You can do a really simple thought experiment that provides a persuasive argument in favor of the many worlds theory. Or at least, it shows that wave function collapse is relative to an observer, not universal everywhere.

Start with Schroedingers cat: the cat is in a box, it’s in a state of quantum superposition. It’s 50/50 alive or dead. Now a researcher opens the box: the cat is found to be either alive, or dead, and according to the Copenhagen interpretation we say that the wave function has collapsed.

But, now take the cat and the researcher, and put the whole first experiment inside another box. Now the researcher inside the second (outer) box opens up the box with the cat. From outside the second box, we still must say that the first researcher is in a state of superposition, but when the second box is opened, the first researcher will say, oh, I’ve known the cat was dead for 10 minutes or whatever.

So, before we open the second box, has the wave function collapsed, or not? It depends on which researcher you are. But Every part of the universe is in some sense “in a box” until information passes to it from another part of the universe. I give you: many worlds.

This is also known as the Wigner's Friend paradox.

Your explanation is not the standard understanding of Many Worlds today, however. The party line is that world splitting happens when decoherence has become "sufficiently irreversible" (for some arbitrary definition of "sufficiently", since there's no evidence of genuine irreversibility anywhere in physics), which in almost all real-world setups happens almost immediately.

Is there actually a "split"? I've always imagined it more like parallel strands that were always distinct but had overlaps. Like in the cat case, it's undetermined whether I'm on the cat dead or cat alive time-line because both are "overlapping" in the part of space-time where my brain is at the time of not knowing the cat's state. Then when I make a measurement, the two strands no longer overlap.

With Wigners friend there are also these two branches all along, but they overlap where the friend hasn't yet looked in the box, they then go apart in the part where the friend makes an observation, overlap where Wigner hasn't looked at his friend yet, but then they are distinct again where Wigner has looked at the friend's state. Similarly to how particles and light take all possible trajectories at the same time (path integral formulation), we could think that in the space-time places where the branches overlap, all compatible pasts are equally valid, at other space-time locations the particular strand extends back through the past to just one of the cat states, so the state seems collapsed, but the stand also led through the overlap area, so you remember the time where you didn't yet have the "which branch you're on" information.

I'm not sure I've digested your whole comment, but the basic idea is this:

A "split" only means that the respective branches of the wave function cannot interfere with each other any longer. Then they will evolve completely separately from each other, and can be considered distinct. The only way you can get branches to interfere is if you have precise control over all particles involved in the entanglement, which is practically impossible for any macroscopic system.

Long before you entangle with the cat system, the two branches have decohered. While technically there is still an enormous superposition there, for all practical purposes it is indistinguishable from being a classical system, with classically indeterminate branches.

I should say that I'm not a fan of this explanation, possibly related to your points. Regardless of "practical purposes," QM does not allow me to say that there is a well-defined answer before it entangles with me (and that there is one afterward). I'm much more comfortable using the word "split" only there. But this skirts too close to solipsism for some, since "my world" (which I also know as "the world") only comes into being upon contact with me.

The answer to this which makes the most sense to me is that the wave function collapses when the information exists that determines the quantum state. In Schrodinger's cat, this would be when the Geiger counter detects the quantum state, but by the time the cat dies/does not die and the researcher opens the box, everything is already in the classical world.

Schrodinger's cat was made to be a thought experiment to show how absurd some of the ideas about the quantum world are. We are unable to prove whether the cat does go into a superposition or does not. Personally, the ideas that superposition is tied to whether consciousness observes it doesn't make sense to me. After reading about some of the stranger experiments such as the quantum eraser experiment, I've gravitated towards information theory being the missing link. If that is true, then the astounding conclusion would be that the fabric of our universe is information!

The problem with many worlds is why do I think I'm in a particular timeline? If many worlds is correct, there is no "I", there are many rapidly branching "I"s, which intuitively doesn't make sense. Of course, each "I" will have the perception of being the sole "I", but the overall view that there is no singular "I" is very counter to my self conception. I prefer to preserve intuitions over conjectures, so I count that as a strike against the many worlds interpretation.

So, it seems like both the Copenhagen and the many worlds interpretations of quantum physics are wrong, given a preference for non arbitrary reference point and preserving intuitions. Is there a reason that those two must be the only possible interpretations?

It's definitely a mindfuck but I'm not sure if you can dismiss MWI based on that.

Everything that was intuitive was already discovered. That's why each new theory is harder to visualise than the previous one: newtonian physics, electromagnetism, relativity, quantum physics, etc.

But perhaps those are valid constraints. Given that all of science is built from fundamental subjective perceptions, it seems inconsistent to extrapolate scientific theories that deny fundamental subjective perceptions.

There is a singular you, the one that is your time-line. There are also different versions of that in other timelines. But as you don't expect to feel the consciousness of the guy spatially next to you, the other parallel universe has also no influence on your conscious feeling of self.

Observation causes collapse, so perhaps because the effect it has is negligible to the system of the universe there's simply no point 'rendering'/wave collapsing. Perhaps there is some other collapse mechanism to explain life before conscious observers, or perhaps quantum behavior emerged as time went on after the big bang, or perhaps there exists some kind of proto-conciousness.

I guess my question is if the conscious observers are themselves part of the universe, what caused their waveform to collapse? And what is the special 'observation' part of their material properties such that it can cause collapse? Why don't other material configurations cause collapse?

I'm inclined more towards a dualistic explanation than a material one. Perhaps a better word to describe these concious and proto-concious entities would be 'collapsing agents', 'collapsants' or something of the sort.

I'd speculate that they exist as some unobservable component of reality, considering we see no hard evidence for it with our existing tools. I'm no expert on string theory, but as I understand it predicts a large number of dimensions, if we think of these as arrangements of data in memory, perhaps this information could be encoded there?

The wavefunction is the thing that is in a state of quantum superposition.

It all depends on whether you think the wavefunction describes the macroscopic world (MWI), or the wavefunction merely calculates amplitudes&probabilities for things that happen in the macroscopic, classical world (Copenhagen)

Is the electron it's wavefunction (MWI), or is the electron a particle/field that is influenced by the wavefunction (Copenhagen)?

Haven't we tested precisly this? The fact that we observe interference patterns shows that the wavefunctions are real.

As far as I can tell, Copenhagen posits that there is a scale at which wave functions stop being real, and instead collapse down to a single "macroscopic" state. Convieniently, this scale happens to be around the scale where our experiments stop being able to test for it in practice.

Put another way, Copenhagen posits that the macroscopic world is something more then simply emergent behavior.

> The fact that we observe interference patterns shows that the wavefunctions are real.

well, not quite - remember that what we actually observe for a given particular event is something that looks like particle that is localized in space.

If you're referring to the double slit experiment, the interference pattern emerges as a property of the whole collection of individual photon strikes. Yes, you need to use the wavefunction to calculate where the photon is likely to land, but each individual photon strike appears to us on the target as a single localized photon.

Whether wavefunctions are "real" is a tricky question, but just the presence in interference patterns doesn't get us to answer that question yet.

Non-physicist perspective, if I’m out of line, please say so.

It seems silly to expect that the wavefunctuon is “real” just like it is silly to assume that the gravitational field is “real” or inches are “real”. They are mathematical constructions we use to make predictions about what we see. Just because the predictions are good doesn’t make the mathematical formalism real.

As a contrived example, imagine an advanced AI program running on Linux. The AI develops all sorts of models and makes very accurate predictions about what will happen in user space when it calls fopen or pipes data from /dev/urandom. But no matter how well it models the output of /dev/urandom, it will never know that the reason why the data seems random is because the kernel is generating it from a camera pointed at a lava lamp. Even if the AI was smart enough to invent models for the lava lamp and the camera optics, etc, this model wouldn’t be real, it’s still some bits in memory /describing/ what is going on outside of the AIs immediate experience.

If we put the sensor in the slits themselves, what we would observe is individual photons passing through one slit or the other. What the double slit experiment shows is that there is still something of a wavefunction (whatever that actually is) at the moment the photons pass the slit, because that is how we get the interference pattern.

As before when we observe the photons, we see them as single discreet particles. However, in principle this seems to be the same situation as when we detect photons passing through a slit.

Right, but in that case we never observe the wavefunction, the quantum state of the particle, or the particle interacting with itself.

In that sense, the wavefunction predicts the outcome but it does not have to "real". It is real in Pilot Wave theory, but it could simply be a mathematical or calculational tool .

In either case, you are right that one has to use it get the right result.

Let me ask you a similar question. Are virtual particles, as described with Feynman diagrams, real?

I think virtual particles are a different question, as it is at least clear what is being asked (I think they are real, but have seen no convincing arguement one way or the other).

With wavefunctions, I am not actually clear on what the object "wavefunction" is. If you are asking if I believe the wave exists independent of the particle, then no. But I also do not think it is merely a calculation tool.

My view is that it is analogous to "velocity" in classical mechanics. In that it is a property of the object in question. I do not know if that qualifies as "real" or not.

No, in a particle model there could be no interference pattern. That is the whole point. You can do the experiment - and it has been done - with a single photon and you still get interference.

Maybe so, but this sentence: "interference pattern emerges as a property of the whole collection of individual photon strikes."

Is incorrect and seems to miss the whole point of the experiment. We don't need a "whole collection" to get interference, we can see it with one photon. If that doesn't mean the wave function is real, how else can it be explained?

There is both the "interference pattern" on the target wall made by individual photons over time, which we can directly observe, and the interference of the photon's wavefunction with itself, which we cannot directly observe. I was referring to the former. We know the photon interferes with itself, via the wavefunction as you describe, but this interaction is entirely unobservable and invisible to us, and in that sense, we don't understand the ontological status of the wavefunction and what it's physical status is.

Sorry, what is it we cannot observe in the latter? We see the pattern on the screen. Do you mean we cannot observe the interference that leads to that pattern? That is quite right since the only way to see a photon is to bounce it off of something.

But surely the first researcher is also in a state of quantum superposition and they are 50/50 opening/not opening the box.

So how does time come into the equation? And does that mean that any observation is localised and there's no need for many worlds except in unobserved areas of the universe?

> But surely the first researcher is also in a state of quantum superposition and they are 50/50 opening/not opening the box.

