Hacker News Reader

Comment by JulianMorrison | Hacker News Reader

Why is it that none of these experiments seem to consider the "many worlds" understanding of "spooky action at a distance" (which is non-spooky, and action only at immediate contact)?

Roughly: both the samples are in both world A (up/down) and world B(down/up), and when you interact with them, you get split and are pulled into both world A and world B, and each copy of you when you interact pulls the thing you interact with also into A and B, and so when you compare results, the A-you compares the A-result and the B-you compares the B-result.

The splitting is not immediate or faster than light. But it moves at exactly the speed of information, so you can't race ahead of it and watch it happen. To interact with it is to become part of it.

View on HN

Because it doesn’t add anything into the discussion.

If the test can't differentiate between QM interpretations, assuming some specific hypothetical interpretation is useless.

Right, exactly. I experience great ennui when people discuss this stuff via popularizations.

Let's start a thread where we talk about how tcp/ip works ... but with a twist. The only admissible references in the discussion are color pictures aimed at non-programmers. Anybody tries to talk about what actually works, we label them the "shut up and program" crowd and decry their lack of imagination.

But it does subtract something from the discussion. It subtracts spookiness, it subtracts action at a distance, and it subtracts anything going faster than light.

And it adds this whole world-collapsing unnecessary machinery that has no practical applications or physical evidence. I myself would prefer Bohmian mechanics if we had to choose a single interpretation. Moreover, it's possible to define a non-stochastic version of the theory (no gods playing dice), which would explain QM somewhat like Classical Mechanics explains Thermodynamics; and it's falsifiable by looking for non-equilibrium conditions (admittedly those conditions will be very hard to achieve).

https://en.wikipedia.org/wiki/De_Broglie%E2%80%93Bohm_theory...

https://en.wikipedia.org/wiki/Quantum_non-equilibrium

My understanding was that MWI allows superpositions of macroscopic objects (the universal wavefunction), whereas the Copenhagen interpretation doesn't (macroscopic objects cause wavefunction collapse), which is in principle testable [1] (although not with current technology).

[1] http://plato.stanford.edu/entries/qm-manyworlds/#5

One surely has to question any claim that is

a) only testable with "unforeseeable future technology" and

b) whose "burden of an experimental proof lies with the opponents".

Win-Win there I think.

a) Nobody said "unforeseeable". Macroscopic superpositions are entirely foreseeable, we just don't have the technology to keep them in a coherent state.

b) The link only says that some people have argued that position. I think all they're saying is that "wavefunction collapse caused by some arbitrary division between microscopic and macroscopic objects" is a stronger claim than MWI

> Nobody said "unforeseeable"

OK. The actual quote from the linked reference was

"These proposals are all for gedanken experiments that cannot be performed with current or any foreseeable future technology".

So I took a little license in quoting. I should ha quoted it more correctly, you're quite right.

That sentence is not talking about MWI, it's talking about wavefunction collapse interpretations.

This experiment[1] seems to rule out "macrorealism" (the principle that objects bigger than a certain size are immune to superposition).

[1] https://physics.aps.org/articles/v8/6

Many worlds interpretation: adding approximately zero explanatory value, at the low low cost of infinite parallel universes.

I think the best way to understand MWI is that a lot of physicists/philosophers are more comfortable with the idea of inhabiting an infinity of forking but "real" (essentially classical) universes than they are with inhabiting a single "unreal" universe in which QM is taken seriously.

> infinite parallel universes

That's a pretty uncharitable interpretation of it. If you look at the actual physics, MWI is incredibly simple: the wavefunction just evolves as the wavefunction evolves. That's it. There's some amplitude over here, and there's some amplitude over there, and this possibility and that possibility are mutually exclusive etc (with some "weird" quantum correlations, of course.)

The Copenhagen interpretation's "wavefunction collapse" is just like a Bayesian update -- the wavefunction you're left with is like a conditional probability. Maybe it makes you happy to throw away the "other universe" bits of the wavefunction and normalise probabilities on your own branch, but you obviously don't have to, and any discussion of when the collapse happens (or worse, whether it's a physical process that propagates across the universe faster than light) is obviously very silly business.

Any other intepretation requires new, undiscovered physics. You have to make additional assumptions in order for the equations of quantum mechanics not to give rise to many world.

Because standard academic discourse in physics journals is not aimed at providing intuitive explanations, it is aimed to provide "shut up and calculate" results and try not to spend much time on questions of what is real. It wouldn't surprise me if the "shut up and calculate" crowd was the dominant object-level belief among physicists, even if most of them could agree that by just assuming quantum mechanics applies universally, that is to say the universal wavefunction is real, then that implies MWI and any collapse postulate on top of that is an unneeded detail stripped by Occam's Razor. To the ideologically motivated this might be frustrating because for the MWI proponent (I'm sort of one) framing everything with MWI seems to simplify a lot of layman explanations. But physicists in the field don't need those explanations framed that way, they need the math and data and engineering setup which I'm not sure is any easier to learn and comprehend and apply if your interpretation belief is one way or another. I'm mainly pro-MWI-for-laymen (practicing physicists can believe what they like, disbelieving can be helpful too if it motivates them to find experimental falsification!) just because MWI trivially dispels pretty much all the pop culture misconceptions about quantum mechanics.

