The rock pattern universe is close to being a reductio ad absurdum for the simulation concept. There's no reason to suppose that placing four rocks in a row creates a sixteen state universe. They're just rocks. There's no other world. No matter how many rocks, how complex the pattern, there are no invisible people living in the rocks.
The possibility of placing four rocks in a row creates (if making certain assumptions) a sixteen state
system. Whether you call it a "universe" or not is a matter of taste.
Saying "another world" is just a metaphor for an encoding. And being "in that world" just means being equipped with a means for decoding (perceiving) the phenomena in question in a system so encoded. Decoding/perception is, of course, itself a computation.
If your four rocks encode a four-bit binary number, then knowing how to read that binary number is the difference between seeing four rocks, and seeing the number 11.
Everything is encoded, in reality as well as in a simulation. We do not perceive any aspect of the real world directly. A graph of the variations in air pressure impinging on your ear looks like a nearly random wavy line, but your ears and your brain perform the computations to perceive three voices, a radio playing an ad jingle, a dog barking, and a truck driving by outside, all at the same time. An alien who had only a pressure sensor, with no algorithm for hearing, would be no more able to perceive those things than you would be able to recognize the ships, winds, currents, and sea floor shape of a bay by watching the changing height of the wavy water surface at the end of a dock.
Program a three-body system as a Turing machine, then code that Turing machine as input to a universal Turing machine, then simulate that universal Turing machine and its inputs in a cyclic tag system, then simulate that cyclic tag system in a Rule 110 cellular automaton, implement it as a row of rocks in a desert with a guy adding rows, and let it run. You'd see a lot of rocks, and not one of them, nor even one contiguous cluster of them, that you could point to and say, "this is the current z velocity of the second of the three bodies in the three body system." Nonetheless, if you follow the layers of encoding (which just requires the right computation), you can perceive the changing states of three bodies. We say the three bodies are "still in there" but they are not actually "in" the rocks, they're represented by certain characteristics of the pattern of rocks that can be perceived by computation.
Very likely, you will object to these examples being comparable to one another because in the case of hearing, the radio you hear is "really there in the real physical universe," while in the three-body simulation, there are no actual three moving bodies anywhere, it was a simulation from the get-go. This is a philosophical point that will not be resolved, but the argument from so-called computationalism is that the only meaningful operational definition of "really there" is something like "can be reliably perceived by performing the appropriate measurements and computation," and that the same could just as well be said for entities in a simulation being able to reliably perceive other entities in the same simulation. Equivalently, it's often pointed out that it is not possible to determine whether or not the real universe is a simulation, and that the question might not even be meaningful.
This reminds me how absurd another thread became, where it was claimed by computationalists that enough people writing enough 1's and 0's on paper could produce consciousness.
Almost as absurd as claiming that enough ions flowing through channels at connections between the cells in a big hunk of tissue could produce consciousness. (But not quite. At least writing numbers on a piece of paper is a method of computation that's been known and understood for millennia. So it's not quite as silly as claiming it's possible to compute -- let alone think -- with meat.)
Respectfully,
Myriad
