I sort of think of it as everything having angular momentum. You could deduce magnetic fields, mass, and gravitational pulls by looking at how things are rotating with one another. Quantum particles seem to be a special case of this, where the gravitational pulls are about equal and have a tendency to act like gears; so if one spins counter-clockwise, then the other must be clockwise, otherwise the gears will grind (a point of friction) and tend to re-orientate to a position with the least amount of friction as possible - having opposite spin.
So QE says that when you separate the particles, they will tend toward having opposite spin from each other, just as they did when they were together. Though of course, if you move particles away from each other, then they also have less of an effect on each other, and are effected more by their new surroundings than each other. So separate them far enough apart with enough time and theoretically they will not have opposite spin.
What's interesting is that Einstein's relativity doesn't seem to account for this, though conceptually it probably does since a changing EM field relates to a changing spin (either in space or speed) and would still fall under light being a reflection of how particles affect and change each other.