Steven Weinberg is no more, since recently. I did appreciate very much his books on quantum mechanics, and also his introduction to quantum field theory.
I have mentioned more than once his work showing that if you delinearise a little bit quantum mechanics, not only you cannot make the parallel worlds/histories disappearing, but somehow, it makes possible to visit those parallel worlds, or, to use an image by Weinberg, to call your doppelgänger with a phone.
This provides a quantum and dual way to refute John Clark argument against the first person indeterminacy in arithmetic. John argue dans le parallel histories in arithmetic allows in principle the doppelgänger to meet, so that it is different from the indeterminacy on the superposition. But the point is that the indeterminacy calculus cannot change based on such counterfactual, unless adding magic to Mechanism, but then Mechanism is false by definition. A dual counterexample based on this work by Weinberg is that delineairsing a little bit the Schroedinger equation, in such a way that the indeterminacy remains unchanged, makes the doppelgänger accessible, like with the classical duplication, and yet does not change the calculus of indeterminacy different, illustrating once again you need to add magic to Mechanism to avoid, like in quantum mechanics, the first person indeterminacy.
A more serious difficulty is to make people understand the original paper of Turing, Church, Post, which shows (along with Gödel) that the arithmetical reality is (more than) Turing complete. This follows from understanding arithmetic, or, at a more formal level, by understanding that all models of arithmetic have the same initials segment in which addition and multiplication stay Turing emulable (which is not the case in the whole non standard models. this requires a bit of mathematical logic, which is not well taught, when taught at all.
Once you grasp this, even without Mechanism, you can understand that the burden of the proof is in the hand of those who add some more axioms to arithmetic, like the existence of some "primitive matter" which have to justify its role in consciousness selection from arithmetic. In deductive theology, it is better to not add any ontological commitment before a reason is provided to it. Up to now, observation confirms mechanism. If they was one fact in favour of non mechanism, or in favour of something more than numbers, I would welcome it, but there are none, as far as I know. On the contrary, Everett QM confirms all prediction of classical Digital Mechanism, and explains furthermore the qualia and consciousness, as notion of knowledge imposed through self-reference and incompleteness.
Bruno