I was being serious. Thanks for completely missing the point.
No, I got your point, because I was being serious, too.
The "man" who is "in the world of the simulation" can only be in my imagination because that's the only place such a world can be.
The state-changes in the computer are real... they are what they are... but they are only that... and you can look at the computer and what it's doing and see that there is no man.
That's because the existence of the simulation
as a simulation -- which is to say, a specific correspondence between the behavior of the simulator and the behavior of one other specific system, real or imagined -- is etirely in my imagination, or the imagination of anyone else who "reads" it correctly, or who programs it.
In the real world, there's only a computer doing the kinds of things it does all the time, sitting there and changing voltage potentials and running a fan and making lights change on a screen, and making a speaker cone vibrate, spitting ink onto paper, and so forth.
And you can examine the behavior of the machine forever, and unless you know it's running a simulation, you will never guess from the state changes of the machine alone. (Outside knowledge, such as experience with simulators, would be required.)
The machine itself would have no way of knowing, were it conscious, that the changes in its body were supposed to represent anything.
If told, it would have no way of knowing which changes in its body were supposed to be involved in the simulation -- voltage changes, temperature changes, fan motion, lights going on and off, and which subsets of these? -- and even if it were told that, it would have no way of knowing what logical (symbolic) value to give to any particular physical computation (change of physical state).
Even if you told the simulator that it was simulating the motion of a creature through a landscape, and even if you told it which group of values were the group that determined the spatial and temporal dimensions, it's a coin flip to the machine which of those values would get assigned to which dimension.
Perfectly simulating the target system won't help, either, because the simulator has no way of knowing that it's dealing with a complete simulation of your target system, or an incomplete simulation of a literally infinite number of larger systems.
And it gets worse... if the simulating machine actually did generate a "world of the simulation" which was real rather than imaginary, and it happened to be simulating a real thing, like a lake, then don't forget that the two systems are mapped to each other. It makes just as much sense to say the lake is a simulation of the computer, so the lake must also create a "world of the simulation" in which the simulator machine really exists somehow.
And inside that simulator inside the "world of the simulator" inside the lake, would be a "world of the simulator" with a lake in it, and in that lake would be a "world of the simulation" with a simulation machine inside it, forever and anon like a mirror box.
In short, if all physical computations in our universe generated a "world of the simulation" corresponding to all possible worlds which could be represented by assigning symbolic values and logic to them, then pretty much any and every thing you care to imagine would be real.
(If you restrict this phenomenon only to physical computations inside things we call simulators, you have to explain how the rest of the universe knows to comply.)
OK, so the "world of the simulation" -- the one that looks like the target world -- isn't real to the simulator, or in the simulator, or anything like that.
So why do simulations work?
They work because people have decided to associate the computer's physical computations (state changes) with symbolic ones, and manipulate the computer's state so that subsequent cascades of state changes (physical computations) mimic the state changes (or the ones we expect) in some other real or imagined system.
This decision by the humans (programmers) has no effect on the nature of the computer's physical computations. In other words, if a change in the computer is supposed to mean "a calf was born" or "it got a little hotter" or "dad got home", or "a photon is in a particular location", this is in no way evident from observing the activity of the machine.
In other words, the symbolic computations (which is what makes the machine's activity a simulation or "information processing" in the first place) are imaginary.
This strikes a lot of people as odd, especially people who spend a lot of time around computers or who are into information theory.
Folks think, no, it's not imaginary, I asked my computer to add these numbers and it did it, that really happened, of course information processing is real, not imaginary!
But wait... that's not really what happened.
What happened is that a programmer, who knew how the whole "adding numbers" thing worked, set up a physical object -- one that can change states real fast with high predictability -- to change those states in ways that match some logic in his head.
He also set up a means for a person to set off a cascade of state changes which would make the computer light up some lights in a pattern that corresponds to a symbol that will correspond to the solution to the addition problem.
What happened in the meantime is simply that the computer changed in ways analogous to how a system would change if groups of those sizes actually merged.
All of that was physical computation. There wasn't any "logic" to it, unless you count the laws of physics as logic, because the laws of physics were the only laws affecting how that computer behaved.
The programmer and the user, however, can use the machine as an information processor, but the symbolic/logical processes (as opposed to the physical ones) are purely imaginary, because only human beings are aware or can be aware of what the computer is supposed to have done informationally.
The physical system is designed to have physical outputs (such as lights) that bounce off our bodies. Our bodies have brains which respond to the patterns of those outputs (like the shape 4,567,892.143) by changing state.
That change of state in the brain of the observer is the "world of the simulation".
If there is a "man" who is "in the simulation" who is "running" from a "tornado", that can only be happening in your (or my or someone's) imagination.
There's no other place for it to happen.
Nor is there any need for any other place for it to happen, since we can explain the entire system of the simulation only with reference to the machine and the programmer/observer.