In this simple thought experiment we're assuming they're not. Imagine that the first researcher is definitely going to open the inner box after 10 minutes and then the second researcher is definitely going to open the outer box after 20 minutes.

> And does that mean that any observation is localised and there's no need for many worlds except in unobserved areas of the universe?

Well the whole problem with non-many-worlds is that it has this idea that observed areas and unobserved areas are different. Which can work in very simple thought experiments but gets crazy as soon as you try to apply it to the real world: any interaction between two particles can be seen as an "observation", every part of the universe is gradually "observing" every other part, how do you draw a line between which parts are in superposition and which parts aren't? Many worlds simplifies all this and says that everyone is always in a superposition, there's never a "collapse", and observing outcome A or outcome B with 50% probability is just what being in a superposition of A or B feels like from the inside.

Started reading this and realized it was Sean... he gave a great talk on this last night at Symphony Space on the UWS. Believe he's doing a full book tour - worth checking out.

I'm not familiar with that book, but I can say that Dr. Carroll's book is far better than any previous book on the subject I've read, in terms of getting rid of the confusion inherent in QM.

That characterization of Woit does not appear quite accurate. Woit's claim may even be bolder, in that he claims that there is no functional difference between MWI and Copenhagen.

> What strikes me when thinking about these two supposedly very different points of view on quantum mechanics is that I’m having trouble seeing why they are actually any different at all.

The main difference is one of rhetoric: Copenhagen insists that experiments have single outcomes. MWI insists that experiments have multiple outcomes in point of metaphysical fact.

Another difference is that the MWI has to explain the Born rule (which is problematic), whereas Copenhagen simply takes it as an axiom.

> The main difference is that the MWI has trouble explaining the Born rule.

To re-phrase, the claim in that article is that once MWI has done away with the Born rule, it has to deal with how the probabilities that we find in experiment are calculated under MWI.

.. Which is not obvious at first glance, because MWI seems to suggest everything happens with probability = 1.

Carroll acknowledges it is one of the challenges against MWI that needs to be answered.

Carroll offers "self-location uncertainty" as a way to recover the probabilities predicted by the Born Rule despite the determinism of the Schrodinger Equation

> But even if both people know the wave function of the universe, there is now something they don’t know: which branch of the wave function they are on. There will inevitably be a period of time after branching occurs but before the observers find out what outcome was obtained on their branch. They don’t know where they are in the wave function. That’s self-locating uncertainty, as first emphasized in the quantum context by the physicist Lev Vaidman. (https://news.ycombinator.com/item?id=20927410)

* edit - OP's post was edited but i'm leaving this post as-is.

> MWI insists that experiments have multiple outcomes in point of metaphysical fact.

That doesn't seem right. under MWI there is only one Wafefunction and one big quantum state. There are not additional 'worlds'

For the record, the original just said (as quoted in the above response) "The main difference is that the MWI has trouble explaining the Born rule." Upon reflection I decided that wasn't really the main difference, that the main difference is the difference in ontology. But the fact that the MWI has to recover the Born rule from the linear dynamics is a major consequence of this difference in ontology.

The jury is still out about whether or not self-location uncertainty actually allows the Born rule to be recovered without begging the question. If it works, it would be a major breakthrough. Personally, I'll give long odds against.

[UPDATE]:

> There are not additional 'worlds'

Um, what do you think that the "MW" in "MWI" stands for?

The label "Many Worlds Interpretation" is a popular label for the theory, but it's not a particularly accurate one.

The original paper by Hugh Everett was titled '"Relative State" Formulation of Quantum Mechanics'.

The underlying principle of Many Worlds isn't actually that there are many worlds. Is that there is a singular wave function, which is the true reality. The things we refer to as "worlds" are massively-complicated entanglements and superpositions within that wave function, of which we—for some reason—can only observe one position of that superposition.

Everything you say is true. Nonetheless, if you read papers written by self-identified advocates of Everett, they invariably refer to it as "many-worlds" and they talk about "the branching structure of the multiverse" and "copies" or "versions" of yourself. So regardless of what you or I think Everett meant, the present-day advocates of his theory are clearly trying to advance the idea that there are, in point of metaphysical fact, many worlds.

Sure, but I would think every one of those advocates would also agree with what I wrote.

That the "world" described by many worlds is a convenient short-hand for a single position in a massively entangled system of particles in a complex-superposition.

It's the "Many Worlds" is common moniker, and the rhetoric does use the concept of worlds all the time, but whenever you get down to the specifics of the theory there really aren't many worlds.

I think it's unfair to hold the shorthand rhetoric against people when they attempt to get to that more precise level. It would be like using the plain english definition of "observation" against someone trying to provide a more rigorous definition of what an "observation" is under the Copenhagen interpretation.

Ultimatley, this discussion all stems from the original statement: "MWI insists that experiments have multiple outcomes in point of metaphysical fact." It's a fairly moot point at this point since that statement was removed.

> "MWI insists that experiments have multiple outcomes in point of metaphysical fact." It's a fairly moot point at this point since that statement was removed.

I'll stand by that with one minor modification: contemporary advocates of MWI insist that experiments have multiple outcomes in point of metaphysical fact.

Sorry for the slow response. Been a busy couple of days :)

> No, they wouldn't. They don't. See my response to lmm here

I don't think lmm and I disagree.

I said "[The underlying principle of Many Worlds] is that there is a singular wave function, which is the true reality". lmm says: "The only thing that MWI demands we take as metaphysically real is the wavefunction".

I think this is expressing the same thing in different terms.

I said: "The things we refer to as "worlds" are massively-complicated entanglements and superpositions within that wave function". lmm says: "Given a wavefunction like 1/sqrt2(|experimenter who observed a live cat> + |experimenter who observed a dead cat>) ...". Again, I think our descriptions are in fairly close agreement.

I think lmm's description is a good summary of how the "worlds" in Many Worlds are viewed by Many Worlds advocates: "in the same way that people who hear a particular pattern of sound might find it convenient to think of it as a chord of two or three notes". The "worlds" aren't actually fundamental to the theory, but rather an emergent phenomena from them.

The reason for the label "Many Worlds" is not because they are the fundamental concept. It's because it's the most... interesting element from a pop-sci marketing angle.

If you agree on that math/science then this point you're adamant about whether it is or is not a "world" reduces to a semantics/ontology/philosophy question, which we're not going to get further clarity on in a thread about trying to gain a proper or better understanding as laypeople of the scientific theories, unless you want to advance the conversation to the understanding of those scientific theories into a semantic/ontological/philosophic realm.

> "worlds" are massively-complicated entanglements and superpositions within that wave function, of which we—for some reason—can only observe one position of that superposition.

Is the collapse of the wave function related to consciousness? How would we construct a test to verify that?

> Um, what do you think that the "MW" in "MWI" stands for?

Ok yes, it is named that, but what the theory postulates is really just that "when Schrodinger opens the box to check on his cat, he himself is in a superposition just like the cat was -- and what it feels like to be in a superposition is that you don't notice you're in a superposition, because your superposed states each only observe one cat."

> your superposed states each only observe one cat

Again, I have to ask: what do you think "many worlds" means? Because what you've just described is exactly what the advocates of the MWI describe when they talk about "many worlds", which they absolutely do. The MWI literature is festooned with references to "copies" or "versions" of yourself, and "the branching structure of the multiverse".

> The MWI literature is festooned with references to "copies" or "versions" of yourself, and "the branching structure of the multiverse".

Many MWI advocates will describe that physical wavefunction in those terms, but that's a metaphor/model/aid to understanding, not a "point of metaphysical fact". The only thing that MWI demands we take as metaphysically real is the wavefunction. Given a wavefunction like 1/sqrt2(|experimenter who observed a live cat> + |experimenter who observed a dead cat>) many people find it convenient to think of this as "two Everett branches with ...", in the same way that people who hear a particular pattern of sound might find it convenient to think of it as a chord of two or three notes. But if you prefer to think of it as simply a unitary (in the everyday sense) wavefunction that's fine, you'll make exactly the same experimental predictions.

Here's a quote from David Deutsch's "The Fabric of Reality":

"When I introduced tangible and shadow photons I apparently distinguished them by saying that we can see the former, but not the latter. But who are ‘we’? While I was writing that, hosts of shadow Davids were writing it too. They too drew a distinction between tangible and shadow photons; but the photons they called ‘shadow’ include the ones I called ‘tangible’, and the photons they called ‘tangible’ are among those I called ‘shadow’.

Not only do none of the copies of an object have any privileged position in the explanation of shadows that I have just outlined, neither do they have a privileged position in the full mathematical explanation provided by quantum theory. I may feel subjectively that I am distinguished among the copies as the ‘tangible’ one, because I can directly perceive myself and not the others, but I must come to terms with the fact that all the others feel the same about themselves.

Many of those Davids are at this moment writing these very words. Some are putting it better. Others have gone for a cup of tea."

And his summary:

"the whole of physical reality, the multiverse, contains vast numbers of parallel universes"

So no, Deutsch, who is arguably the leading advocate of MWI, does not consider parallel universes to be a "metaphor/model/aid to understanding". He insists that they are real. He even defends this position at extraordinary length in chapter 10 of TFOR.

MWI provides a clear, comprehensive explanation for all of the “weird” quantum phenomenon. Beyond that, this quote from FoR [1] convinced me:

“To those who still cling to a single-universe world-view, I issue this challenge: explain how Shor’s algorithm works. I do not merely mean predict that it will work, which is merely a matter of solving a few uncontroversial equations. I mean provide an explanation. When Shor’s algorithm has factorized a number, using 10^500 or so times the computational resources than can be seen to be present, where was the number factorized? There are only about 10^80 atoms in the entire visible universe, an utterly minuscule number compared with 10^500. So if the visible universe were the extent of physical reality, physical reality would not even remotely contain the resources required to factorize such a large number. Who did factorize it, then? How, and where, was the computation performed?”

"Interpretations of quantum mechanics, unlike Gods, are not jealous, and thus it is safe to believe in more than one at the same time. So if the many-worlds interpretation makes it easier to think about the research you’re doing in April, and the Copenhagen interpretation makes it easier to think about the research you’re doing in June, the Copenhagen interpretation is not going to smite you for praying to the many-worlds interpretation. At least I hope it won’t, because otherwise I’m in big trouble."

