None, it is an interpretation.

The Many Worlds interpretation has been introduced to avoid the puzzling collapse of a quantum state induced by the measurement that is at the basis of the Copenhagen interpretation.

The Many Worlds interpretation postulates a single quantum state of the world that evolves continuously. When an experiment is performed, there is no collapse of a quantum state, as said by the Copenhagen school, but it happens that the world bifurcates in to two separate braches each of them is a new world. The observer belongs to one of the branches and cannot know or say anything about the other branch.

Besides the effort required to my imagination, on my opinion the only possible, (negative), implication of the Many Worlds interpretation, is the consequence that we cannot use the Born Rule to evaluate the measurable probability of a given quantum state

Thanks Dr.Gian Bentini for your clarification. My doubt is that if MWI is going to be correct, then we do not require a conscious observer to collapse a wave function, isn't it? In that case, the profound philosophical implication could be that "the moon does not disperse into a mere probability function, regardless of an observer looking at it or not". This again takes us back to the Newtonian mechanistic interpretation of the universe, wherein the universe is out there undisturbed by the presence of the observer. I somehow could not digest this mechanistic interpretation of the universe, though physicists like the universe to be that way (objective way!). There are quiet some good theories such as "Biocentric Universe" have been formulated on the basis of Copenhagen way of thinking. Can you comment on these aspects?

I have to disagree with the position that if MWI is found to be true that there would be no implication.At base an interpretation of quantum mechanics is a statement about the nature of reality.The presence of branching universes will have deep seated philosophical and physical implications.

I am of the opinion that considering Heisenberg's uncertainty principle strict determinism has to be rejected.I don.t think classical laplacian determinism will ever berealised in a quantum system.

In my opinion, the many worlds interpretation is just a smoke screen of many words which hides the interesting questions: how to reconcile the deterministic evolution of the wave function of a closed quantum system according to Schrödinger's equation with the random branching of a quantum system interacting with the macroscopic outside world according to Born's law (von Neumann collapse). Just saying that all the branches are all equally real does not help. We need to know the probability of different branches, and we need to know the laws of physics as they hold at the moment on the branch we are living on right now. There have been many attempts to *derive* Born's law inside of a strict many-worlds approach but as far as I can see they are all either flawed or circular. Many of them fall into the "FAPP-trap" - "for all practical purposes". Practically minded physicists are not bothered by logical consistency of the mathematical model(s) they are using, and as long as they have a tool box of approaches and experience ("physical intuition") to choose the right tools to get the right answer, they are happy. This comes down to "treating a bug as a feature". The Schrödinger cat paradox shows that using the standard quantum formalism in different ways gives different answers. For someone concerned with logical consistency, which is more or less the possibility to imagine a "model universe" which actually runs according to some clear set of mathematical rules and reproduces the features of our own which we are interested in, this possibility to get several contradictory answers at the same time is a clear sign that the foundations of QM are not in order yet. But a practical minded person doesn't worry since they are only interested in getting the right answer (and they know by physical intuition how to get that).

I think that Slava Belavkin's eventum mechanics provides a consistent and elegant mathematical solution to these problems. It treats the random branching as real and it treats "reality" as real. Wave functions are just an intermediary step: they tell us the probability of different future possibilities, given a particular past actuality.

In my opinion, there are no *observable* consequences of the validity of the Many-Worlds Interpretation, and this is an important feature of this intepretation, which allows it to stand. Indeed, if it did make predictions different from that of conventional quantum mechanics, these predictions would easily be shown to be false. The reason it makes no different predictions is that decoherence effectively makes the overlap of two macroscopically different states extremely small, so we cannot experimentally distinguish between the coherent superposition of two such states from the incoherent one.

As for Born's law, it is clear that it cannot be an exact law within a purely quantum mechanical framework. That, however, is not necessarily a problem. It makes no sense within a single universal wave function to ask which branch has what probability. But as humans, who always observe correlations within one part of this wave function, an approximate law of Born type can indeed be shown to hold. And this, I believe, is what we need to account for observations, which again is what physics is about. Of course,this does not mean that the Many-worlds intepretation is "true", whatever that means. But I personally have found it useful to understand, for example, such issues as non-locality in experiments involving Bell inequalities.

