Suppose Alice and Bob perform a test of a Bell inequality. They start together and produce an entangled quantum state. Then the parties move apart such that when they make their measurements on the entangle quantum state they are spacelike separated and are at rest with respect to each other’s motion. Let’s suppose that Alice and Bob perform their measurements at the same time (simultaneous) according to the two party’s rest frames. Now the standard paradox arises: the correlations Alice and Bob produce cannot be explained by a local hidden variable model: one party needs to produce an outcome which is predicated on the measurement choosen by the other party. So this seems very bad: if Alice sends a signal to Bob at the speed of light, immediately as she does her measurement, then the signal will arive at Bob at a time equal to the distance between the parties divided by the speed of light. But now look at exactly just this scenario from the perspective of a party who is traveling in a reference frame which is moving with a velocity in the same direction as the separation of Alice and Bob. If this frame is moving at nearly the speed of light with respect to rest frame of the parties, then the paradox seems almost triffling: in this frame of reference, Bob would get the information immediately after he would normally need it. Extrapolating as this party’s frame’s velocity goes to the speed of light and thinking that there is a limit to how sharply we can define when a measurement occurs, we see that “There exists a reference frame in which the paradox of Bell violation for all practical purposes disappears.”
What does this tell us? Well it certainly doesn’t resolve the Bell paradox. One thing to get is that the paradox is reference dependent. OK, not too interesting either. But what I think is neat about this thought experiment is it gives us a point of little or no paradox upon which we can begin to think about how to bootstrap our way into the reference frame where there is a very measurable Bell paradox. How do we do this? Heck if I know: the point is mostly that the solution should be a SMOOTH map from this little or no paradox reference frame to the frame where the information propogation is ridiculous.