Can a super-positioned human be used to differentiate between the Copenhagen interpretation and many-worlds? I'm pretty new to quantum physics, so there's a good chance that I have this all backwards.
However, by my understanding of the Copenhagen interpretation, the wave function 'collapses' upon observation. In the many worlds interpretation, the observer becomes decohered and collapse never occurs.
I understand that it's fairly difficult to super-position big things. However, if many-worlds holds correct, then it should be really easy to super-position a human. For example, they observe the result of an experiment. Upon seeing result A they take a step back. Upon seeing result B they take a step forward. Bam, we have a massive super-positioned object.
So the question is, is it possible to super-position a human such that they can interfere with themselves in some way? Or at least such that we can measure whether or not they are super-positioned?
For example, perhaps the super-positioned person carries a "photon gun", and shoots it at where his super-positioned self would be. He may not see anything special, but could it be set up such that an outside observer (unaware of where the super-positioned human is standing) could measure the interference between the photons shot by the photon gun?
If so, then it seems that there is an experiment which can differentiate between the Copenhagen interpretation and many-worlds. I've been told that this is not the case, so I'm wondering where my logic failed.
 A: If you believe quantum mechanics is a complete and accurate description of "reality", it is all too easy to set up a superposition of humans. As the OP mentioned, measure a superposed qubit, and if A, step forward. Otherwise, step backward. The problem lies with decoherence which prevents us from measuring any interference.
A: Of course you can superposite a human in theory.
But this will not help you to differentiate between MWI and Copenhagen: in both interpretations the expected result is the same.
A: Take a quantum supercomputer. Simulate a human in it, but don't keep any records or measurements of the simulation in any way. Get the human to form a superposition, then unsimulate by uncomputing. The human will have to interfere with itself. Do we end up with the same initial condition at the end of the unsimulation? If there are no records of the human, did the simulated human even exist? How come in the case where where we can actually interfere a human with itself, its existence is itself in question? Is it Orwellian with the human having existed and then vanished from the contents of existence into the metaphorical memory hole? Or is it Stalinesque with the human never ever in the contents of consciousness? Positivistically, can we ever tell the difference? Why is that? If a property exists, can that property simultaneously be in a superposition?
A: Anything in the past which is forgotten by having all memories and records of it scrambled irreversibly thermodynamically reverts back into a decohered superposition.
Turn around a minute ago and observe a fly resting on your bedstand. You record that fact in your short term memory. Turn around for a minute. Turn around again to see the fly resting on the window sill. Assume there are no other observers in the room, and no surveillance cameras either. When your back was turned around for a minute, the fly was in a decohered (by the environment, i.e. air molecules, photons, dust, etc) superposition over all likely trajectories it could have taken from the bedstand to the window sill according to the sum-over-histories formalism. It is true the ambient light, air molecules, dust etc. record the location of the fly, but the air molecules revert back to thermal equilibrium, the dust to thermal brownian motion, and the ambient photons get absorbed by the walls which then thermalize. We can't tell the difference anymore after even a very mild coarse graining.
When you slept last night, you tossed and turned, but being unconscious during your sleep, your brain stored no memories of your trajectory of tossing and turning. So, you took an uncollapsed sum-over-histories over all likely tossing and turning trajectories when you slept last night.
What did you eat for lunch 10 years ago on March 28th 2002? If you forgot, you were eating an uncollapsed sum-over-all the likely things you ate 10 years ago.
What was the name of your great, great, ... grandmother (matrilineal lineage) 200 generations ago in prehistory? It was in an uncollapsed sum-over-histories of likely names!
What are you doing right "now", whenever "now" is? If no one else is observing, and no one else but your brain is taking note, and later today, without having told anyone or written down what you did, you suddenly die of a heart attack, you would be doing an uncollapsed superposition of things right now.
Isn't it time to start keeping a journal, or writing an autobiography of yourself? ;-)
