I didn't watch the linked video. For the record, quantum mechanics does not say the multiverse is real or is not. Not sure what 'physics' is wrong or conflicts with QM.

I think different people have different definitions of what 'is real'.
One can view the universe from within or from without and yield different answers to the same questions, and nobody is wrong or right.
So from within, only what we can sense or what possibly can effect us is real. It is not wrong to say there is no multiverse. From an objective view, there may be multiple worlds, but still one universe. These two different ways of looking at it are completely valid for not just Type 3, but also types 1 and 2.

So it seems there is no multiverse, but only because that is a poor choice of words. Everett proposed one universe with different solutions to measurements, but no 'collapse' that makes one of those solutions the actual one. That's multi-world perhaps, but not multiverse.

For those of us who prefer not to sit through a video and who have only a nodding acquaintance with the topic, can you please remind us what a level 3 multiverse is?

I think it is time to take physics seriously. — tom

On the contrary, it seems that in the past few years it's becoming time to stop taking physics seriously. Theories that can not be experimentally verified or refuted are not science.

The Level III Multiverse (Tegmark) gets a lot of stick, despite the fact that it adds zero complexity to our conception of Reality according to known physics. — tom

It seems a stretch to say the many worlds interpretation doesn't add an unnecessary amount of complication to our metaphysics.

I still prefer to hope for a physics that fixes the MWI's unbounded world-line branching by incorporating the actuality of a generalised wavefunction collapse.

But with decoherence maths - QM+statistical mechanics - we get a workable collapse anyway. So there's no burning need to embrace the extravagant ontology of MWI. Decoherence delivers a quasi-collapse that makes the theory fit with the world we observe quite adequately while a "proper" theory is still under development.

So while you claim that MWI adds zero complexity to the maths, you have to admit that the same equations produce too many answers. It is only natural to expect the maths therefore misses some constraint on its unbound fecundity. And now just such a thermodynamic limit has been tacked on with decoherence.

So MWI has become more realistic, more in accord with the world as we know it classically. Yet the very fact that QM+limits is an improvement should be evidence that multiverse thinking was always wrong-headed.

MWI has already begun the job of denying itself, even though many people think it is the "decoherence interpretation". We need something extra by way of a world-constraining mechanism so as to actually reduce the complexity of the QM maths.

For those of us who prefer not to sit through a video, and who have only a nodding acquaintance with the topic, can you please remind us what a level 3 multiverse is? — fishfry

It is a Tegmark designation, and I'm not sure how much the video gets into it. Type (or level) 1: Places that are too distant to causally interact with here, ever. There is a duplicate Earth out there if you go far enough. Type 2 is other bubbles in eternal inflation theory. Those bubbles all have the same QM, but different dimensions, light speed, and other physical constants. There is a duplicate Earth there as well. Type 3 is Everett multi-worlds, essentially parallel-here. Type 4 is other structures altogether, and it is hard to argue that they're not separate universes, but ours exists no more than those others.

1: Places that are too distant to causally interact with here, ever. There is a duplicate Earth out there if you go far enough. — noAxioms

How does the second sentence follow from the first? Do the universes share the same history? Why should they do that?

How does the second sentence follow from the first? Do the universes share the same history? Why should they do that? — fishfry

They're not separate universes (especially types 1 and 3), just separate worlds in this universe. For type 1, the distant Earth is a true duplicate. The space is infinite, but the possible states in a finite space (say that of Earth) are not, so each state much eventually be duplicated given enough distance.

The same objection that Apo applies to Type 3 is relevant to type 1. Those distant hubble-spheres are just other (sometimes identical) solutions to the same equations. In both cases, there is no separate ontology to it. There is only the one universe, and not a multiverse of separately existing noninteracting things.

The space is infinite, but the possible states in a finite space (say that of Earth) are not, so each state much eventually be duplicated given enough distance. — noAxioms

That would not seem to follow, since no distance is infinite.

That would not seem to follow, since no distance is infinite. — Janus

Models say otherwise. For the distance to be finite, there would need to be an edge where there is stuff only on one side, and not uniform as we see it. This is true of a subjective model (one with a frame and a 'current event'), but not of any objective view. Other-worlds is necessarily a description from the outside.