If one die can have 6 values, how is it possible that two identical dice have 21 configurations? Where are those additional configurations coming from if there are only 12 possible values in the pair of dice?

This is a stupid argument, comparable to "if humans evolved from apes why are there still apes?"

An analogous question: when three objects are orbiting each other in space, where is the computation being performed that provides the solution to the three-body problem that represents their motions?

Well, obviously, the computation in the latter case is being "performed" by the three bodies themselves, and the Shor computation is being performed inside a quantum computer that exists in our universe. It's just that the stuff that is performing the computation isn't atoms, it's entangled qbits. There's no need to bring entire parallel universes into it.

I think Deutsch is doing a very good job of explaining exactly the concept I'm trying to explain.

But you still have to peel back the words to think about what kind of mechanism they are describing, which Deutsch is trying to communicate. You can't just complain that he says "reality contains parallel universes" without accepting that he's taken the time to map "parallel universes" to a well-defined concept.

I'm not complaining about anything. I'm simply observing that Deutsch and other prominent advocates of the MWI do not consider multiple worlds to be metaphorical. They insist in no uncertain terms that they are real, and go to great lengths to justify this position. There is absolutely no ambiguity with respect to what they believe.

Deutsch may believe that. (Heck, I believe it). It's a natural interpretation of the MWI wavefunction. But it is an interpretation, a layer on top of the physical claims, a model - not a mandatory part of MWI. Deutsch may interpret a given wavefunction as describing a particular ensemble of 'tangible' and 'shadow' objects, but someone who interprets that same wavefunction a different way (but agrees on the same wavefunction, and expects it to evolve according to the same laws of quantum mechanics) will still generate the same experimental predictions and, more to the point, still be doing MWI.

That makes MWI a "no true parallel Scotsman" interpretation.

Either the other versions of reality are tangible, or they're not. If they're just abstractions, you have't solved the problem - you've taken a long way around a circular argument that starts with the wavefunction and ends with it doing something that results in a measurement, maybe.

Unfortunately explaining how that happens is the problem you're trying to solve. It's not an answer to the problem.

If they are real, there are all kinds of other issues, the Born Rule being the most obvious.

> Either the other versions of reality are tangible, or they're not. If they're just abstractions, you have't solved the problem - you've taken a long way around a circular argument that starts with the wavefunction and ends with it doing something that results in a measurement, maybe.

The wavefunction is physically real. Whatever you interpret it as is your business. We've solved a real problem: we can explain the evolution of the wavefunction and the results of experiments without needing some weird collapse/decoherence concept that breaks all the usual rules. The Born Rule is still something that remains to be explained, agreed, but it's a much smaller something than the full Copenhagen style collapse.

There seems to be some confusion here, based around the usage of words.

You seem annoyed by the terminology of words like "worlds" or "multiverse", but are you also thinking about what a MWI advocate is referring to by the usage of those words? They are shorthand for "the observable wave function" and "the global wave function".

There being "copies" of Schrodinger observing his cat is the most intuitive and expressive way to summarize Schrodinger being in a linear superposition of two states in a (cat alive, cat dead) basis. It's not any different from how a Copenhagenist would treat a photon passing through a polarizating beamsplitter as being in a superposition of having passed through and having been reflected; if the photon then goes and interacts with other things, they would keep track of what would have happened in each "branch" of the superposition.

If you have an issue with those, why don't you have an issue with the idea of superposition itself? It's exactly the same concept.

(UPDATE)

In particular, when MWI people say "multiverse", it's not a sleight of hand where we are being deceptive when we actually mean "observable and unobservable components of the wave function". The term "multiverse" had no prior meaning in physics; MWI is defining it to mean the global wave function, using that word to draw attention to the fact that it has unobservable components.

These words are being given rigorous definitions that align with their intuitive meaning. You seem to be arguing against that practice?

Should I complain that when a Copenhagenist says "collapse", nothing is actually falling down and so it's misleading?

No it isn't. Copenhagenists insist that the wave function "collapses" upon "measurement" and so measurements only have one outcome in point of metaphyisical fact. Many-worlders insist that the wave function does not collapse, and hence measurements have many outcomes in point of metaphyiscal fact.

You may consider this a distinction without a difference, but the fact of the matter is that people do argue about this. A lot.

I agree that it's an important difference! And it matters a lot. I am just saying that the concept of "copies" is no more mysterious than the concept of "superposition", and if you can accept the latter then I don't see why you complain about the former.

As for collapse, Copenhagenists can't seem to explain why an interaction between a photon and electrons in a polarizing filter does not collapse the wave function, but an interaction between a photon and electrons in my retina does. They're made of the same stuff and should be following the same quantum rules; why does one of them perform a unitary transformation and the other a lossy projection?

The reason I can accept superposition is that I have first-hand access to experimental evidence of it. I can see interference fringes with my own eyes. By way of contrast, not only do I not have any direct evidence of the existence of copies of me, I cannot possibly have such evidence. And this is not a technological limitation. This is a limit predicted by the theory itself.

Your usage of "many outcomes" is what may be causing contention/confusion for those looking for answers in what is described by the mathematical equations, and scientific theories. If you want to justify "many outcomes" on metaphysical grounds, that is not a scientific conversation.

Confused. Consider a light beam with only one photon at a time (very dim) measured by a photoreceptor. Put a 2nd photoreceptor after the first - it detects no photons.

So detecting the first, definitely 'collapsed the wave function'. This is not an unknown result in physics.

It just applies the absolute square of the amplitudes to the chance of being in a given world rather than the chance of an event. The thing that needs explaining is why either probability is given by the square of the amplitudes.

Fair enough - I'm not a physicist or read papers on physics. I'm just trying to re-tell the best version of the theories as I've understood them by steel-manning arguments laid out (by scientists that have some media exposure), rather than try to straw-man each theory with prima-facie criticisms that have already been answered.

I know MWI has significant gaps, this appears to be one of the biggest.

Ta. As far as I can tell a lack of an explanation for the Born rule is a feature of all quantum interpretations, not just many worlds. It's pretty much the central mystery of quantum mechanics.

Conversely, Copenhagen-like explanations need to explain (or postulate) a decoherence/collapse operation that violates all the usual rules of QM: it would be non-time-symmetric, non-unitary, likely non-local (propagating faster than light), ...

Non-locality is practically axiomatic if you understand Bell's Inequalities, and doesn't require true FTL signalling.

MWI doesn't really solve either the locality or the FTL problem. If you're going to believe in MWI you have to explain how an entire universe appears instantaneously, identical in every way to another universe, except for one almost insignificant difference in exactly one quantum interaction.

That's a rather larger and more serious signalling and information transfer problem than the much smaller-scale non-locality required to explain entanglement.

> Non-locality is practically axiomatic if you understand Bell's Inequalities, and doesn't require true FTL signalling.

My complaint is exactly that the Copenhagen interpretation introduces that unnecessary FTL signalling requirement. Under the MWI, nothing weird is going on in the EPR thought experiment. But if you believe that a collapse occurs when we measure system A, then that collapse has to somehow occur non-locally and/or be FTL-transmitted between A and B.

> If you're going to believe in MWI you have to explain how an entire universe appears instantaneously, identical in every way to another universe, except for one almost insignificant difference in exactly one quantum interaction.

You don't; to talk about it in those terms is putting the cart before the horse. You have to believe that a wavefunction evolves in the usual way and represents the thing that it usually represents. You have to believe that when a pure system interacts with a system in superposition then the result is two systems that are entangled and in superposition - something you presumably already believe if you believe in any kind of QM at all. You only have to believe that a new universe appears instantaneously if your interpretation of putting a particle into superposition is that a new particle appears instantaneously, which is not an approach I've ever heard advocated.

I don't think MWI claims, exactly, that an entire universe 'appears instantaneously'.

Couldn't MWI represented with a probability cloud? Maybe the many-worlds exists as a fixed-volume probability on all states in the universe.

From this perspective, MWI would be similar to a multi-universal annealing system, where different quantum outcomes result in fractures within the probability cloud (as opposed to an expansion of the tree-graph).

Carroll postulates that all the necessary degrees of freedom are already just sitting there in hilbert space - it's only when the states become orthogonal to each other (e.g. "branching") that we no longer have access to them. https://www.youtube.com/watch?v=jHLfMXvQqX8

Actually most interpretations have the same explanations for the Born rule, it's just that people who study many worlds are more inclined to consider them. People who believe in Copenhagen don't care about this stuff.

Right, most of them just take it as an axiom. But the MWI explicitly rejects the Born rule as an axiom, insisting that the Schrödinger equation is all the actual physics there is.

MWI is not incompatible with postulating the Born rule if you want to. It just doesn't elevate it to an axiom of the theory. It's like doing constructive mathematics: you're not denying the law of excluded middle, you can postulate it and it will have the usual consequences, but the core theory doesn't rely on it, and it's possible to hope for a more elegant alternative.

Yes, you can just add the Born rule as an additional axiom, but that undermines the whole point of adopting the MWI in the first place, which is to remain mathematically "clean" with no ad-hoc assumptions.

Yes and no. I'd argue that the Born rule as a rule for subjective probabilities in a MWI context is much cleaner than a full Copenhagen-style collapse operation.

You are, of course, free to argue that. But most advocates of the MWI will disagree with you. That's why they put so much effort into trying to derive the Born rule.

Not that it really matters but I disagree that MWI advocates like it because of the born rule derivations. Actually most born rule derivations apply equally to all interpretations.

I’m with Vaidman on this, who is one of the best known supporters of MW, which is that the clearest solution is just to postulate the connection between psi magnitude and subjective experience. It’s an empirical discovery that works.

There's a significant gap between "cleaner than a Copenhagen-style collapse" and "clean enough to be a good part of our fundamental theory of nature" :).

He didn't say to re-add it as an axiom. He's pointing out that the "born rule" (or something that gives you probabilities in the way that you need to agree with experiment) can potentially "emerge" i.e. be recovered (if they can find the right proof for it), without having had to assume that up front.

It's like when GR can "recover" F=MA, despite not having assumed F=MA that as an axiom.

