My thinking on this is basically, hypotheses don't have to be falsifiable, but they aren't very useful if they aren't, because you can't design an experiment to

My thinking on this is basically, hypotheses don't have to be falsifiable, but they aren't very useful if they aren't, because you can't design an experiment to test them. Neither the determinism nor nondeterminism hypotheses are falsifiable, so there's no reason to prefer one over the other right now. However, non-determinism will never be falsifiable, whereas determinism is only currently non-falsifiable because a plausible, testable mechanism has not yet been hypothesized. So while there's no reason to favor determinism as something we believe, there's reason to favor it in research, because research has the potential to yield further information.