Take a black stone and a white stone, shake up in your hand, put one in one sealed container and one in a second sealed container. Take the two containers a hundred light year apart. Open one and "suddenly" you know the contents of another container light years away.
I don't know how "spooky action" differs from this except in the way the process gets described. You could say "wow, until you know whether there's a black stone or a white stone in there, it's like the stone is half white and half black." You could but how does that help? I mean quantum seems to take a "because it absolutely can't be measured till, it absolutely doesn't happen till later" but that seems to just contradict ordinary understanding in gratuitously unnecessarily fashion.
Because, when you put the white stone in one jar and the black stone in the other, it's decided. One of them does contain a white stone. One of them does contain a black stone. The universe "knows" which is which. The decision has been made.
With the entangled particles, it's not decided. You separate them, and you do NOT have one particle in "up" state and one particle in "down" state. That decision has not yet been made. It's not the case that one of them is "up" and you just haven't checked yet; one of them is not "up". One of them is not "down". The decision has not been made. One will be "up" and the other "down" when the decision is made, but that random decision has not yet been made.
You then move them a long way apart, and look at one of them; the decision is now made for BOTH of them, even though they're very far apart.
Here it is; because that's how the universe works.
That decision doesn't get made until "something" checks it/depends on it/interacts with it (the English language, at heart a way of telling macroscopic monkeys about tigers and bananas, does struggle a bit here). If you go digging yourself, you'll be able to go as deep as you like. I apologise for not presenting a better answer, but I'd really be doing you a disservice.
Edit: It's really not a good answer at all. I'm digging through old reddit threads to see if RobotRollCall ever talked about this. Her explanations were generally very accessible.
I asked the original question and I've harassed people with this question enough that I vaguely recall the answer.
In terms of the boxes, the answer is you wind-up breaking open the boxes over time in different ways that wind-up with being incapable with the perspective that "just" a black stone or "just" a white stone is in each box.
http://en.wikipedia.org/wiki/EPR_paradox#Mathematical_formul...
Unfortunately, too many years passed and I don't remember the counter-argument that explains why you can't apply that logic to entanglement but it involved Bell's theorem[1] and the EPR Paradox[2]. Actually I think Einstein himself made a counter-argument similar to that (just not with stones/balls but with actual particles) and was disproved later on. Sorry if I can't give more details, I'm not a physicist.
[1] https://en.wikipedia.org/wiki/Bell%27s_theorem [2] https://en.wikipedia.org/wiki/EPR_paradox
Bell's inequality paper in 1964 proved that the predictions of a local reality system and a quantum entanglement system are in fact different for certain polarization experiments.
Subsequent experimentation proved the two theories to be mutually exclusive and results back up FTL communication. An FTL mechanism has not been found experimentally, however we know that special relativity is not violated because quantum uncertainty prevents information from being transmitted via entanglement at FTL speeds.
You also don't just have two worlds, you have infinite. In the coin-flipping example, whenever a coin is flipped there are many other universe-splitting this-or-that events happening elsewhere in the universe at the same instant. So each instant there are infinite worlds created (or infinite x infinite to describe each combination), and again a split in each universe in the next instant for every next possibility. Mmm. Dodgy.
My favourite: John Cramer's Transactional Interpretation. Doesn't require infinite universes' worth of energy, doesn't require an observer to collapse multiple possible waves. Does require you to accept backwards-in-time communication (breaks causality) but brings some very interesting what-if questions.
edit: removed my question after reading joe_the_user's question.
So, "I flip a coin and whichever it comes up, heads or tails, the other entangled coin flipped at the same instant comes up the the opposite, with instantaneous communication" is the wrong view. Rather, the coins are entangled (in the past), and observing them pulls the observers into the entanglement one atom at a time as the informational contagion spreads, so now you have world A, where scientists saw "heads, tails" and world B where the same scientists saw "tails,heads". The scientists got slurped in, but they are simply observing an entanglement that existed and still exists and which was established in the past. Only now they are observing it from inside.
Quantum spooky action "at a distance" is local in "many worlds" and never needs to be faster than light.