it's microstructure.

The fact that two observers expect different probabilities, means their actions will reflect this - ie. the prediction lies again at the rational player assumption that you can infere from how someone acts, what they know. Or from how someTHING acts, what it's state of information about it's environment is.

A possible prediction could be that, the reaction from the environment on an action that is "inconsistent" with what actually happens, implies physical forces on the system, that serves to "revise it's state", and sometimes even revise it's state space! (ie it's microstructure, not just microstate withing agiven microstructure)

Somehow this is the basis of howto infere physical interactions from a few assumption of all action systems of matter beeing in a loose sense "rational players". And that when systems becomes smaller, the possible interactions are highly constrained (unification).

This is how my view, will come with predictions, and with it not just a reinterpretation but also a reFORMULATION of QM. The idea taken seriously, is not just an interpretation since it makes predictions of the actions of matter. Except of course it's not a well developed program, and I'm still pondering many problems. But from a basic conceptual point of view, the problems I think you refer to looks like solvable.

/Fredrik

 

[Fra]

1. in MWI different branches have different probabilities, or different weight

A suspect construction IMO.

What is the physical process from where a real observer can infer/compute these probabilities?

It's exactly objections like that, that has lead me to my preferred view. If we just picture "probabilities" like that, without justification, then there is a birds view involved. This makes absolutely no sense to me at least.

So my alternative is, not multiple universes, but multiple observers, and instead the problem of "computing the probabilities of branches" becomes that of finding the population of the one universe. The population of observers, and thus different views, does effectrively represent these weights you're looking for I think.

In my view, the population of the universe, is a result of evolution. There is diversity from uncertainty, and selection from equilibration. To predict the population of the universe, also becomes the same as to find the properties of matter. Ie. what are the properties of the matter that does populate our universe?

I don't have a final answer, but the reasoning outline suggests a research program towards one. To be able to constrain the evolution into something managable and not get lost in infinite landscapes like other programs, the trick is to use the complexity scaling of the observer. When the complexity of the habitant systems is low, the laws of physics and the complexity of interactions are similarly simple.

/Fredrik

 

[kote]

The many worlds interpretation is based on 2 pretty poor premises in my opinion. The first premise for MWI is that any formula we can come up with to predict the outcome of experiments must represent a basic element of reality. The assumption is that instrumentalism in science is false and predictive formulas represent reality.

The second premise for MWI comes from scientists giving up. "We can't seem to do better than QM right now, and QM is based on probabilities, so nature must be random." It seems to me FAR easier to give up the absolute basic accuracy of QM as the be-all-end-all theory than it does to give up holding on to a deterministic (as opposed to random) nature/reality.

If you accept that not only will we never have a better theory than QM, but a basic randomness in nature itself is the reason that we can't get any better, and if you accept that unfalsified scientific theories necessarily represent basic reality, then you can get to MWI.

That's a lot to accept. I think MWI is pretty ridiculous for the above reasons.

To put it another way... is it more likely that QM is incomplete and there are underlying deterministic laws that we don't know yet? That's the conclusion everyone in history with a probabilistic theory has come to. Or is it more likely that QM must necessarily represent basic reality in its simplest form, that nature is random, and that there are infinite multiple universes constantly being created? Only someone extremely tied to QM as a theory (and many professional physicists are) would ever suggest the possibility of the absurd second option.

 

[yossell]

@Kote,

I agree that 'any formula we can come up with to predict the outcome of experiments must represent a basic element of reality' is naive - but why do you think MWI people make this assumption? I understood MWI to be motivated by the Measurement Problem, the desire to make sense of a QM that avoided the collapse postulate, and discontent with say Copenhagen style interpretations of QM.