"Let me address another issue with many-worlds. It is a deterministic theory, even a hyper-deterministic theory, i.e. determinism applies to everything in the entire universe. Indeed, since there can’t be any influences coming from outside and since the Schroedinger equation - the only dynamical equation of the theory - is deterministic, everything that happens today, e.g. what I am writing, the way each reader reacts, the details of all solar eruptions, etc, was all encoded in some “quantum fluctuations” of the initial state of the universe. Given the complexity of the (many-) worlds, it had to be encoded in some infinitesimal digits of some quantum state, possibly in the billionths of billionths decimal place. I am always astonished that some people seriously believe in that."

This seems wrong to me? A lot of "things that happen" seem to be created by randomness generated along the way, from only observing subsystems.

I.e., suppose I create a random bit generator by shining photons on a half-silvered mirror and then measuring which way they go. I might then decide what to write depending what the bit is---if it's 0 I will write a HN comment, if it's 1 I will reply to an email.

In this case, the things that happen are determined deterministically, but it's not the case that there is some quantum fluctuation in the initial state of the universe which determines what I write. Rather, when I do the measurement of the bit, I create two branches, one with a HN comment and one with an email. Observers entangled with the first branch can only see the HN comment, but this complexity was created at the time of the measurement, not at the time of the big bang...

There is no "randomness generated along the way" in the MWI. Schroedinger's equation is deterministic and according to many-worlders is the only thing that "makes things happen". Everything that happens in every "world" was bound to happen since the beginning of the evolution of the "multiverse".

Right, but my point is that this doesn't mean that the initial state contains very much information.

Maybe a different example makes this clearer. Suppose you use a quantum random number generator to choose letters from A to Z, and write out a full book this way. If you consider the entire (multiple) world, it's basically Borges' Library of Babel: it has non-zero amplitudes for every possible book. So one possible "thing that can happen" is that Shakespeare's Hamlet gets written.

But that doesn't mean that the initial state of the universe somehow encodes the text of Hamlet. The initial state of the universe doesn't have any information at all, in quantum fluctuations or otherwise. You only see Hamlet emerge if you restrict your attention to some particular branches of the world. It was "bound to happen" that Hamlet would get written, but only because every book would get written.

The initial state contains as much information as the current state if the evolution is unitary (only Schroedinger’s equation matters according to the MWI, right?).

And surely the wavefunction of a system of 10^80 particles (or whatever) does contain some information about what happens... or do you think that “every book would get written” independently of the form of the universal wavefunction?

But the "current state" of the entire wavefunction doesn't necessarily contain very much information. You only get most of the complexity (solar flares, novels) if you look at subsystems of it.

I think the situation is exactly analogous to the Library of Babel in Borges' story. If someone sends you the entire library in the mail, they have given you zero bits of information. If they only send you a particular room of it you get more information, and narrowing it down to sending a particular book gives you more bits still.

It's the same with the multiverse. If we start out with the photon about to go through the half-silvered mirror, the state looks like |photon>⊗|me>, and then over time it evolves into

(1/√2)|photon went left>⊗|me observing left> + (1/√2)|photon went right>⊗|me observing right>

and the latter state contains no more information than the former one (you can run the time evolution backwards to recover the original state). However, if you restrict your attention and only consider one of the branches, e.g. |photon went right>⊗|me observing right>, then you have added (one more bit) of information. And most macroscopic processes can be explained "one branch at a time", since in practice decoherence means that there are no noticable interactions between branches. I think in the MWI this is the source of most complexity in the universe: not detailed information encoded in the initial state, and not objective collapse, but the bits of information you implicitly create if you only consider one branch of an experimental outcome.

I imagine the same thing happens in general: we start with some uniform soup (with ~zero information), in a bunch different MWI branches it self-gravitates slightly differently so we get a different set of galaxies and stars in each branch, then for each branch that has a planet, different quantum randomness may cause different intelligent life to develop, and eventually they may write different books. Of course, it's also possible that some small initial state gets magnified, as a separate mechanism, but that's the same in objective-collapse and MWI. The random bits created by objective collapse are the same same the ones you get by looking at particular branches in MWI.

> If we start out with the photon about to go through the half-silvered mirror, the state looks like |photon>⊗|me>,

But in the MWI what we start with is the universal wavefunction |Psi>. Say there are 10^80 particles in the universe and they may have a spin and could be in infinite locations. That makes for an ininite-dimensional Hilbert space. If we consider that the precision in the location variables is limited by Planck's length there are still around 10^180 possible locations in the universe for each particle. The Hilbert space where the quantum state of the universe is defined is huge.

Now, you say "the state looks like |photon>⊗|me>". To keep things simple and in the realm of Schroedinger's equation I will consider an electron instead of a photon, prepared in the state spin z=+1/2.

Is "|up>⊗|me>" shorthand for "(1 |up> + 0 |down>)⊗|me, the electron at that position, the rest of the universe>" which would be the "in-branch" quantum state of the universe?

If indeed |Psi> can be written as a superposition of that state and an inconceivable huge number of orthogonal states, are all those states (when their coefficient is non-zero) the multiple worlds?

But given |Psi> there are infinitely many (or at least a lot of) possible basis to express the pure state as a decomposition (including of course the one where the only non-zero "world" in the superposition is the one described by the wave function |Psi>). Do we count as "worlds" as well the states appearing in other basis?

Is this slicing of |Psi> into other states just a mathematical trick or does it have a physical meaning? Even if we can redo the operation at different times on the evolving |Psi(t)>, how are the "emergent substates" at different times related?

The answer to all these questions is far from trivial.

Indeed not, but is any of this relevant to the original quote, that deterministic dynamics means that the initial state of the universe must have lots of information hidden in the far-off decimals? I would hope that one could consider a simplified setting for that and then extrapolate back to the continuous case.

I agree the original quote doesn’t present a very strong argument against the MWI. In the same way that arguments against determinism in classical physics are not very convincing. But seemed worth pointing out.

In the MWI case it’s not clear how sensitive the “worlds” are to the in initial state. It’s not even clear how sensitive they are to the current state! Better understanding of how (and how many) worlds “emerge” would be needed to say if the “problem” is or not as bad as in the classical case.

Is the number of worlds stable? Is it increasing? In any case, those worlds are the inevitable consequence of the initial conditions even though we can’t even estimate how much fine-tuning is needed. I don’t think “randomness generated along the way” is a satisfactory counter-argument.

The benefit is that instead of two rules for quantum behavior 1) the Schrodinger equation 2) the born rule , that contradict each other, you just have 1) the Schrodinger equation, which you already accepted.

There is also a higher level goal that also motivates him

> the quest for quantum gravity is being held back by physicists’ traditional strategy of taking a classical theory (such as Albert Einstein’s general relativity) and ‘quantising’ it. Presumably nature doesn’t work like that; it’s just quantum from the start.

2. The benefit is that it's a very direct translation of the math into concepts. If you actually build the mathematical model, the "other worlds" are right there in the model, as the other half of a superposition.

It doesn't go away. You just can't get to it. Nor can you ignore it, because it's sitting there as a term in the equation, persisting as you evolve it over time. It just becomes smaller and smaller without ever becoming zero. That's why you observe a classical universe even though it's really quantum: the odds of the other parts of a superposition having an effect are on the order of Avogadro's-number-to-1 against.

Copenhagen works by just arbitrarily defining that number as zero, which is easy to work with but conceptually incomplete. MWI keeps it there in another "world", which is more in keeping with the math but unedifying because it sounds so unfalsifiable.

Hope that helps. (Which is always code for "... but it probably doesn't", I'm afraid.)

Quantum Immortality. Basically, if you keep splitting and you make an assumption like "My perception of the world is relative to me", then when a split causes you to die you don't really split - because you must ignore branches where you don't exist.

So then your view always follows the branching path where you live and you should expect to live 1M years from now. If there is only one world you shouldn't expect that.

As scld touched on, this "article" is really an excerpt from Sean Carroll's new book, Something Deeply Hidden: Quantum Worlds and the Emergence of Spacetime.

## 121 Comments:

You can do a really simple thought experiment that provides a persuasive argument in favor of the many worlds theory. Or at least, it shows that wave function collapse is relative to an observer, not universal everywhere.

Start with Schroedingers cat: the cat is in a box, it’s in a state of quantum superposition. It’s 50/50 alive or dead. Now a researcher opens the box: the cat is found to be either alive, or dead, and according to the Copenhagen interpretation we say that the wave function has collapsed.

But, now take the cat and the researcher, and put the whole first experiment inside another box. Now the researcher inside the second (outer) box opens up the box with the cat. From outside the second box, we still must say that the first researcher is in a state of superposition, but when the second box is opened, the first researcher will say, oh, I’ve known the cat was dead for 10 minutes or whatever.

So, before we open the second box, has the wave function collapsed, or not? It depends on which researcher you are. But Every part of the universe is in some sense “in a box” until information passes to it from another part of the universe. I give you: many worlds.

This is also known as the Wigner's Friend paradox.

Your explanation is not the standard understanding of Many Worlds today, however. The party line is that world splitting happens when decoherence has become "sufficiently irreversible" (for some arbitrary definition of "sufficiently", since there's no evidence of genuine irreversibility anywhere in physics), which in almost all real-world setups happens almost immediately.

Is there actually a "split"? I've always imagined it more like parallel strands that were always distinct but had overlaps. Like in the cat case, it's undetermined whether I'm on the cat dead or cat alive time-line because both are "overlapping" in the part of space-time where my brain is at the time of not knowing the cat's state. Then when I make a measurement, the two strands no longer overlap.

With Wigners friend there are also these two branches all along, but they overlap where the friend hasn't yet looked in the box, they then go apart in the part where the friend makes an observation, overlap where Wigner hasn't looked at his friend yet, but then they are distinct again where Wigner has looked at the friend's state. Similarly to how particles and light take all possible trajectories at the same time (path integral formulation), we could think that in the space-time places where the branches overlap, all compatible pasts are equally valid, at other space-time locations the particular strand extends back through the past to just one of the cat states, so the state seems collapsed, but the stand also led through the overlap area, so you remember the time where you didn't yet have the "which branch you're on" information.

I'm not sure I've digested your whole comment, but the basic idea is this:

A "split" only means that the respective branches of the wave function cannot interfere with each other any longer. Then they will evolve completely separately from each other, and can be considered distinct. The only way you can get branches to interfere is if you have precise control over all particles involved in the entanglement, which is practically impossible for any macroscopic system.

Long before you entangle with the cat system, the two branches have decohered. While technically there is still an enormous superposition there, for all practical purposes it is indistinguishable from being a classical system, with classically indeterminate branches.

