Hacker News Reader

The objects in boxes scenario you have described is useful to think about when coming to grips with QM and entanglement. However, one needs to be careful, because there is an additional subtlety: QM can have measurement scenarios that are not only entangled, but also non-local [1]. This means that it is provably impossible to endow the variables that you are measuring with prior, local assignments, e.g. like in the boxes scenario. So there is actually another phenomenon present in QM that makes it weird, and is strictly speaking distinct from entanglement.

[1] - https://en.wikipedia.org/wiki/Quantum_nonlocality

locality is looking suspect for other reasons too, as it's one of the only ways to avoid 'black hole firewalls' [1] (the other way to avoid firewalls involves giving up conservation of information which would be bad for QM theory).

[1] Black hole Firewalls with Sean Carroll and Jennifer Ouellette https://www.youtube.com/watch?v=_8bhtEgB8Mo&t=3634s

So to me, that doesn't suggest that something "traveled" to the other box, just like in the Classical version. Rather, somehow I only "knew" the system at the beginning without knowing "any of it's parts" (due to entanglement & superposition). Then, just observing a degree of freedom in one the parts finally reveals what the corresponding degree of freedom in the other part was. This (intuitively) makes more sense to me than saying "information traveled" and "action at a distance"

But exactly this is not the case, exactly this is the difference between classical physics and quantum physics. Classically you place one coin in each of the two boxes, heads in one, tails in the other. Nothing strange happens here, the outcome is determined when setting up the boxes. When you look in one of the box, you will see either heads or tails as determined by the setup. The other box will have the coin showing the opposing side, also determined by the setup.

The quantum case is very different. You again place one coin in each of the two boxes, again one heads and one tails, but this time the coins are constantly flipping. It is no longer predetermined whether you will get heads or tails and only when you look in one of the boxes does the coin in there stops flipping and you see either heads or tails. But more importantly the other coin stops flipping at the very same time so that you will see the opposing side when you look into the second box.

The correlation is in both cases established when the boxes are setup, the coins will always show opposing sides. The quantum cases of course also requires both coins flipping synchronously. But in the classical case the exact outcome is also established during the setup while in the quantum case the exact outcome is only established when one looks in the first box.

> the coins are constantly flipping ... the other coin stops flipping at the very same time

I'm not sure this corresponds to my understanding of the wave function before observation.

Just because the probabilities for each state is less than 1, to me doesn't mean that it's "constantly flipping". Only that upon measurement there's a probability that one of those states gets defined/constrained.

This is what i think susskind meant by understanding the whole (probabilities of the wave function of the whole system), without understanding the parts "the particles" i.e. it's kind of not right to think of just the parts here. It's only the system+parts that must be thought of together (which of course is why one of the reasons no-one truly understands it).

I'm not sure this corresponds to my understanding of the wave function before observation.

This was only meant as a classical model how the outcome is not predetermined, I did not want to imply that the wave function oscillates back and forth between two states or something like that. I could also have said that you place two coins in each box and one randomly explodes when you look in the box. And of course the corresponding coin in the second box also explodes simultaneously.

I'm not sure your description of the difference between classical and quantum contradicts what I said. In fact I think they are compatible

I didn't use any notion of "predetermination" when describing the quantum mechanical version. I said that "there could be several different degrees of freedom to measure on" I meant this to exactly correspond to "the exact outcome is only established when one looks in the first box".

The only notion of "predetermination" I ascribe to is that of the whole system, which is completely described by the quantum states before observation, but as i said this tells us nothing of the parts it's composed of.

(I am choosing in my description to put less emphasis on the particles/parts (which normally get all the attention), and more on the quantum system)

I was especially responding to this part.

This (intuitively) makes more sense to me than saying "information traveled" and "action at a distance"

You can not get away from non-locality in quantum mechanics. Because the outcome is only determined when a measurement is performed, this »information« has to travel to the entangled partner when the measurement is made. And this must happen instantaneously. In the classical example you can avoid this because the outcome is already determined during the setup.

I said that "there could be several different degrees of freedom to measure on" I meant this to exactly correspond to "the exact outcome is only established when one looks in the first box".

This are three different things. You can of course decide to measure the spin along any axis you like, that is you can choose the observable. When you do this, then you are not measuring different degrees of freedom, you are measuring the same degree of freedom, the spin, in different bases. But, and that is what I meant, even if you always measure the spin along the same axis, say along the positive x axis, then the spin of an electron in a superposition of spin either +1 or -1 along this axis is only determined at the time of the measurement. Before that measurement the electron had no definite spin along this axis.

> You can not get away from non-locality in quantum mechanics.

I didn't/am not claiming that the effects are local.

> only determined at the time of the measurement... Before that measurement the electron had no definite spin along this axis.

That is what I'm saying as well. I didn't claim otherwise.

Well, then I just misunderstood you. I should probably better have quoted this sentence in my last comment.

So to me, that doesn't suggest that something "traveled" to the other box, just like in the Classical version.

I read this as you saying nothing travels from the measurement to the entangled partner. But this is what happens, »Someone just looked inside me and saw heads, please show tails when someone looks inside you!« travels from the measurement to the entangled partner. And if that happens instantaneously, then this is a non-local effect.

That is why I thought you wanted to say nothing travels, neither slowly nor instantaneously and therefore non-locally.

Very little about QM is intuitive at the ontological level; what matters is that it's logical at the mathematical and practical level. Lasers work, the math works, the ontology doesn't.

sure, but I'm not sure we should be saying that it doesn't have an ontology (even if it is a very weird one), just because the only model that we've been able to come up with thus far is mostly a calculational one (as QM "calculates" probabilities).

It doesn't have an ontology with a current prayer of being falsifies, which is the same thing if you're interested in the science and not the popsci. Exceptions exist, hence AdS/CFT correspondence and M-Theory, but they are exceptions. Even then, all that can be said is various tools have been created with value beyond their fields, not that they're in any way falsifiable yet.

At any rate, "intuitive" strikes me as a bad test for QM.

Yeah, that's why QM and I parted ways. I'm too committed to intuition. So it goes.

I mean, when you say that you parted ways, did you go over to something like deBB?

Biology :)

Ahhh, that makes a lot of sense; big warm systems that are still fascinating, but definitely not dominated by QM.

Yeah. But in retrospect, QM chaos seems to underlie a lot of biology, from evolution to neural development to immune response. But there's been no need to think much about QM except as a source of randomness. So far, anyway.

Fingers crossed At the very least, if QM emerges into your end of things, it should do so in clearly defined ways, and at scales/energies which have some hope of being probed.

I do love Stephenson's Anathem :)

Honestly, I love everything he writes... The Diamond Age in particular I think is a bit underrated.