For type 1, the distant Earth is a true duplicate. The space is infinite, but the possible states in a finite space (say that of Earth) are not, so each state much eventually be duplicated given enough distance. — noAxioms

First, why is the space infinite? Your premise is that the universe is finite but sufficiently large that there are regions inaccessible to each other due to light not having had enough time to get from one place to the other. That doesn't require an infinite universe.

Secondly, even if the universe is infinite AND the possible states in a given region of space are finite, you STILL are not guaranteed a duplicate earth. Say there are two states, and infinitely many universes:

0, 1, 1, 1, 1, 1, 1, ...

So SOME state occurs infinitely often. But the 0 state is never duplicated. In other words if there are infinitely many universes, then SOME rock or bacteria or planet is duplicated for sure. But you personally, or the earth, are not necessarily duplicated.

Again though, you're betraying your original premise. You started with TWO regions of the universe that are causally and informationally isolated from one another. That's two, not infinitely many.

First, why is the space infinite? Your premise is that the universe is finite but sufficiently large that there are regions inaccessible to each other due to light not having had enough time to get from one place to the other. That doesn't require an infinite universe. — fishfry

Indeed, it doesn't require infinite space. It (a type 1 world, not a duplicate) does at least require an expanding universe, else eventually light would have time to cross the distance. The dup-Earth requires space big enough to form duplicates of something, which could in theory be close enough to be visible from here once light had time to make the trip. That is more probable than what you show below where it by chance just never happens.

Secondly, even if the universe is finite AND the possible states in a given region of space are finite, you STILL are not guaranteed a duplicate earth. Say there are two states, and infinitely many universes:

0, 1, 1, 1, 1, 1, 1, ...

Not guaranteed, no. A coin may flip tails forever. It's just a probability after enough distance.

Edit: I think that mathematically, a coin cannot come up tails forever. There cannot not be a dup Earth given infinite space. The probability of of that is 0.000... which is zero.

Again thought, you're betraying your original premise. You started with TWO regions of the universe that are causally and informationally isolated from one another. That's two, not infinitely many.

Not sure I follow this. If there is a duplicate, there would seem to be an infinite number of them. There is simply a probability as to how far away (measured in non-linear units) the nearest one is. There are infinite type-1 worlds given infinite universe size. In fact, each point in space is centered on such a world, so you are in a different one than I am. Light can in theory reach (immortal) you from your slightly-left-shifted world but never reach me in mine. That means that by the time the light gets to you, the expansion of the universe separates us so far that we're forever isolated from each other.

I think you misunderstood; the point is that there is no actual infinite distance. Even if you traveled away from Earth, for example, forever you would never reach an infinite distance from earth.

I think it is time to take physics seriously. — tom

On the contrary, I think it is time to acknowledge that physicalism as a natural philosophy has utterly failed, and is now leading us up 10⁵⁰⁰ garden paths.

Models say otherwise. For the distance to be finite, there would need to be an edge where there is stuff only on one side, and not uniform as we see it. This is true of a subjective model (one with a frame and a 'current event'), but not of any objective view. Other-worlds is necessarily a description from the outside. — noAxioms

A 3-manifold, e.g. a 3-sphere, is finite with no edge.

For those of us who prefer not to sit through a video and who have only a nodding acquaintance with the topic, can you please remind us what a level 3 multiverse is? — fishfry

I didn't watch it either, I just fast-forwarded to the parts in which the physicists were talking. David Deutsch is particularly worth listening to.

As an aside, there is a video where DD is being interviewed by Dutch TV in the same scruffy front room of his house. They ask to see his office, so he takes them there. They mention that his office is very spartan, it is empty. He replies that he had never been there before. Hilarious!

On the contrary, it seems that in the past few years it's becoming time to stop taking physics seriously. Theories that can not be experimentally verified or refuted are not science. — fishfry

Quantum mechanics has been around since 1925, and if you count Einstein's paper on the photo-electric effect as the start (which it was) then since 1905. I have no idea what you mean by "the last few years", but over the last 100 years QM has been the most tested theory ever conceived, and has passed every test. More than that, it has also revealed several astonishing and novel features of reality - interference of matter particles, superposition, entanglement, ...

And by the way, the existence of the other Worlds is experimentally verified.