I should say that I'm not a fan of this explanation, possibly related to your points. Regardless of "practical purposes," QM does

notallow me to say that there is a well-defined answer before it entangles with me (and that thereisone afterward). I'm much more comfortable using the word "split" only there. But this skirts too close to solipsism for some, since "my world" (which I also know as "the world") only comes into being upon contact with me.The answer to this which makes the most sense to me is that the wave function collapses when the information exists that determines the quantum state. In Schrodinger's cat, this would be when the Geiger counter detects the quantum state, but by the time the cat dies/does not die and the researcher opens the box, everything is already in the classical world.

Schrodinger's cat was made to be a thought experiment to show how absurd some of the ideas about the quantum world are. We are unable to prove whether the cat does go into a superposition or does not. Personally, the ideas that superposition is tied to whether consciousness observes it doesn't make sense to me. After reading about some of the stranger experiments such as the quantum eraser experiment, I've gravitated towards information theory being the missing link. If that is true, then the astounding conclusion would be that the fabric of our universe is information!

The problem with many worlds is why do I think I'm in a particular timeline? If many worlds is correct, there is no "I", there are many rapidly branching "I"s, which intuitively doesn't make sense. Of course, each "I" will have the perception of being the sole "I", but the overall view that there is no singular "I" is very counter to my self conception. I prefer to preserve intuitions over conjectures, so I count that as a strike against the many worlds interpretation.

So, it seems like both the Copenhagen and the many worlds interpretations of quantum physics are wrong, given a preference for non arbitrary reference point and preserving intuitions. Is there a reason that those two must be the only possible interpretations?

> but the overall view that there is no singular "I" is very counter to my self conception

And the idea that the Earth being round and rotating around the sun is also very counter to my intuition.

Think of it this way. Instead of asking "Does this feel correct to me?", ask, "If this was correct, would I feel any different?"

It's definitely a mindfuck but I'm not sure if you can dismiss MWI based on that.

Everything that was intuitive was already discovered. That's why each new theory is harder to visualise than the previous one: newtonian physics, electromagnetism, relativity, quantum physics, etc.

>"I" is very counter to my self conception

it'a because of human nature, ego and self-worth. It's hard to accept that "I" might have many siblings living in different branches.

But perhaps those are valid constraints. Given that all of science is built from fundamental subjective perceptions, it seems inconsistent to extrapolate scientific theories that deny fundamental subjective perceptions.

There is a singular you, the one that is your time-line. There are also different versions of that in other timelines. But as you don't expect to feel the consciousness of the guy spatially next to you, the other parallel universe has also no influence on your conscious feeling of self.

This whole idea of observers just brings to mind things like culling in graphics rendering, perhaps that's what's happening

What is doing the culling?

Depends who you ask, I'd say God, others might say the computer we exist in and others might say The Universe

But in a quantum universe everything is an uncollapsed wave. It doesn't seem anything can cause the collapse to happen.

Observation causes collapse, so perhaps because the effect it has is negligible to the system of the universe there's simply no point 'rendering'/wave collapsing. Perhaps there is some other collapse mechanism to explain life before conscious observers, or perhaps quantum behavior emerged as time went on after the big bang, or perhaps there exists some kind of proto-conciousness.

I guess my question is if the conscious observers are themselves part of the universe, what caused their waveform to collapse? And what is the special 'observation' part of their material properties such that it can cause collapse? Why don't other material configurations cause collapse?

I'm inclined more towards a dualistic explanation than a material one. Perhaps a better word to describe these concious and proto-concious entities would be 'collapsing agents', 'collapsants' or something of the sort.

Interesting idea, where would the dualist agents exist?

I'd speculate that they exist as some unobservable component of reality, considering we see no hard evidence for it with our existing tools. I'm no expert on string theory, but as I understand it predicts a large number of dimensions, if we think of these as arrangements of data in memory, perhaps this information could be encoded there?

YES

The wavefunction is the thing that is in a state of quantum superposition.

It all depends on whether you think the wavefunction describes the macroscopic world (MWI), or the wavefunction merely calculates amplitudes&probabilities for things that happen in the macroscopic, classical world (Copenhagen)

Is the electron it's wavefunction (MWI), or is the electron a particle/field that is influenced by the wavefunction (Copenhagen)?

Haven't we tested precisly this? The fact that we observe interference patterns shows that the wavefunctions are real.

As far as I can tell, Copenhagen posits that there is a scale at which wave functions stop being real, and instead collapse down to a single "macroscopic" state. Convieniently, this scale happens to be around the scale where our experiments stop being able to test for it in practice.

Put another way, Copenhagen posits that the macroscopic world is something more then simply emergent behavior.

> The fact that we observe interference patterns shows that the wavefunctions are real.

well, not quite - remember that what we actually observe for a given particular event is something that looks like particle that is localized in space.

If you're referring to the double slit experiment, the interference pattern emerges as a property of the whole collection of individual photon strikes. Yes, you need to use the wavefunction to calculate where the photon is likely to land, but each individual photon strike appears to us on the target as a single localized photon.

Whether wavefunctions are "real" is a tricky question, but just the presence in interference patterns doesn't get us to answer that question yet.

Non-physicist perspective, if I’m out of line, please say so.

It seems silly to expect that the wavefunctuon is “real” just like it is silly to assume that the gravitational field is “real” or inches are “real”. They are mathematical constructions we use to make predictions about what we see. Just because the predictions are good doesn’t make the mathematical formalism real.

As a contrived example, imagine an advanced AI program running on Linux. The AI develops all sorts of models and makes very accurate predictions about what will happen in user space when it calls fopen or pipes data from /dev/urandom. But no matter how well it models the output of /dev/urandom, it will never know that the reason

whythe data seems random is because the kernel is generating it from a camera pointed at a lava lamp. Even if the AIwassmart enough to invent models for the lava lamp and the camera optics, etc, this model wouldn’t be real, it’s still some bits in memory /describing/ what is going on outside of the AIs immediate experience.I think this is the view that scientists if they are interested more in the task of prediction than anything else https://www.youtube.com/watch?v=ka9KGqr5Wtw

If we put the sensor in the slits themselves, what we would observe is individual photons passing through one slit or the other. What the double slit experiment shows is that there is still something of a wavefunction (whatever that actually is) at the moment the photons pass the slit, because that is how we get the interference pattern.

As before when we observe the photons, we see them as single discreet particles. However, in principle this seems to be the same situation as when we detect photons passing through a slit.

Right, but in that case we never observe the wavefunction, the quantum state of the particle, or the particle interacting with itself.

In that sense, the wavefunction predicts the outcome but it does not have to "real". It is real in Pilot Wave theory, but it could simply be a mathematical or calculational tool .

In either case, you are right that one has to use it get the right result.

Let me ask you a similar question. Are virtual particles, as described with Feynman diagrams, real?

I think virtual particles are a different question, as it is at least clear what is being asked (I think they are real, but have seen no convincing arguement one way or the other).

With wavefunctions, I am not actually clear on what the object "wavefunction" is. If you are asking if I believe the wave exists independent of the particle, then no. But I also do not think it is merely a calculation tool.

My view is that it is analogous to "velocity" in classical mechanics. In that it is a property of the object in question. I do not know if that qualifies as "real" or not.

No, in a particle model there could be no interference pattern. That is the whole point. You can do the experiment - and it has been done - with a single photon and you still get interference.

You've misunderstood my point.

Maybe so, but this sentence: "interference pattern emerges as a property of the whole collection of individual photon strikes."

Is incorrect and seems to miss the whole point of the experiment. We don't need a "whole collection" to get interference, we can see it with one photon. If that doesn't mean the wave function is real, how else can it be explained?

There is both the "interference pattern" on the target wall made by individual photons over time, which we can directly observe, and the interference of the photon's wavefunction with itself, which we cannot directly observe. I was referring to the former. We know the photon interferes with itself, via the wavefunction as you describe, but this interaction is entirely unobservable and invisible to us, and in that sense, we don't understand the ontological status of the wavefunction and what it's physical status is.

Sorry, what is it we cannot observe in the latter? We see the pattern on the screen. Do you mean we cannot observe the interference that leads to that pattern? That is quite right since the only way to see a photon is to bounce it off of something.

But surely the first researcher is also in a state of quantum superposition and they are 50/50 opening/not opening the box.

So how does time come into the equation? And does that mean that any observation is localised and there's no need for many worlds except in unobserved areas of the universe?

> But surely the first researcher is also in a state of quantum superposition and they are 50/50 opening/not opening the box.

In this simple thought experiment we're assuming they're not. Imagine that the first researcher is definitely going to open the inner box after 10 minutes and then the second researcher is definitely going to open the outer box after 20 minutes.

> And does that mean that any observation is localised and there's no need for many worlds except in unobserved areas of the universe?

Well the whole problem with non-many-worlds is that it has this idea that observed areas and unobserved areas are different. Which can work in very simple thought experiments but gets crazy as soon as you try to apply it to the real world: any interaction between two particles can be seen as an "observation", every part of the universe is gradually "observing" every other part, how do you draw a line between which parts are in superposition and which parts aren't? Many worlds simplifies all this and says that everyone is always in a superposition, there's never a "collapse", and observing outcome A or outcome B with 50% probability is just what being in a superposition of A or B feels like from the inside.

If a tree falls in the woods, and no one is around to hear it, does it make a sound?

nope

Sean Carroll's new book on quantum physics that came out yesterday is the clearest laymen's exposition of quantum physics I have ever read: https://www.audible.com/pd/Something-Deeply-Hidden-Audiobook...

Started reading this and realized it was Sean... he gave a great talk on this last night at Symphony Space on the UWS. Believe he's doing a full book tour - worth checking out.

Another book I'm enjoying is "Einstein's Unfinished Revolution" [1] by Lee Smolin, which also has a deep dive into foundations of QM.

[1] https://www.amazon.com/Einsteins-Unfinished-Revolution-Searc...

I'm not familiar with that book, but I can say that Dr. Carroll's book is far better than any previous book on the subject I've read, in terms of getting rid of the confusion inherent in QM.

Halfway through his latest Mindscape podcast episode that drills down into this topic in a very approachable way.

