I have a slight problem with it… are the many worlds well defined?
Here's the crux. The quantum state of the universe is a vector in Hilbert space which evolves deterministically.
That is a completely empty sense if there is no example available that demonstrates how that is done.
Even this sentence: "The physical state of the universe evolves deterministically" because there exist no clear defintion of the words deterministic and indeterministic.
We don't even need to go to the extent of considering an entire universe. Any closed quantum system can be represented as a vector in a Hilbert space.
How is that be done?
However, a vector can be decomposed into projections on a measurement basis. If the projection requires more than a single basis state, the state vector can be considered to be in a superposition.
What does that physical mean?
A measurement will select a particular basis state with a certain probability. Each selected state corresponds to a “world". However, I can choose to make a different measurement. Then the entire basis is different. In fact I can choose a basis aligned with the state vector. Then there can be only a single measurement result. So I, as the observer can arbitrarily change the many world mapping?! What about all the other observers? What are these observers anyway?
This answers leave the observers baffled behind.
If these many worlds were fixed and immutable, then I'd feel a little easier. However, we seem to have agency in determining them…
I am much happier with the state vector being representative of an underlying reality. Then no matter what measurements are made, they only serve to probe this vector. That makes the worlds only something apparent, not real. Then the many worlds just represent measurement possibilities of something well-determined.
Also how that is done should be explained in detail. Being not, this makes this whole discussion rather meaningless.
I could be way off with this criticism, but maybe this will serve to clear up some confusion among a more lay audience. It seems that many MWI adherents actually do consider the state vector as the most important aspect of the theory (Everett certainly did). It would then just be an unfortunate event that the term “many worlds" interpretation was coined.
So going with these worlds only being apparent, and representative of something definite, we arrive at a far more reasonable interpretation of things. We really don't have a multiverse, but rather something intangible that we only probe with measurements. In that sense, all possible measurement results are required to specify the quantum state.
Does that mean that the writer of this answer, rejects the Many World interpretation?
I find this far more reasonable. However, it moves the ontology to the state vector. Some people won't like that. Furthermore, the role of the observer is still ill-defined. In other words, this interpretation still suffers from the measurement problem.