Definition and Rules of Quantum State Observation

[Jarvis323]

I was wondering how the rules work for observation in a quantum system. Particularly, about what happens if two separate entities try measuring at the same time. And also, what kinds of interactions are happening all the time that are considered measurements, for example in quantum thermodynamics, or quantum gravity, is the state considered to be measured as part of the process? And would this be discrete measurements, e.g. of bunch of different entities requesting measurement throughout time? Or more a continuous measurement of the quantum system, or one that is somehow (shared) so to speak (distributed continuously in some sense)? Would there be a rate at which distinct measurements can occur? Like say multiple entities want to measure, they must do so one at a time, and then it would take longer for all to complete than one? If so, is this plausible as some kind of basis for space time curvature, or is this consideration part of some other unified theory?

Hope these questions are not too far off base, my background is not in physics. Thanks

 

[Strilanc]

"Measurement" is a high-level abstraction on top of decoherence, which is a continuous phenomena where information about the system gradually makes its way into the environment. When you impose a measurement onto a system, you are attempting to force specific information to leak out quickly.

If multiple parties are all trying to impose different measurements, the most likely outcome is that they all trip over each others' toes. Pragmatically, this would be due to obvious stuff like one party wanting to put a beamsplitter before the photon counter while another party wants there to be no beamsplitter. More abstractly, you could suppose each party is able to add dynamics into the system and the total dynamics is the sum of all the contributions (i.e. they each specify a Hamiltonian and you just add up the Hamiltonians). The resulting process would be well-defined, but probably wouldn't do what anyone wanted. For example, if you take a Hamiltonian for measuring horizontal-vs-vertical polarization and just add it into a Hamiltonian for measuring clockwise-vs-counterclock polarization, you don't get a Hamiltonian that causes both to be measured simultaneously. You get something that measures upleft-vs-upright polarization (or other stuff; it depends on the details).