Wigner friend experiment

Question

Let's supposed we take the Wigner's friend experiment from the metaphysical arena and try to implement it as an actual physical experiment

Assuming Wigner's lab friend is kept as a coherent superposition of states until the external experimenter tries to open the decoherence-insulated lab door, then the following check to detect coherence should work, even after the enclosed physicists has interacted with the dead/alive cat, but before the sealed lab is allowed to interact with the external experimenter:

• the sealed lab has two lasers (A and B) inside, that mix at a half-mirror, and then the laser goes out of the lab from the half-mirror. This is the only way the state in the inner lab can affect the external universe. If both lasers are on, only one of the external paths has constructive interference, while the other has full destructive interference.
• if the cat is dead, the sealed lab experimenter turns on laser A and keeps B off. Both branches of the outward paths of the half-mirror give half intensities when the single laser A is on while laser B is off
• if the cat is alive, the sealed lab experimenter turns on laser B and keeps A off. Both branches of the outward paths of the half-mirror give half intensities when the single laser B is on while laser A is off
• After some synchronization time, the external experimenter measures photons from each outward path of the half-mirror, and he sees two possible outcomes:

1) one of the paths has 100% intensity, the other has 0% (coherence is kept)

2) both paths have 50% intensity (coherence was lost previously)

Thoughts? What you expect to be the result of this experiment? why?

Answer 1

As long as the external experimenter does not attempt to measure the phase of each of the arms - and indeed makes a measurement that does not 'touch' this information - he will observe no contradiction. In each of the definite outcomes of the cat experiment inside the lab, the experimenter inside the lab signals this definiteness by outputting 50/50 intensities out of the lab's two windows.

He also signals out what the outcome was, in the phase of the light, but the external experimenter does not look at this signal. (Whatever pieces of equipment are sensitive to this phase will, of course, be entangled with the inside of the lab.)

Thus, all that the external experimenter can tell is that a definite outcome was obtained inside the lab, by virtue of the light being 50/50 over both windows.