Peter Woit in his blog "Not Even Wrong" [0] sometimes deals with the Many Worlds interpretation. He doesn't seem to be a fan.

He had a recent post [1] about Everett that's interesting.

[0] https://www.math.columbia.edu/~woit/wordpress/ [1] http://www.math.columbia.edu/~woit/wordpress/?p=11202 [2] https://www.math.columbia.edu/~woit/wordpress/?p=10522

That characterization of Woit does not appear quite accurate. Woit's claim may even be bolder, in that he claims that there is no functional difference between MWI and Copenhagen.

> What strikes me when thinking about these two supposedly very different points of view on quantum mechanics is that I’m having trouble seeing why they are actually any different at all.

The main difference is one of rhetoric: Copenhagen insists that experiments have single outcomes. MWI insists that experiments have multiple outcomes in point of metaphysical fact.

Another difference is that the MWI has to explain the Born rule (which is problematic), whereas Copenhagen simply takes it as an axiom.

http://blog.rongarret.info/2019/07/the-trouble-with-many-wor...

> The main difference is that the MWI has trouble explaining the Born rule.

To re-phrase, the claim in that article is that once MWI has done away with the Born rule, it has to deal with how the probabilities that we find in experiment are calculated under MWI.

.. Which is not obvious at first glance, because MWI seems to suggest everything happens with probability = 1.

Carroll acknowledges it is one of the challenges against MWI that needs to be answered.

Carroll offers "self-location uncertainty" as a way to recover the probabilities predicted by the Born Rule despite the determinism of the Schrodinger Equation

> But even if both people know the wave function of the universe, there is now something they don’t know: which branch of the wave function they are on. There will inevitably be a period of time after branching occurs but before the observers find out what outcome was obtained on their branch. They don’t know where they are in the wave function. That’s self-locating uncertainty, as first emphasized in the quantum context by the physicist Lev Vaidman. (https://news.ycombinator.com/item?id=20927410)

* edit - OP's post was edited but i'm leaving this post as-is.

> MWI insists that experiments have multiple outcomes in point of metaphysical fact.

That doesn't seem right. under MWI there is only one Wafefunction and one big quantum state. There are not additional 'worlds'

> OP's post was edited

For the record, the original just said (as quoted in the above response) "The main difference is that the MWI has trouble explaining the Born rule." Upon reflection I decided that wasn't really the

maindifference, that the main difference is the difference in ontology. But the fact that the MWI has to recover the Born rule from the linear dynamics is a major consequence of this difference in ontology.The jury is still out about whether or not self-location uncertainty actually allows the Born rule to be recovered without begging the question. If it works, it would be a major breakthrough. Personally, I'll give long odds against.

[UPDATE]:

> There are not additional 'worlds'

Um, what do you think that the "MW" in "MWI" stands for?

The label "Many Worlds Interpretation" is a popular label for the theory, but it's not a particularly accurate one.

The original paper by Hugh Everett was titled '"Relative State" Formulation of Quantum Mechanics'.

The underlying principle of Many Worlds isn't actually that there are many worlds. Is that there is a singular wave function, which is the true reality. The things we refer to as "worlds" are massively-complicated entanglements and superpositions within that wave function, of which we—for some reason—can only observe one position of that superposition.

Everything you say is true. Nonetheless, if you read papers written by self-identified advocates of Everett, they invariably refer to it as "many-worlds" and they talk about "the branching structure of the multiverse" and "copies" or "versions" of yourself. So regardless of what you or I think Everett meant, the present-day advocates of his theory are clearly trying to advance the idea that there are, in point of metaphysical fact, many worlds.

Sure, but I would think every one of those advocates would also agree with what I wrote.

That the "world" described by many worlds is a convenient short-hand for a single position in a massively entangled system of particles in a complex-superposition.

It's the "Many Worlds" is common moniker, and the rhetoric does use the concept of worlds all the time, but whenever you get down to the specifics of the theory there really aren't many worlds.

I think it's unfair to hold the shorthand rhetoric against people when they attempt to get to that more precise level. It would be like using the plain english definition of "observation" against someone trying to provide a more rigorous definition of what an "observation" is under the Copenhagen interpretation.

Ultimatley, this discussion all stems from the original statement: "MWI insists that experiments have multiple outcomes in point of metaphysical fact." It's a fairly moot point at this point since that statement was removed.

> I would think every one of those advocates would also agree with what I wrote.

No, they wouldn't. They don't. See my response to lmm here: https://news.ycombinator.com/item?id=20941803

> "MWI insists that experiments have multiple outcomes in point of metaphysical fact." It's a fairly moot point at this point since that statement was removed.

I'll stand by that with one minor modification: contemporary advocates of MWI insist that experiments have multiple outcomes in point of metaphysical fact.

Sorry for the slow response. Been a busy couple of days :)

> No, they wouldn't. They don't. See my response to lmm here

I don't think lmm and I disagree.

I said "[The underlying principle of Many Worlds] is that there is a singular wave function, which is the true reality". lmm says: "The only thing that MWI demands we take as metaphysically real is the wavefunction".

I think this is expressing the same thing in different terms.

I said: "The things we refer to as "worlds" are massively-complicated entanglements and superpositions within that wave function". lmm says: "Given a wavefunction like 1/sqrt2(|experimenter who observed a live cat> + |experimenter who observed a dead cat>) ...". Again, I think our descriptions are in fairly close agreement.

I think lmm's description is a good summary of how the "worlds" in Many Worlds are viewed by Many Worlds advocates: "in the same way that people who hear a particular pattern of sound might find it convenient to think of it as a chord of two or three notes". The "worlds" aren't actually fundamental to the theory, but rather an emergent phenomena from them.

The reason for the label "Many Worlds" is not because they are the fundamental concept. It's because it's the most... interesting element from a pop-sci marketing angle.

If you agree on that math/science then this point you're adamant about whether it is or is not a "world" reduces to a semantics/ontology/philosophy question, which we're not going to get further clarity on in a thread about trying to gain a proper or better understanding as laypeople of the scientific theories, unless you want to advance the conversation to the understanding of those scientific theories into a semantic/ontological/philosophic realm.

> "worlds" are massively-complicated entanglements and superpositions within that wave function, of which we—for some reason—can only observe one position of that superposition.

Is the collapse of the wave function related to consciousness? How would we construct a test to verify that?

> Um, what do you think that the "MW" in "MWI" stands for?

Ok yes, it is named that, but what the theory postulates is really just that "when Schrodinger opens the box to check on his cat, he himself is in a superposition just like the cat was -- and what it feels like to be in a superposition is that you don't notice you're in a superposition, because your superposed states each only observe one cat."

> your superposed states each only observe one cat

Again, I have to ask: what do you think "many worlds" means? Because what you've just described is exactly what the advocates of the MWI describe when they talk about "many worlds", which they absolutely do. The MWI literature is festooned with references to "copies" or "versions" of yourself, and "the branching structure of the multiverse".

> The MWI literature is festooned with references to "copies" or "versions" of yourself, and "the branching structure of the multiverse".

Many MWI advocates will describe that physical wavefunction in those terms, but that's a metaphor/model/aid to understanding, not a "point of metaphysical fact". The only thing that MWI demands we take as metaphysically real is the wavefunction. Given a wavefunction like 1/sqrt2(|experimenter who observed a live cat> + |experimenter who observed a dead cat>) many people find it convenient to think of this as "two Everett branches with ...", in the same way that people who hear a particular pattern of sound might find it convenient to think of it as a chord of two or three notes. But if you prefer to think of it as simply a unitary (in the everyday sense) wavefunction that's fine, you'll make exactly the same experimental predictions.

Here's a quote from David Deutsch's "The Fabric of Reality":

"When I introduced tangible and shadow photons I apparently distinguished them by saying that we can see the former, but not the latter. But who are ‘we’? While I was writing that, hosts of shadow Davids were writing it too. They too drew a distinction between tangible and shadow photons; but the photons they called ‘shadow’ include the ones I called ‘tangible’, and the photons they called ‘tangible’ are among those I called ‘shadow’.

Not only do none of the copies of an object have any privileged position in the explanation of shadows that I have just outlined, neither do they have a privileged position in the full mathematical explanation provided by quantum theory. I may feel subjectively that I am distinguished among the copies as the ‘tangible’ one, because I can directly perceive myself and not the others, but I must come to terms with the fact that all the others feel the same about themselves.

Many of those Davids are at this moment writing these very words. Some are putting it better. Others have gone for a cup of tea."

And his summary:

"the whole of physical reality, the multiverse, contains vast numbers of parallel universes"

So no, Deutsch, who is arguably

theleading advocate of MWI, does not consider parallel universes to be a "metaphor/model/aid to understanding". He insists that they are real. He even defends this position at extraordinary length in chapter 10 of TFOR.MWI provides a clear, comprehensive explanation for all of the “weird” quantum phenomenon. Beyond that, this quote from FoR [1] convinced me:

“To those who still cling to a single-universe world-view, I issue this challenge: explain how Shor’s algorithm works. I do not merely mean predict that it will work, which is merely a matter of solving a few uncontroversial equations. I mean provide an explanation. When Shor’s algorithm has factorized a number, using 10^500 or so times the computational resources than can be seen to be present, where was the number factorized? There are only about 10^80 atoms in the entire visible universe, an utterly minuscule number compared with 10^500. So if the visible universe were the extent of physical reality, physical reality would not even remotely contain the resources required to factorize such a large number. Who did factorize it, then? How, and where, was the computation performed?”

[1] https://www.newyorker.com/magazine/2011/05/02/dream-machine/...

"Interpretations of quantum mechanics, unlike Gods, are not jealous, and thus it is safe to believe in more than one at the same time. So if the many-worlds interpretation makes it easier to think about the research you’re doing in April, and the Copenhagen interpretation makes it easier to think about the research you’re doing in June, the Copenhagen interpretation is not going to smite you for praying to the many-worlds interpretation. At least I hope it won’t, because otherwise I’m in big trouble."

- Peter Shor (https://www.scottaaronson.com/blog/?p=218#comment-5753)

If one die can have 6 values, how is it possible that two identical dice have 21 configurations? Where are those additional configurations coming from if there are only 12 possible values in the pair of dice?

This is a stupid argument, comparable to "if humans evolved from apes why are there still apes?"

