Circular. You have to already accept many worlds to take this as evidence of worlds. You can argue for Many-Worlds on grounds of parsimony and clarity, but that's it.

No I don't think so. Many worlds is basically the hypothesis that we can put arbitrarily large systems in superposition. Like schroedingers cat that is both dead and alive, but instead it could be the whole lab, the whole planet, the whole galaxy that is simultaneously in two different states. We could test this by trying to actually do it, by putting really large and complex systems in superposition. If they both evolve as we'd predict and cause the interference patterns we predict, then we must conclude that these "other-labs", "other-planets", "other-galaxies", "other-worlds" evolve just like ours, and have a causal effect on ours.

Saying that sure, they evolve like ours, and sure they affect ours, but even so they are not real - that is I suppose one stance, but then you are getting pretty close to solipsism.

> Saying that sure, they evolve like ours, and sure they effect ours, but even so they are not real - that is I suppose one stance, but then you are getting pretty close to solipsism.

No, you'd get something like Bohmian mechanics, or Rovelli's relational interpretation. Which is my point: you can only take this as evidence of other worlds if you basically smuggle those assumptions in.

> We could test this by trying to actually do it, by putting really large and complex systems in superposition.

Is there any hope that one could actually do this? The article mentions that large systems may be self-observing (via gravity).

I thought there's a quasi-classical approximation of quantum gravity, it doesn't work only for event horizon due to infinities there, but should work fine for diffuse matter, like how most calculations for electron orbitals use classical electromagnetic field.