So is quantum entanglement actually FTL “communication” or is it mundane pre-determination?
In short:
- It is not "mundane pre-determination".
- Whether it is "actually" some form of FTL communication depends on which interpretation of quantum mechanics you prefer.
- However, regardless of interpretation, if there is some form of FTL communication between the parties, then it is completely inaccessible to us.
That said, you're certainly correct in one respect: there's a lot of very poor descriptions of what entanglement is and how it works in the popular-science press. In particular, these two paragraphs do a very good job of summing up a very common misconception that arises from poorly-written material:
The usual analogies I hear are things like, well if you have a blue and red marble, you scramble them in a bag and then take one blindly, and give the other to your friend, if you look at your marble and see blue, you know your friend must have red.
To me this sounds like "boringly classical pre-determination." Like this whole spooky action at a distance is really just "the spins were already opposite to begin with and we just separated them," as if we had a machine that spat out pairs of opposite-colored marbles and now we're surprised that they're always opposites regardless of how far apart we are when we see them.
Your interpretation of that description is indeed correct. However, quantum-mechanical entanglement goes much further than that property, and this is exactly the content of Bell's theorem.
To be more precise, Bell's theorem is a description of systems that use "boringly classical pre-determination" (known in the technical lingo as Hidden-Variable Theories) to produce correlated outcomes, and it makes quantitative statements about which kinds and what amounts of correlations you can expect from such a system.
Bell's bigger argument then goes on to construct quantum-mechanical states that break those bounds, and which therefore (provably) cannot be explained with "boringly classical pre-determination" at all. And, when we talk about Bell-test experiments, we refer to experiments that implement those states and show that you do indeed get more correlations than classical pre-determination can produce.
This does raise the question about how your initial model,
you have a blue and red marble, you scramble them in a bag and then take one blindly,
fails to describe the quantum systems at play. The answer is that this classical model is unable to describe superposition states between the blue-marble and red-marble states of each bag, and for the quantum-mechanical description to work at a level where it can exceed the classical bounds on correlations, it needs to be able to measure on the superposition state
$$
|\text{blue marble}⟩ + |\text{red marble}⟩
$$
and to distinguish it from the 'conjugate' superposition state
$$
|\text{blue marble}⟩ - |\text{red marble}⟩,
$$
where that $-$ sign is a truly new, fully quantum-mechanical ontological feature that doesn't exist in classical mechanics.
OK, so it's not classical pre-determination. But is it faster-than-light communication, then?
Well... also no, this time because of another theorem of quantum mechanics - the No-Communication Theorem. This theorem states that, if you have any arbitrary entangled states linking two parties, and you allow them to perform any arbitrary physical operation on them, it is provably impossible to transmit a message in any way other than classical communication.
But that's at the level of the (ostensibly human) operators of the experiment, though. Do the particles themselves communicate?
And the answer here is that yes, it's not inconceivable that they do. More to the point, it is a consistent interpretation of quantum mechanics to suppose that there is a hidden-variable theory that underpins QM. Since Bell's theorem rules out hidden-variable theories which are 'local' and 'realist', the price for doing so is that those variables need to be 'non-local', which basically means that changes in those variables can propagate at FTL speeds.
However, the price for that is that the No-Communication Theorem requires that, to be consistent with QM, such a hidden-variable theory needs to somehow make that FTL communication completely inaccessible above the quantum-mechanical layer of the description. And that then means that, when you build those theories, they come out looking extremely contrived and artificial.
And here is where it gets subjective: most working quantum mechanicists are extremely uncomfortable with those theories as models of what reality is "really like". There are very good reasons to be uncomfortable with them, but $-$ as in all things where the interpretations of quantum mechanics are concerned $-$ this is ultimately a subjective thing.
Now as for the bulk of the text that you've written $-$
Normally this discussion goes nowhere and someone inevitably says "Well you just have to learn quantum mechanics." I always find this answer deeply unsatisfying.
Your observation is correct: this is indeed unsatisfying. But the hard truth is that, while it would be wonderful to know what's "really going on" between two entangled particles, we simply don't know. There's a broad array of proposals of how one should understand this, but they all have serious drawbacks and for each of them there's multiple reasons to think that it's completely bonkers.
We do know a lot about entanglement:
- we know how to formulate the quantum-mechanical laws that define it, and which are unbeaten in their ability to match experiments
- we know how to operate on those quantum-mechanical laws and the concepts within them
- we know how to perform experiments that are able to probe those laws and their differences to more 'reasonable' theories
- we know for sure that it goes well beyond 'classical pre-determination', and we have verified that experimentally multiple times
- we also know for sure that it does not allow us, as operators, to use it for FTL communication
- we also know that if it does involve some "back-end" FTL communication between the systems, then that requires additional "protective" layers in the theory that can fairly be described as completely bonkers
- (but then again, every way to interpret QM does things that can fairly be described as completely bonkers)
But as to what's "really going on", we just don't know.
Judging from your text, it sounds like you already have a pretty decent understanding of the epistemology around entanglement, and that you've reached the kind of grounded frustration at the weirdness of the theory that's widely shared among working physicists. We know it's weird. We agree that it ultimately makes very little sense. We absolutely would like a better answer. We are actively searching for better answers, and we are making decent progress $-$ we are indeed advancing the state of the theory, bit by bit, and we're getting better and better at examining single quantum systems in ways that allow us to test QM in finer and finer ways. But we've yet to find that better answer.
Best Answer
Going off of WillO's answer, while this scheme would work it would be no more effective than using a printer and two pieces of paper. Yes, your scheme is different in that it involves quantum nonlocality, but nevertheless it does not constitute faster-than-light communication because no information is being transferred between the two leaders. Their respective observations are correlated, but are nevertheless random. Hence, there's no problem. Is it weird? Yes. Is it a threat to causality? No. :)