I'm not sure what conclusion you tihnk everyone in history with a probabilistic theory has come to - that QM is incomplete and there are underlying deterministic laws? But a number of people have argued the opposite - it's the hidden variable theories, such as Bohm theory, which the majority regard with suspicion. Einstein's 'God doesn't play dice' has hardly been embraced by all. For my own part, though I think there are genuine problems with interpreting QM in its present form, its indeterminism is not a problem. I don't see why it shouldn't turn out that the underlying laws are probabilistic: it's an empirical matter, a genuine way the world might be - insisting that it's always ignorance of underlying deterministic laws looks like an anti-empirical unjustified assumption to me.

yossell

 

[kote]

@Kote,

I agree that 'any formula we can come up with to predict the outcome of experiments must represent a basic element of reality' is naive - but why do you think MWI people make this assumption? I understood MWI to be motivated by the Measurement Problem, the desire to make sense of a QM that avoided the collapse postulate, and discontent with say Copenhagen style interpretations of QM.

I'm not sure what conclusion you tihnk everyone in history with a probabilistic theory has come to - that QM is incomplete and there are underlying deterministic laws? But a number of people have argued the opposite - it's the hidden variable theories, such as Bohm theory, which the majority regard with suspicion. Einstein's 'God doesn't play dice' has hardly been embraced by all. For my own part, though I think there are genuine problems with interpreting QM in its present form, its indeterminism is not a problem. I don't see why it shouldn't turn out that the underlying laws are probabilistic: it's an empirical matter, a genuine way the world might be - insisting that it's always ignorance of underlying deterministic laws looks like an anti-empirical unjustified assumption to me.

yossell

Ok, yossell, you got me :). I should have said that every realist scientist in history has automatically accepted the incompleteness of probabilistic theories (until now). Physics, historically, has presumed deterministic realism - this is the entire classical paradigm. For an instrumentalist, questioning the completeness of any theory is meaningless because the claim is that it is inherently impossible to know anything about the underlying basic reality anyways. You know if your equations fit experiments, and that's the end of it.

The indeterminism does become a problem in a realist framework. And only if QM is accepted as an absolutely correct theory deserving of a realist interpretation within that framework. What does it even mean to have an indeterminate (random) but objectively real nature? How can something be real but not have a definite position, mass, etc? Multiple universes are one possibility that comes out of these premises. Multiple universes, however, do more in my mind just to disprove the premises (QM is complete and/or theories are real).

The details of the measurement problem etc that are the intermediate steps in getting to this conclusion aren't really relevant to the basic premises or eventual conclusion.

Overall, I agree with you that it's an empirical matter. We have no empirical evidence of multiple universes. Problem solved right there, premise 1.

 

[Dmitry67]

The second premise for MWI comes from scientists giving up. "We can't seem to do better than QM right now, and QM is based on probabilities, so nature must be random." It seems to me FAR easier to give up the absolute basic accuracy of QM as the be-all-end-all theory than it does to give up holding on to a deterministic (as opposed to random) nature/reality.

If you accept that not only will we never have a better theory than QM, but a basic randomness in nature itself is the reason that we can't get any better, and if you accept that unfalsified scientific theories necessarily represent basic reality, then you get to MWI.

Wait, wait.
MWI denies randomness: it is deterministic (in the birds view)

 

[yossell]

Hi Kote,

I can see why one might prefer instrumentalism to MWI. And I think you're right that, historically, before QM the paradigm seemed to be deterministic realism. It wasn't until QM, a theory that many have found difficult to interpret realistically, that indeterminism was embraced.

But if QM is problematic for the realist, then I do not think this is *because* of its indeterminism. Here, by 'indeterminism', I just mean the idea that, given the laws, given the conditions, there's more than one way for the world to evolve. This arises in QM because the fundamental laws involve probability.

For instance, I think that even Newtonian mechanics isn't strictly deterministic
(see Earman 'A Primer on Determinism' - though the scenarios where determinism fails are rather artificial) and there are questions about the determinism of GR near singularities. But neither of these theories seems anti-realist. And it seems coherent to me that a theory have probabilistic laws, and that this is entirely objective and realist matter. In short, I don't see why God shouldn't play dice (after all - he's God! he can do what he likes. That's what's so good about being God :).

Of course, if that were the only non-classical feature of QM, I don't think anyone today would be worrying about it. Its 'Indeterminacy' understood as implying that quantities somehow lack a determinate value until the right measuring device comes along, and giving rise to apparently odd things like S's cat, wave particle duality - plus the apparently unavoidable use of the notion of measurement in formulating the laws are harder for the realist to deal with.

