At the request of @grokkingStuff, a pontification thread about MWI.
First, my (lack of) credentials: QM, adv classical mech, E&M in undergrad, many books (pop- and text-) and papers, focus on foundations.
So, more than a typical interested layman, but less than a curious physics BS with an interest in foundations.
The good: MWI is one of two obvious ways to take the Copenhagen "interpretation" and turn it into a real ontology without major changes.
(the other being de Broglie–Bohm) (obvious is meant loosely, Everett deserves a lot of respect for coming up with it)
Drop the baggage of wavefunction collapse and complementarity and Schrödinger evolve the wavefunction forever. Beautiful. Simple.
("the" wavefunction? yeah, there are some deep issues there, but we don't even need to touch them)
The major problem here is that with *only* the wavefunction, there's no longer any way to connect the system empirical results.
This is what that diagram is about: There's nothing at all in 3+1D space in MWI. There's *only* the wavefunction in configuration space.
(What even is configuration space when there are no particles whose configurations we're representing, anyway???)
Without modifications, MWI is inescapably a *no*-worlds interpretation. It's coherent, unlike Copenhagen, but empirically inadequate.
(They acknowledge this when they insist that in MWI Bell's theorem doesn't imply non-locality. True, because there's no spacetime!)
Some claim (or assume) that splitting along the position basis falls out naturally. But it doesn't!
arxiv.org/abs/1210.8447
arxiv.org/abs/1210.8447
And even if it did, there's no real justification from there to the 3+1D world actually existing in some real sense.
If you take it as an additional postulate, you're back at de Broglie–Bohm.
The other alternative, which IMO is actually more respectable, is many minds or something like it.
Basically, the existence of structure in configuration space isomorphic to brains in 3+1D space *causes* consciousness in that space.
This could in principle resolve the issue and not require us to admit the fundamental existence of 3+1D space.
At the low cost of requiring a solution to the hard-problem of consciousness, of course. And it's *definitely* an addition to base MWI.
A more technical problem, though still pretty embarrassing IMO, is that no one has been able to get the Born rule out of the theory.
Even accepting decomposition along the position basis, nothing actually requires us to consider the square of the amplitude as meaningful.
(The traditional examples obscure this, since you can get the relevant special case of the Born rule from symmetry considerations there)
This is a big deal because in many ways, deriving the Born rule is the whole *point* of MWI. If you have to postulate it, you've lost.
To wrap up: I get the appeal, and it's certainly an improvement over Copenhagen. But the elegance is at too high a cost.
It can be a useful framing to think about things in, and helps clarify some of the big issues. But as a serious interpretation, it fails.