As has been stated above, the canonical view is that since MW (Many Worlds) theory has been shown to be equivalent to standard QM, then there can be no discrepancy. Historically, however, it is claimed that the desire to find testable differences is what led Deutsch to start looking at quantum computing.

Hence to address this question in a meaningful way we need to look towards extensions to either non-MW QM, or MW.

The first point to stress is that the Universes in MW need not be actual 'universes', and there is therefore no issue about non-conservation of mass. Instead the universes should be thought of as essentially states 'tagged' by a set of parameters, and when the parameters are distinct enough, then interference can no longer occur.

Let me give a concrete example to illustrate what I believe is the important observational effect we want to explain.

The big question is why don't we see macroscopic interference, or more precisely macroscopic quantum effects? We postulate in canonical Copenhagen (and indeed many other approaches) some kind of measurement process, but in fact this is IMO metaphysics as we have no physical description of what constitutes a measurement, merely the consequences of measurement (which are well-understood and entirely testable, especially in such approaches as weak measurement).

One of the standard interpretations behind MW is that it is in fact entanglement that prevents this interference. So let me clarify.

Consider an interferometer. A particle travels through this interferometer, where there is an initial splitter that allows the particle to follow either of the two paths, but provides no information as to which path it has travelled. Then there is another splitter to recombine the paths, and the final path of the particle determined by interference.

To keep things simple, we choose to align the interferometer so that there is a bright output and a dark output. This means that if the particle showed interference, then every time the particle is measured, it will be found at the bright output. Therefore (ignoring noise) a click in the dark output shows that which-path information has been gleaned (not that a click in the bright output does NOT negate the possibility that a measurement has occurred).

In the no-measurement case, this is easy and we retrieve quantum mechanics. So let's make the system more complicated.

Imagine that there is a single atom in the system that can measure if the particle has flown past it via some interaction, but where it does not destroy the particle. So we have which path information now. This means the interference signature must be destroyed (it is) and we will find that 50% of the particles are now detected at the dark port.

But a single atom is still a quantum object, isn't it? Of course it is. So if it is possible to erase this information, for example by measuring the atom in some state that provides no information about whether or not the particle has passed it by, then the interference should be retrieved.

This experiment is called quantum erasure and it works. We can understand this problem as being the result of entanglement my writing down the problem mathematically.

Consider two states for the particle, |0> and |1>, which correspond to the two paths of the interferometer. The atom has the states |a>, |b> , and we suppose that it only interacts with a particle in mode |0>.

First, no atom, particle in mode |0>

|0> -> (1/\sqrt{2})(|0> + |1>)

Second splitter performs the operation

(1/\sqrt{2})(|0> + |1>) -> |0>

Note that here |0> is our bright port, and |1> is the dark port.

The system is written initially as |0>|a> (particle in mode |0>, atom in |a>). The splitter takes this to

(1/\sqrt{2})(|0> + |1>)|a>

This can be rewritten as

(1/\sqrt{2})(|0>|a> + |1>|a>)

The interaction takes this to (this is the entangling part)

(1/\sqrt{2})(|0>|a> + |1>|b>)

The final splitter then gives us

(1/2)(|0>|a> + |1>|a> + |0>|b> - |1>|b>).

I cannot cancel the |1>|a> and the |1>|b> because they are different states, and so if I just look at the particle, now I have a 50% probability of detecting the particle at |1>.

However, if I can erase the information, for example by measuring the atom in the basis

(1/\sqrt{2})(|a> \pm |b>), then the interference is recovered (albeit up to some random sign correction which must be known to prove the interference).

Now, many worlds is simply the extension of the above example to many atoms. The measurement interaction is simply the result of many such entanglements, and this tells you that pretty soon, as the number of particles increases, the amount of entanglement gets so large that erasure becomes practically (although not theoretically) impossible. This mechanism breaks the interference and gives us functionally, distinct worlds.

So where would I look for exceptions?

Well the obvious one is to look for intrinsic decoherence, or some other mechanism that enforces measurement.

Another possibility is that the total amount of entanglement might be bounded, so that in a finite system, non-classical interference might be unavoidable. (It is amusing to think that bounding entanglement might actually make bizarre quantum effects more manifest).

But at present, we really can't distinguish between the different cases as the interpretations are equivalent.