They're not separate universes (especially types 1 and 3), just separate worlds in this universe. For type 1, the distant Earth is a true duplicate. The space is infinite, but the possible states in a finite space (say that of Earth) are not, so each state much eventually be duplicated given enough distance. — noAxioms

It is psychologically interesting that people generally accept the existence of the Type 1 multiverse, when there is absolutely no evidence for it. I'm not sure if evidence for it is even possible?

Accepting the Type 1 multiverse, which is a consequence of an infinite universe subject to the Bekenstein bound, then the Type 2 adds zero complexity, but many people recoil at the idea, even Penrose!

Indeed, it doesn't require infinite space. It (a type 1 world, not a duplicate) does at least require an expanding universe, else eventually light would have time to cross the distance. The dup-Earth requires space big enough to form duplicates of something, which could in theory be close enough to be visible from here once light had time to make the trip. That is more probable than what you show below where it by chance just never happens. — noAxioms

The argument for the existence of the Type 1 multiverse requires an infinite, ergodic, and expanding universe.

I think you misunderstood; the point is that there is no actual infinite distance. Even if you traveled away from Earth, for example, forever you would never reach an infinite distance from earth. — Janus

True, but I don't understand how this is relevant to what noAxiom was saying.

For those interested, the argument that, as a generic consequence of inflationary cosmology, there almost certainly exist exact duplicates of Earth (among other interesting things) is given here: Many worlds in one, J. Garriga, A. Vilenkin, Phys.Rev. D64 (2001). (This is still within the parameters of "level-I multiverse.")

And the classification of multiverses is given in Tegmark's Parallel Universes.

True, but I don't understand how this is relevant to what noAxiom was saying. — SophistiCat

Just that it would seem there could be no actual infinite number of planets, and hence the argument that a finite number of possible planets would lead to a duplicate earth would seem to fail.

Thanks for the link anyway; I'll take a look.

Just that it would seem there could be no actual infinite number of planets, and hence the argument that a finite number of possible planets would lead to a duplicate earth would seem to fail. — Janus

How does it seem that there can be no infinite number of planets?

Because there is no infinitely large number, just as there can be no infinite distance; the idea is untintelligble.

The idea of an infinitely large number is not as unintelligible as you think - don't underestimate the intelligence of mathematicians ;) But we don't even need to invoke infinitely large numbers here. It is enough to suppose that there is no limit to the number of planets: for any number that you may choose, there are (supposedly) more planets than that. That's all we need to assume for Vilenkin's argument.

Because there is no infinitely large number, just as there can be no infinite distance; the idea is untintelligble. — Janus

I see. You are proscribing reality based on your intellectual abilities. Not a good idea!

Indeed, it doesn't require infinite space. It (a type 1 world, not a duplicate) does at least require an expanding universe, else eventually light would have time to cross the distance. The dup-Earth requires space big enough to form duplicates of something, which could in theory be close enough to be visible from here once light had time to make the trip. — noAxioms

Correct. Even if the universe is finite there might well be type-1 multiverses, which are simply regions of space too far apart for light to have gotten from one end to the other since the beginning. Which is good, because contemporary physics holds that the universe is finite. There's no sensible theory of an infinite universe, since then set theory would become an experimental science. And there are no grant applications from physics postdocs proposing to investigate the Continuum hypothesis and the axiom of choice. That's because infinity is a mathematical concept and not a physical one. This could change tomorrow, but by contemporary physics, the universe is finite.

Edit: I think that mathematically, a coin cannot come up tails forever. There cannot not be a dup Earth given infinite space. The probability of of that is 0.000... which is zero. — noAxioms

In infinite probability spaces, probability zero events may still happen. Suppose you flip infinitely many coins and they come up in any sequence whatsoever: hthhthththththhthttthhthththt... say. A completely random sequence. What's the probability? Well, the prob that flip 1 is h is 1/2. The prob that flip 2 is t is 1/2. Etc. The prob of the first n flips being exactly what they are is 1/2^n, and that goes to zero as n goes to infinity. Every particular sequence has probability zero. Do you follow that point? All heads is just as likely as alternating heads and tails which is just as likely as the random sequence above. The probability is zero. Yet SOME probability zero sequence must occur.

