In principle, could the Many-Worlds Interpretation be tested via interference?

Question

My understanding is that the Many-Worlds interpretation essentially says that the entire universe is one big wavefunction, and what we interpret as "measurement" or "collapse" is just when our conscious mind splits into a particular branch. I've also heard that this is just an interpretation, and makes no testable predictions to differentiate it from, say, the Copenhagen interpretation.

But if the universe is one big wavefunction, shouldn't some of the branches be able to interfere with each other via quantum interference? Would this be a prediction that contradicts the predictions of other interpretations?

For example, under Many-Worlds, could we, in principle, put a human in some kind of capsule and send them through some version of a double-slit experiment? My understanding is that Copenhagen would predict that capsule would always go through one slit or the other because the human inside is observing it, but under Many-Worlds we'd see interference.

Obviously, there's a lot of practical obstructions to doing this, but I'm wondering if there's something I'm missing with the principle itself.

Answer 1

It is a feature of all interpretations of QM that they are indistinguishable from one another. “Many-worlds” is no different from “Copenhagen” except the words they use to describe what’s happening; they make the exact same predictions.

The few circumstances where many-worlds claims to be different from more-standard theories, e.g. quantum suicide and immortality rely on assumptions like “conscious observers are actually a quantum mechanical thing”, which stems out of taking the word “observer” too literally when taking QM/SR/GR classes. If many-worlds actually offered a prediction different from Copenhagen’s, and if we could conduct such an experiment, it would already have been done.

On that note:

For example, under Many-Worlds, could we, in principle, put a human in some kind of capsule and send them through a double-slit experiment? My understanding is that Copenhagen would predict that capsule would always go through one slit or the other because the human inside is observing it, but under Many-Worlds we'd see interference.

“Observer” does not equal “human”. An observation is an event whence the quantum-mechanical “information” about a particular property of a particle is released, not when a human looks at it or knows about it or something. This is a common mistake for newcomers to QM.

Answer 2

In principle, the answer to the question is yes. What you are describing is a version of the so called Wigner's friend experiment, which was used to argue a similar point. See in particular this paragraph from Wikipedia:

Several authors, including Everett, John Archibald Wheeler and David Deutsch, call many-worlds a theory or metatheory, rather than just an interpretation.Everett argued that it was the "only completely coherent approach to explaining both the contents of quantum mechanics and the appearance of the world." Deutsch dismissed the idea that many-worlds is an "interpretation", saying that to call it an interpretation "is like talking about dinosaurs as an 'interpretation' of fossil records."

Note however that observing interference between macroscopic systems is extremely unpractical. For example, one will have first to calculate the full wave function of the macroscopic system with all of its degrees of freedom. For comparison, representing on a computer the full wave function of a single iron atom with its 26 electrons, will require something like $10^{78}$ bits (3 spatial degrees of freedom per electron), which is only slightly less then the total number of protons in the universe.

Another caveat is that the possibility of interference between macroscopic systems might not be unique to MWI. There are probably as many different versions of each interpretation of QM as the number of physicists practicing the theory!

Answer 3

The many worlds interpretation (MWI) is what you get if you take the equations of motion of quantum theory seriously as a description of how the world actually works and work out their implications as you would for any other scientific theory.

In classical physics the equations of motion for the $x$ position of a particle are written in terms of a function $x(t)$ such that if you measure $x$ at time $t$, you get the answer $x(t)$.

In quantum physics the equations of motion are written in terms of matrices called observables. The eigenvalues of those observables are the possible results of measuring the relevant quantity. Quantum physics predicts the probability of each of those possible outcomes.

In many experiments, the result of an experiment depends on what happens to all of the possible values of the relevant observable: this is called quantum interference. For an example see Section 2 of this paper

https://arxiv.org/abs/math/9911150

When you walk through a doorway you don't have to take account of all of the possible routes through the doorway. A commonly stated explanation for this fact is that somehow all of the possible values of the observables describing your trajectory vanish except the one you see: this is called collapse. Collapse is incompatible with the equations of motion of quantum theory. Some physicists have tried to modify the equations of motion of quantum theory to include collapse:

https://arxiv.org/abs/2310.14969

It is more common in textbooks to simply state that collapse happens and to give no explanation of how it happens. It's difficult to test a theory like this because it is extremely vague.

Quantum theory without collapse models measurements as interactions that produce records of some property of the measured system. A record is a piece of information that can be copied and copying information out of a quantum system suppresses interference. This is called decoherence:

https://arxiv.org/abs/1911.06282

Any object you can see in everyday life is undergoing such interactions on scales of space and time smaller than those over which they change significantly, e.g. - light reflecting off an object conveys information about its position to other systems, as does the pressure an object exerts on whatever it is resting on and so on. As a result of decoherence such objects tend to obey the laws of classical physics to a good approximation.

Decoherence doesn't eliminate the other values of the monitored observable, it just prevents interference between them. As a result quantum theory describes a reality in which all of the systems around you exist in multiple versions that form layers each of which acts approximately like the world as described by classical physics:

https://arxiv.org/abs/1111.2189

https://arxiv.org/abs/quant-ph/0104033

Hence the name of the MWI.

You couldn't interfere multiple versions of a human being because we're extremely large complicated sacks of chemicals that have to constantly exchange material with the outside world to stay alive, e.g. - breathing, heat exchange etc.

In Section 8 of his 1985 paper "Quantum theory as a universal physical theory" David Deutsch gives an account of an experiment where an AI implemented on a quantum computer could do an interference experiment on himself and know that he was in an unsharp state during that experiment:

https://boulderschool.yale.edu/sites/default/files/files/Deutsch.pdf

Deutsch interprets this as a test of the many worlds interpretation (MWI) versus the Copenhagen interpretation (CI) since the CI wouldn't allow an observer to undergo interference and the MWI would.

There are a lot of Wigner's friend type experiments on the quantum theory of observers such as the Frauchiger and Renner paper:

https://arxiv.org/abs/1604.07422

and there are other more recent papers along similar lines:

https://arxiv.org/abs/2209.06236

https://arxiv.org/abs/2407.06279

More generally the MWI is testable in the same way as other scientific theories:

https://arxiv.org/abs/1508.02048

I should also note that the idea of collapse isn't a good fit for the description of many kinds of quantum measurements, such as repeated, continuous and unsharp measurements:

https://arxiv.org/abs/1604.05973

Collapse theories (and pilot wave theories) also don't reproduce the predictions of relativistic quantum theories:

https://arxiv.org/abs/2205.00568

In practice we don't have to do exotic experiments involving interfering AIs or anything like that. The other interpretations are underdeveloped modifications of quantum theory and should be understood in that light.