> But Leibniz's idea was more like that there is an arbitrarily large number of computers running the same simulation, while some of them also run artificial intelligences which just watch the simulation unfold.
I think this was just Leibniz awkward attempt at explaining modal realism. The idea of many possible worlds is much more mainstream now in physics, so I wouldn't call it "completely nuts".
I wasn't previously acquainted with the term "modal realism" (been avoiding contemporary philosophy for a while now), but there doesn't seem to be an immediate connection in this case. There aren't actually many worlds being simulated here, as each computer simulates the same world and they all get the same results. Leibniz does introduce the idea of many possible worlds (§53), but as a set of designs for this whole thing (computers plus the stuff running inside them) from which God draws one particular design to be the actual world.
He might be closer to modal realists in other regards, though. I don't know.
I think this was just Leibniz awkward attempt at explaining modal realism. The idea of many possible worlds is much more mainstream now in physics, so I wouldn't call it "completely nuts".