 

[kote]

Wait, wait.
MWI denies randomness: it is deterministic (in the birds view)

I'll agree that in a way MWI is deterministic. However, the path taken by the world we live in (which is the only world we can have any evidence of) is randomly chosen from all possible paths. Randomness is still just as prevalent in MWI, its place in the theory has just been shifted slightly, albeit without any empirical or practical result.

In MWI our universe is just as random and arbitrary as in any other interpretation of QM.

 

[Dmitry67]

I'll agree that in a way MWI is deterministic. However, the path taken by the world we live in (which is the only world we can have any evidence of) is randomly chosen from all possible paths.

Absolutely not.
We always follow all possible paths.

This is a key: we don't 'chose' at some point what path to take. Let's say there are 2 outcomes: A and B. After an experiment there are 2 copies of an observer, each saying:
"But why the outcome is A, not B? It has been randomly chosen!"
and another one
"But why the outcome is B, not A? It has been randomly chosen!"

As we can remember events in the past, we perceive history as LINE while in MWI it is more like a TREE. But again, randomness in MWI is just a frog's view illusion.

 

[kote]

But if QM is problematic for the realist, then I do not think this is *because* of its indeterminism. Here, by 'indeterminism', I just mean the idea that, given the laws, given the conditions, there's more than one way for the world to evolve. This arises in QM because the fundamental laws involve probability.

Yossel,

When you get down to the logical consequences, I think there is really a lot to the idea that there's more than one way for the world to evolve.

There is a causal deterministic quality to the classical objects and properties of waves and particles. For something to have a mass, it is implied that it will behave a certain determined way in the presence of a gravitational field. An object's mass is defined by its behavior in gravitational fields (and collisions etc). Similarly, for something to have a spin, it is implied that it will behave a certain determined way in the presence of a magnetic field.

Denying that this determined causal chain exists is to deny the persistent existence of classical properties in basic reality. If a massive object may not necessarily behave in a determined way in the presence of a gravitational field, then mass must not be a basic property of matter.

So what's left for reality in an indeterminate realist theory? Well... we have no idea. But we know whatever it is, the properties of location, mass, velocity, polarization, spin, etc, are all emergent from it and are not "real" (or intrinsic) themselves.

The indeterminacy you mention with regard to classical properties is a logical consequence of QM denying the determined causal implications of these properties in the first place. The reason that momentum only has a determined value during interactions (measurements) is that momentum cannot, by definition, be a basic persistent property of basic quantum matter.

As a side note, by replacing the word "measurement" with the word "interaction," I think the whole measurement problem falls apart. Why shouldn't measurable properties be features of the interactions between substances as opposed to being intrinsic and persistent in the substances themselves? Being red is not an intrinsic property of an apple, but it only exists in the interaction between the apple and my eyes and brain, as mediated by underlying physical forces. In the same way, why shouldn't momentum only exist in the interaction between two colliding objects, as mediated by underlying quantum forces?

No experiment ever performed has measured anything except the result of interactions. This is not a new feature of QM.

P.S. Wow, intelligent conversation on the internet! I appreciate not getting the pointless response of "have you even studied QM?" like some of the "experts" have given on other threads.

 

[knoppix]

I have put some thought into this and have come to the conclusion that there is no conflict between probability theory and MWI.

MWI branches do not have 'weights'. However, the probabilities we experience are in fact combined probabilities from millions of quantum states. That means that if we roll a dice 6 million times we will get around 1 million 6's in almost all of the branches. The combined probability results in that most branches experience the most probable outcomes most of the time. It is true that a single branch will end up with 6 million 6's but it is extremely unlikely that we will end up in such a branch - in fact as unlikely as if we were to experience the event in a non-MWI universe.

That is why we often experience the most probable outcome.

All of the above is just my opinion, of course.

Now a question - does MWI predict infinite variations or is it a finite number of branches?

 

[Dmitry67]

However, the probabilities we experience are in fact combined probabilities from millions of quantum states.

Yes, consciousness has less distinct states than the quantum system (brain) where it resides, hence different quantum states are mapped into the same observers (consciousness) state.

Still I don't think it is a whole story and an 'explanation'.