An analogous question: when three objects are orbiting each other in space, where is the computation being performed that provides the solution to the three-body problem that represents their motions?

Well, obviously, the computation in the latter case is being "performed" by the three bodies themselves, and the Shor computation is being performed inside a quantum computer that exists in our universe. It's just that the stuff that is performing the computation isn't atoms, it's entangled qbits. There's no need to bring entire parallel universes into it.

Can you explain how qbits work then? How exactly do they take on multiple values at the same time?

Not in an HN comment, but there's a lot of material available on the internet. Search for "introduction to quantum computing."

I'm sorry to downvote you, but I think this is both a misrepresentation of MWI and a distraction from the conversation that we're trying to have.

I think Deutsch is doing a very good job of explaining exactly the concept I'm trying to explain.

But you still have to peel back the words to think about what kind of mechanism they are describing, which Deutsch is trying to communicate. You can't just complain that he says "reality contains parallel universes" without accepting that he's taken the time to map "parallel universes" to a well-defined concept.

I'm not

complainingabout anything. I'm simply observing that Deutsch and other prominent advocates of the MWI do not consider multiple worlds to be metaphorical. They insist in no uncertain terms that they are real, and go to great lengths to justify this position. There is absolutely no ambiguity with respect to what they believe.Deutsch may believe that. (Heck, I believe it). It's a natural interpretation of the MWI wavefunction. But it

isan interpretation, a layer on top of the physical claims, a model - not a mandatory part of MWI. Deutsch may interpret a given wavefunction as describing a particular ensemble of 'tangible' and 'shadow' objects, but someone who interprets that same wavefunction a different way (but agrees on the same wavefunction, and expects it to evolve according to the same laws of quantum mechanics) will still generate the same experimental predictions and, more to the point, still be doing MWI.That makes MWI a "no true parallel Scotsman" interpretation.

Either the other versions of reality are tangible, or they're not. If they're just abstractions, you have't solved the problem - you've taken a long way around a circular argument that starts with the wavefunction and ends with it doing something that results in a measurement, maybe.

Unfortunately explaining how that happens is the problem you're trying to solve. It's not an answer to the problem.

If they are real, there are all kinds of other issues, the Born Rule being the most obvious.

> Either the other versions of reality are tangible, or they're not. If they're just abstractions, you have't solved the problem - you've taken a long way around a circular argument that starts with the wavefunction and ends with it doing something that results in a measurement, maybe.

The wavefunction is physically real. Whatever you interpret it as is your business. We've solved a real problem: we can explain the evolution of the wavefunction and the results of experiments without needing some weird collapse/decoherence concept that breaks all the usual rules. The Born Rule is still something that remains to be explained, agreed, but it's a much smaller something than the full Copenhagen style collapse.

There seems to be some confusion here, based around the usage of words.

You seem annoyed by the terminology of words like "worlds" or "multiverse", but are you also thinking about what a MWI advocate is referring to by the usage of those words? They are shorthand for "the observable wave function" and "the global wave function".

There being "copies" of Schrodinger observing his cat is the most intuitive and expressive way to summarize Schrodinger being in a linear superposition of two states in a (cat alive, cat dead) basis. It's not any different from how a Copenhagenist would treat a photon passing through a polarizating beamsplitter as being in a superposition of having passed through and having been reflected; if the photon then goes and interacts with other things, they would keep track of what would have happened in each "branch" of the superposition.

If you have an issue with those, why don't you have an issue with the idea of superposition itself? It's exactly the same concept.

(UPDATE)

In particular, when MWI people say "multiverse", it's not a

sleight of handwhere we are being deceptive when we actually mean "observable and unobservable components of the wave function". The term "multiverse" had no prior meaning in physics; MWI isdefiningit to mean the global wave function, using that word to draw attention to the fact that it has unobservable components.These words are being given rigorous definitions that align with their intuitive meaning. You seem to be arguing against that practice?

Should I complain that when a Copenhagenist says "collapse", nothing is actually falling down and so it's misleading?

No it isn't. Copenhagenists insist that the wave function "collapses" upon "measurement" and so measurements only have one outcome in point of metaphyisical fact. Many-worlders insist that the wave function does not collapse, and hence measurements have many outcomes in point of metaphyiscal fact.

You may consider this a distinction without a difference, but the fact of the matter is that people do argue about this. A lot.

I agree that it's an important difference! And it matters a lot. I am just saying that the concept of "copies" is no more mysterious than the concept of "superposition", and if you can accept the latter then I don't see why you complain about the former.

As for collapse, Copenhagenists can't seem to explain why an interaction between a photon and electrons in a polarizing filter does not collapse the wave function, but an interaction between a photon and electrons in my retina does. They're made of the same stuff and should be following the same quantum rules; why does one of them perform a unitary transformation and the other a lossy projection?

The reason I can accept superposition is that I have first-hand access to experimental evidence of it. I can see interference fringes with my own eyes. By way of contrast, not only do I not have any direct evidence of the existence of copies of me, I

cannot possiblyhave such evidence. And this is not a technological limitation. This is a limit predicted by the theory itself.Your usage of "many outcomes" is what may be causing contention/confusion for those looking for answers in what is described by the mathematical equations, and scientific theories. If you want to justify "many outcomes" on metaphysical grounds, that is not a scientific conversation.

Also "metaphysical fact" is a questionable term.

Confused. Consider a light beam with only one photon at a time (very dim) measured by a photoreceptor. Put a 2nd photoreceptor after the first - it detects no photons.

So detecting the first, definitely 'collapsed the wave function'. This is not an unknown result in physics.

> the main difference is the difference in ontology

Sure, I'm not sure how much Science has to say about ontology but I'm interested in the outcome.

"self-location uncertainty" seems to still assume what is basically the Born rule:

>MWI probability postulate: The probability of self-location in a world with a given set of outcomes is the absolute square of that world’s amplitude. (http://philsci-archive.pitt.edu/14389/1/defence-location-unc...)

It just applies the absolute square of the amplitudes to the chance of being in a given world rather than the chance of an event. The thing that needs explaining is why either probability is given by the square of the amplitudes.

Fair enough - I'm not a physicist or read papers on physics. I'm just trying to re-tell the best version of the theories as I've understood them by steel-manning arguments laid out (by scientists that have some media exposure), rather than try to straw-man each theory with prima-facie criticisms that have already been answered.

I know MWI has significant gaps, this appears to be one of the biggest.

thank you for citing that paper directly.

Ta. As far as I can tell a lack of an explanation for the Born rule is a feature of all quantum interpretations, not just many worlds. It's pretty much the central mystery of quantum mechanics.

Conversely, Copenhagen-like explanations need to explain (or postulate) a decoherence/collapse operation that violates all the usual rules of QM: it would be non-time-symmetric, non-unitary, likely non-local (propagating faster than light), ...

Non-locality is practically axiomatic if you understand Bell's Inequalities, and doesn't require true FTL signalling.

MWI doesn't really solve either the locality or the FTL problem. If you're going to believe in MWI you have to explain how an entire universe appears instantaneously, identical in every way to another universe, except for one almost insignificant difference in exactly one quantum interaction.

That's a rather larger and more serious signalling and information transfer problem than the much smaller-scale non-locality required to explain entanglement.

> Non-locality is practically axiomatic if you understand Bell's Inequalities, and doesn't require true FTL signalling.

My complaint is exactly that the Copenhagen interpretation introduces that unnecessary FTL signalling requirement. Under the MWI, nothing weird is going on in the EPR thought experiment. But if you believe that a collapse occurs when we measure system A, then that collapse has to somehow occur non-locally and/or be FTL-transmitted between A and B.

> If you're going to believe in MWI you have to explain how an entire universe appears instantaneously, identical in every way to another universe, except for one almost insignificant difference in exactly one quantum interaction.

You don't; to talk about it in those terms is putting the cart before the horse. You have to believe that a wavefunction evolves in the usual way and represents the thing that it usually represents. You have to believe that when a pure system interacts with a system in superposition then the result is two systems that are entangled and in superposition - something you presumably already believe if you believe in any kind of QM at all. You only have to believe that a new universe appears instantaneously if your interpretation of putting a particle into superposition is that a new particle appears instantaneously, which is not an approach I've ever heard advocated.

I don't think MWI claims, exactly, that an entire universe 'appears instantaneously'.

Couldn't MWI represented with a probability cloud? Maybe the many-worlds exists as a fixed-volume probability on all states in the universe.

From this perspective, MWI would be similar to a multi-universal annealing system, where different quantum outcomes result in fractures within the probability cloud (as opposed to an expansion of the tree-graph).

Carroll postulates that all the necessary degrees of freedom are already just sitting there in hilbert space - it's only when the states become orthogonal to each other (e.g. "branching") that we no longer have access to them. https://www.youtube.com/watch?v=jHLfMXvQqX8

It should be noted though that few interpretations have an explanation for the born rule.

Actually most interpretations have the same explanations for the Born rule, it's just that people who study many worlds are more inclined to consider them. People who believe in Copenhagen don't care about this stuff.

For a recent high-quality example: https://www.nature.com/articles/s41467-019-09348-x

Right, most of them just take it as an axiom. But the MWI explicitly

rejectsthe Born rule as an axiom, insisting that the Schrödinger equation is all the actual physics there is.MWI is not incompatible with postulating the Born rule if you want to. It just doesn't elevate it to an axiom of the theory. It's like doing constructive mathematics: you're not

denyingthe law of excluded middle, you can postulate it and it will have the usual consequences, but the core theory doesn't rely on it, and it's possible to hope for a more elegant alternative.Yes, you

canjust add the Born rule as an additional axiom, but that undermines the whole point of adopting the MWI in the first place, which is to remain mathematically "clean" with no ad-hoc assumptions.Yes and no. I'd argue that the Born rule as a rule for subjective probabilities in a MWI context is much cleaner than a full Copenhagen-style collapse operation.

You are, of course, free to argue that. But most advocates of the MWI will disagree with you. That's why they put so much effort into trying to derive the Born rule.

Not that it really matters but I disagree that MWI advocates like it because of the born rule derivations. Actually most born rule derivations apply equally to all interpretations.

I’m with Vaidman on this, who is one of the best known supporters of MW, which is that the clearest solution is just to postulate the connection between psi magnitude and subjective experience. It’s an empirical discovery that works.