Not sure I follow this. If there is a duplicate, there would seem to be an infinite number of them. — noAxioms

In your first paragraph you agreed that level-1 multiverses only need a finite universe. And now you are claiming an infinite universe. I ask you (again), what is your evidence for such a claim? Contemporary physics has no such theory other than pure speculation.

There is simply

a probability as to how far away (measured in non-linear units) the nearest one is. There are infinite type-1 worlds given infinite universe size. — noAxioms

Yes but the universe is finite as far as we know. So you're saying, "Well, if Santa Claus is real, he gets to every house in finite time on Christmas eve." That's true, simply by virtue of the premise being false. You say that such and so happens in an infinite universe, but your premise is contradicted by known physics.

In fact, each point in space is centered on such a world, so you are in a different one than I am. Light can in theory reach (immortal) you — noAxioms

So now there are immortal beings in an infinite universe? This is not the science fiction forum. What is the evidence for your premises that the universe is infinite AND that it contains immortal beings?

You're just making this up.

A 3-manifold, e.g. a 3-sphere, is finite with no edge. — tom

Good point.

It is psychologically interesting that people generally accept the existence of the Type 1 multiverse, when there is absolutely no evidence for it. I'm not sure if evidence for it is even possible? — tom

Isn't just a curvature measurement enough? If flat enough, there are places sufficiently separated to never interact. Yes, expansion is required for that, but not infinite space.

I didn't watch it either — tom

Well then I'm glad I didn't watch it! It's bad form to post a vid you didn't watch IMO. But now that you pointed us to the good parts I'll take a look when I get a chance.

I have no idea what you mean by "the last few years", — tom

Multiverse theory, string theory, speculative theories that have zero experimental basis and that can't be experimentally verified even in theory.

Which is good, because contemporary physics holds that the universe is finite. — fishfry

If so, the dup-Earth bit kinda falls apart, eh?

In infinite probability spaces, probability zero events may still happen. Suppose you flip infinitely many coins and they come up in any sequence whatsoever: hthhthththththhthttthhthththt... say. A completely random sequence. What's the probability? Well, the prob that flip 1 is h is 1/2. The prob that flip 2 is t is 1/2. Etc. The prob of the first n flips being exactly what they are is 1/2^n, and that goes to zero as n goes to infinity. Every particular sequence has probability zero. Do you follow that point? — fishfry

No, disagree with this. A finite sequence has a nonzero probability. An infinite one is not a specific one, and has probability zero and does not happen.
Every particular sequence has half the probability of the flip sequence of one-less flip. That is not zero. Some probability-zero sequence is not a particular sequence and thus does not in fact occur.

If so, the dup-Earth bit kinda falls apart, eh? — noAxioms

Well yeah, that's my point. You claimed there's a dup earth, I asked you to support your claim, and you have apparently backtracked. Yes?

No, disagree with this. A finite sequence has a nonzero probability. An infinite one is not a specific one, and has probability zero and does not happen.
Every particular sequence has half the probability of the flip sequence of one-less flip. That is not zero. Some probability-zero sequence is not a particular sequence and thus does not in fact occur. — noAxioms

You said all heads has probability zero hence can't happen. I pointed out that ANY specific infinite sequence has probability zero. Are you unclear on this point? By coding H as 1 and T as 0 and putting a binary point in front of the sequence, an infinite sequence of coin flips represents some real number in the unit interval. If you randomly pick one, the probability that you picked it was zero. But SOME number gets picked.

Forgive my novice question; but, given that the Level III multiverse theory relies on the MWI, then how does one explain the fact that decoherence doesn't contradict the evolution of the wavefunction in every world? Meaning, how is causality preserved in each world and what concept of time is professed wrt. to each world in the Level III Multiverse? In other words, it seems like time is absolute and not relative with respect to each world in the Level III multiverse and decoherence being able to happen at all. Or am I understanding the wavefunction wrong here, as if it were multiple entities existing in each world and not one dictating the behavior of all worlds?

I can see now that an infinitely large number of planets is not needed for the argument, so thanks for correcting me on that.

I remained unconvinced, though, that an infinite number (can there be more than one?) could be specified; because it would seem that any specifiable number must be finite. This is not to deny that an unspecifiable number might be useful for mathematical operations. In any case I see no reason to believe there are infinitely many planets; but admittedly I am no expert on cosmology.