To communicate you would need either of two possibilities (let’s overlook the fact that there may be no one to communicate with):
1: A telephone. Okay, maybe not a telephone, but something that could interpret the signals (whatever) we send and make it intelligible at their end (again I’m assuming they are there and they’re intelligent enough to build, let alone operate a telephone).
2: A bomb. Again, I’m being cute, but something that we can use to influence they’re environment in such a way that they can interpret as intelligible. Although, on the surface this may appear to be similar to option one, in that interpretation is involved, a bomb is a little more like a one-sided conversation, wherein, it is hoped they can’t communicate back.
Personally, I opt for the 2nd option. You get lots of free land as a side effect.
Some supporters of the many-worlds interpretation (as on this FAQ) would say that the existence of multiple worlds can be derived from the mathematical formalism of quantum mechanics, and that the only metaphysical assumption they add is to take that formalism seriously. In reality, the situation is more complicated: it’s more accurate to say that if the interpretation is true, it must in principle be derivable from the mathematical formalism.
Assuming the equations we use to describe quantum mechanics are accurate, this means the interpretation is testable, at least in a weak way: it must be possible given sufficient computing power (or, more realistically, knowledge of sufficiently good approximations) to demonstrate via simulation that quantum systems brought into contact with a complex environment that is also modeled quantum-mechanically either do or do not cause that environment to split into different “worlds.”
A significant amount of work has been done along those lines in the past 20 years. The phenomenon of “decoherence” is a recognized physical effect that was first discovered in the theoretical study of quantum computers. Roughly speaking, it refers to how the wave function of a system, when interacting with a much larger system, can split into distinct components that no longer interfere quantum-mechanically with one another in a measurable way. This causes quantum computation to fail, and it is also considered to be a physical description of the dynamics via which worlds can “split.”
The much harder part is to show that these distinct branches of the wave function will under normal conditions resemble classical worlds, but recently there has been great progress along these lines. This paper is an excellent (but pretty technical) summary of the present state of our understanding. It describes the effect of decoherence from the perspective of several different interpretations, including the MWI and others. From the abstract:
I really must call bullshit here. The predictions of observations are entirely within the scope of the mathematical structure of the theory, and MWI makes no different predictions than the old-school CI. The only difference is pure metaphysics. Try reading, say, anything by Jeff Bub.
Decoherence is perfectly explicable by an addition (which has been done) to basic axiomatic QM. In particular, it’s explicable without reference to worlds “splitting off”. To say that worlds split over decoherence makes the argument no more convincing than to say they split over the Von Neumann nonunitary evolution. Occam’s Razor makes quick work of all those stubbly alternate worlds.