There's a significant gap between "cleaner than a Copenhagen-style collapse" and "clean enough to be a good part of our fundamental theory of nature" :).

He didn't say to re-add it as an

axiom. He's pointing out that the "born rule" (or something that gives you probabilities in the way that you need to agree with experiment) can potentially "emerge" i.e. be recovered (if they can find the right proof for it), without having had to assume that up front.It's like when GR can "recover" F=MA, despite not having assumed F=MA that as an axiom.

> He didn't say to re-add it as an axiom

Yes, he did:

> MWI is not incompatible with postulating the Born rule if you want to

"Postulate" (as a noun) and "axiom" are synonyms.

Ok - well then he and I meant different things.

Though Everett in his disertation doesn't call it an axiom he kind of assumes it:

>It is convenient at this point to introduce some notational conventions...

>For a complete orthonormal set we define... [the Born rule]

>In the probabilistic interpretation this distribution represents the probability distribution over the results of a measurement...

https://www-tc.pbs.org/wgbh/nova/manyworlds/pdf/dissertation... page 34

Fun fact, Hugh Everett is “E” Mark Oliver Everett’s father, lead singer of The Eels.

Good documentary on that. (https://en.wikipedia.org/wiki/Parallel_Worlds,_Parallel_Live...)

Came here for this :-D.

:-D

"Let me address another issue with many-worlds. It is a deterministic theory, even a hyper-deterministic theory, i.e. determinism applies to everything in the entire universe. Indeed, since there can’t be any influences coming from outside and since the Schroedinger equation - the only dynamical equation of the theory - is deterministic, everything that happens today, e.g. what I am writing, the way each reader reacts, the details of all solar eruptions, etc, was all encoded in some “quantum fluctuations” of the initial state of the universe. Given the complexity of the (many-) worlds, it had to be encoded in some infinitesimal digits of some quantum state, possibly in the billionths of billionths decimal place. I am always astonished that some people seriously believe in that."

- Nicolas Gisin "Collapse. What else?" https://arxiv.org/abs/1701.08300

This seems wrong to me? A lot of "things that happen" seem to be created by randomness generated along the way, from only observing subsystems.

I.e., suppose I create a random bit generator by shining photons on a half-silvered mirror and then measuring which way they go. I might then decide what to write depending what the bit is---if it's 0 I will write a HN comment, if it's 1 I will reply to an email.

In this case, the things that happen are determined deterministically, but it's not the case that there is some quantum fluctuation in the initial state of the universe which determines what I write. Rather, when I do the measurement of the bit, I create two branches, one with a HN comment and one with an email. Observers entangled with the first branch can only see the HN comment, but this complexity was created at the time of the measurement, not at the time of the big bang...

There is no "randomness generated along the way" in the MWI. Schroedinger's equation is deterministic and according to many-worlders is the only thing that "makes things happen". Everything that happens in every "world" was bound to happen since the beginning of the evolution of the "multiverse".

Right, but my point is that this doesn't mean that the initial state contains very much information.

Maybe a different example makes this clearer. Suppose you use a quantum random number generator to choose letters from A to Z, and write out a full book this way. If you consider the entire (multiple) world, it's basically Borges' Library of Babel: it has non-zero amplitudes for every possible book. So one possible "thing that can happen" is that Shakespeare's Hamlet gets written.

But that doesn't mean that the initial state of the universe somehow encodes the text of Hamlet. The initial state of the universe doesn't have any information at all, in quantum fluctuations or otherwise. You only see Hamlet emerge if you restrict your attention to some particular branches of the world. It was "bound to happen" that Hamlet would get written, but only because every book would get written.

The initial state contains as much information as the current state if the evolution is unitary (only Schroedinger’s equation matters according to the MWI, right?).

And surely the wavefunction of a system of 10^80 particles (or whatever) does contain some information about what happens... or do you think that “every book would get written” independently of the form of the universal wavefunction?

But the "current state" of the entire wavefunction doesn't necessarily contain very much information. You only get most of the complexity (solar flares, novels) if you look at subsystems of it.

I think the situation is exactly analogous to the Library of Babel in Borges' story. If someone sends you the

entirelibrary in the mail, they have given you zero bits of information. If they only send you a particular room of it you get more information, and narrowing it down to sending a particular book gives you more bits still.It's the same with the multiverse. If we start out with the photon about to go through the half-silvered mirror, the state looks like |photon>⊗|me>, and then over time it evolves into

(1/√2)|photon went left>⊗|me observing left> + (1/√2)|photon went right>⊗|me observing right>

and the latter state contains no more information than the former one (you can run the time evolution backwards to recover the original state). However, if you restrict your attention and only consider one of the branches, e.g. |photon went right>⊗|me observing right>, then you

haveadded (one more bit) of information. And most macroscopic processes can be explained "one branch at a time", since in practice decoherence means that there are no noticable interactions between branches. I think in the MWI this is the source of most complexity in the universe: not detailed information encoded in the initial state, and not objective collapse, but the bits of information you implicitly create if you only consider one branch of an experimental outcome.I imagine the same thing happens in general: we start with some uniform soup (with ~zero information), in a bunch different MWI branches it self-gravitates slightly differently so we get a different set of galaxies and stars in each branch, then for each branch that has a planet, different quantum randomness may cause different intelligent life to develop, and eventually they may write different books. Of course, it's also possible that some small initial state gets magnified, as a separate mechanism, but that's the same in objective-collapse and MWI. The random bits created by objective collapse are the same same the ones you get by looking at particular branches in MWI.

> If we start out with the photon about to go through the half-silvered mirror, the state looks like |photon>⊗|me>,

But in the MWI what we start with is the universal wavefunction |Psi>. Say there are 10^80 particles in the universe and they may have a spin and could be in infinite locations. That makes for an ininite-dimensional Hilbert space. If we consider that the precision in the location variables is limited by Planck's length there are still around 10^180 possible locations in the universe for each particle. The Hilbert space where the quantum state of the universe is defined is huge.

Now, you say "the state looks like |photon>⊗|me>". To keep things simple and in the realm of Schroedinger's equation I will consider an electron instead of a photon, prepared in the state spin z=+1/2.

Is "|up>⊗|me>" shorthand for "(1 |up> + 0 |down>)⊗|me, the electron at that position, the rest of the universe>" which would be the "in-branch" quantum state of the universe?

If indeed |Psi> can be written as a superposition of that state and an inconceivable huge number of orthogonal states, are all those states (when their coefficient is non-zero) the multiple worlds?

But given |Psi> there are infinitely many (or at least a lot of) possible basis to express the pure state as a decomposition (including of course the one where the only non-zero "world" in the superposition is the one described by the wave function |Psi>). Do we count as "worlds" as well the states appearing in other basis?

Is this slicing of |Psi> into other states just a mathematical trick or does it have a physical meaning? Even if we can redo the operation at different times on the evolving |Psi(t)>, how are the "emergent substates" at different times related?

The answer to all these questions is far from trivial.

Indeed not, but is any of this relevant to the original quote, that deterministic dynamics means that the initial state of the universe must have lots of information hidden in the far-off decimals? I would hope that one could consider a simplified setting for that and then extrapolate back to the continuous case.

I agree the original quote doesn’t present a very strong argument against the MWI. In the same way that arguments against determinism in classical physics are not very convincing. But seemed worth pointing out.

In the MWI case it’s not clear how sensitive the “worlds” are to the in initial state. It’s not even clear how sensitive they are to the current state! Better understanding of how (and how many) worlds “emerge” would be needed to say if the “problem” is or not as bad as in the classical case.

Is the number of worlds stable? Is it increasing? In any case, those worlds are the inevitable consequence of the initial conditions even though we can’t even estimate how much fine-tuning is needed. I don’t think “randomness generated along the way” is a satisfactory counter-argument.

I don't think we know enough to presume what was going on in the "initial state" of the universe, except perhaps that it was very low entropy.

I'm not sure if the Aeon article explains what Everett really meant, at least I still have questions.

Suppose we're eternally branching, as in the example about the up/down spin. If there's no way to visit other branches, how will we ever confirm this?

Second, what is the benefit of the branching interpretation? It seems to simply be a way to think about probabilities?

I'd recommend reading some of the explanations given by Sean Carroll on the subject. https://www.preposterousuniverse.com/

....and I didn't realize that Sean Carroll actually wrote this article! Still, he does some good explanations on his podcast.

The benefit is that instead of two rules for quantum behavior 1) the Schrodinger equation 2) the born rule , that

contradicteach other, you just have 1) the Schrodinger equation, which you already accepted.There is also a higher level goal that also motivates him

> the quest for quantum gravity is being held back by physicists’ traditional strategy of taking a classical theory (such as Albert Einstein’s general relativity) and ‘quantising’ it. Presumably nature doesn’t work like that; it’s just quantum from the start.

1. No, there is no way to visit other branches.

2. The benefit is that it's a very direct translation of the math into concepts. If you actually build the mathematical model, the "other worlds" are right there in the model, as the other half of a superposition.

It doesn't go away. You just can't get to it. Nor can you ignore it, because it's sitting there as a term in the equation, persisting as you evolve it over time. It just becomes smaller and smaller without ever becoming zero. That's why you observe a classical universe even though it's really quantum: the odds of the other parts of a superposition having an effect are on the order of Avogadro's-number-to-1 against.

Copenhagen works by just arbitrarily defining that number as zero, which is easy to work with but conceptually incomplete. MWI keeps it there in another "world", which is more in keeping with the math but unedifying because it sounds so unfalsifiable.

Hope that helps. (Which is always code for "... but it probably doesn't", I'm afraid.)

Quantum Immortality. Basically, if you keep splitting and you make an assumption like "My perception of the world is relative to me", then when a split causes you to die you don't really split - because you must ignore branches where you don't exist.

So then your view always follows the branching path where you live and you should expect to live 1M years from now. If there is only one world you shouldn't expect that.

As scld touched on, this "article" is really an excerpt from Sean Carroll's new book,

Something Deeply Hidden: Quantum Worlds and the Emergence of Spacetime.https://www.amazon.com/Something-Deeply-Hidden-Emergence-Spa...

If you wish to read further, you may want to consider picking up a copy. Mine is waiting for me at home, I'm looking forward to it!

Haha... Anything is in some sense "in a box" until information passes to it from another thing. Philosophical points only...