An extension of this is "shut up and test it". MWI has an experiment. I'm ambivalent about it being feasible since on my (layman) reading it sounds like you need to turn back time by proxy to execute it.

I'm not sure your explanation for the lack of spookiness in many worlds works. The essence of most Bell's Theorems experiments is you start with entangled photons (or similar), they travel a long way apart and then go through polarisers at either end and into detectors. The odd thing is that the number getting through both slits depends on the angle between the polarisers. But how does the photon at polariser 1 know what the orientation of polariser 2 is at the time it's pair photo gets there? You can spin the polariser and the chances of it getting through depends on the orientation when the photon gets there, not when it set off. You'd think it can't but experimentally it does. I don't see how that weirdness goes away with many worlds.

(Bell's theorem is basically maths to figure out if the photons have to 'know' the angle between the polarisers rather than just say start off with the same polarisation.)

For as far as I can tell, the solution is that whenever there is a choice to be made, all choices are made and you get separate branches/parallell universe bubbles that expand at the speed of light, interacting (or not) with other universe bubbles. The bubble in which entangled electron A had spin up interacts with the other in which electron B had spin down and vice versa.

So essentially you have an ever growing set of merging and splitting history branches, always growing for every point in space, continously.

I've got no idea how to make that work for your example of polarizers, or for the example of delayed quantum eraser experiments, etc... The surface of the bubbles apparantly have to carry all of that information and there must be some mechanism for them to decide which ones interact with which others.

They aren't exactly bubbles expanding at the speed of light (necessarily). They are more like: when you get a particle interaction whose outcome is determined by a fact which has more than one answer, you get more than one outcome. And that applies recursively as those particles whizz off and interact with anything else.

It's possible though unlikely that they could be expanding no faster than the speed of a walking scientist - if nothing else in the universe is yet causally determined by the outcome.

How does this not enable ftl information transmission?

It seems to me that you have a valve that instantly opens/close a different valve at an arbitrary distance.

Many worlds makes it clear how it can't: because there is no FTL transmission. There is only stuff that has been pulled into the superposition (seeing a definite answer) and stuff that hasn't yet (for which both answers are still superposed). When the latter comes in contact with the former, it enters the superposition, which from the outside perspective is a split. From the inside perspective, it looks like "the superposition collapsed". From the outside, one side gets "open" and the other gets "closed". From the inside, this looks like getting a random answer.

You can see intuitively why this can't convey information. You always get both answers (in different "worlds").

That doesn't really explain the mechanics of the splitting process.

It's "And to explain this, we postulate that a whole new universe appears fully formed out of nowhere, with an identical history but a distinguishable future, because it does."

I suppose it's possible. But it doesn't seem very parsimonious - because you effectively have to double the energy in the multiverse for every distinguishable quantum event.

Does conservation of energy not apply in the multiverse? Are universes - as many as you need - free at point of sale?

Splitting is just a metaphor for a superposition of macroscopic states. There's no extra mechanics involved.

Because whether the spin is up or down when you observe it is completely random. Both observers at either end would believe they looked at it first without any other information, and the stream of bits would be completely random AND 100% correlated with other.

Ok, so you will always have perfect correlation and since you spun the polarizer after the photons started moving that implies spooky action at a distance.

right?

You can't control what the value will be or when it will open/close. It isn't a detectable "push event" like in computer systems.

Because interpretations are just that, interpretations. They are intended to let us understand it intuitively, but the math behind them is always the same.

Interpretations affect how you apply that math to model reality, and hence the predictions you make. Math does not make any predictions about the outcome of an experiment on its own, it's just a formal system.

What??

It doesn't make any different predictions from the Copenhagen interpretation. That makes the question of which of the two (or of the various other interpretations) is "right" untestable, and therefore unscientific.

Probably because the many-world hypothesis seems fundamentally untestable?

I'm not sure any of the interpretations of quantum mechanics are actually distinguishable in any way we can measure, so the best thing to do would probably be to pick the one which makes things easiest to work with.

I think Occam's razor is the cause. Many worlds suffers from the appearance of violating Occam's razor. Of course Occam's razor is just a rule of thumb, but it seems to inhibit physicists being comfortable with the many worlds interpretation.

Objects exist that we cannot see.

Those words provoke little controversy when spoken about matter beyond the visible universe. Why? Because the simplest mathematical description we have of the objects we can see also includes those which in principle we cannot. We grant them the same ontological status as the Earth we stand on -- few would claim them as mere fictions to ease our task of prediction.

Occam's Razor properly applied limits the amount of information in the description, not the amount of things the description describes. For example, pi's infinite decimal expansion is spared by this information theoretic razor. If you reject the multiverse, your full description of the universe must contain one bit for every binary "choice" the universe made, every qubit that collapsed. Quite a few bits that turns out to be. The multiverse description suffers no such flaw, as it contains only the initial conditions and the rules by which those conditions change over time.

It isn't, though.

It doesn't create worlds. There's always only exactly one fuzzy-quantum-thing. It's just that some parts of it no longer interact with other parts of it.

I know that is why I said appeared to violate, not violate :)

It's an interesting thought for sure. Another is that there is the one-electron universe: https://en.wikipedia.org/wiki/One-electron_universe

I have have no idea.

Reply on news.ycombinator.com