I've been thinking about this sort of thing.<p>It seems more reasonable to me that the act of measurement actually does not change the state of the quantum system, but entangles the observer with the system.<p>It is the observer and not the pre-existing entanglement that is subject to change as a result of measurement.<p>If a man in a box at some stage measures the state of a cat in a box, then my perspective of the box is a superposition of:
a man who has not yet measured the state of the cat;
a man who measured the state of the cat, and found it alive;
and a man who measured the cat and found it dead.<p>I can expand this superposition with arbitrary possibilities - such as whether or not the cat is yet to enter rigor mortis.<p>If without seeing the cat myself - or the man's reaction to seeing the cat - I watch as the man looks at the cat, my superposition is reduced to the possibilities in which the man has measured the cat.<p>This is not about the state of reality - it's about the localisation of information regarding the state of reality.<p>(Additionally, if I saw the man's reaction to his observation of the cat it would skew my superposition based on the likelihoods that he would react that way if the cat were dead/alive).<p>Spooky action at a distance is another one where I don't think the common interpretation holds up - perspectives on quantum systems should be absolutely free to contradict each other (in terms of absolute measurement) it is only when the light cones of the quantum systems meet that contradictions are excluded.