Been thinking lately about the simulation hypothesis, which I continue to regard as a downright fascinating idea.
One thing that struck me about Nick Bostrom's original thoughts on it was how his emphasis was specifically on a certain type of simulation: the "Ancestor Simulation", which would simulate not only a world, but also all of their evolutionary development, etc, starting from the Origin of Life or the Big Bang.
At the time I thought, "Huh. Well, THAT is awfully specific." I mean, there's a lot of programs one could code up with the amount of processing power Bostrom's imagining could be kicked around. And a lot of programs which could feature conscious beings (if you accept Bostrom's premises, of course). Why assume that the ancestor simulation is the only relevant one to think about? Why not any other number of possible simulations that would include simulated worlds and conscious beings? From what I read, those possibilities went unmentioned by Bostrom. Even by Chalmers, with his Matrix paper.
A few days ago, it finally hit me as to why.
Imagine the NPCs in World of Warcraft were conscious. This would qualify as a simulated universe, I think, in the sense Bostrom is talking about: Conscious beings, interacting in a virtual world, etc. Now, admittedly there is an in character storyline for the origins of various things. Typical fantasy fare: Created by godlike beings, corrupted by demonlike beings. Dwarves originally carved from stone and starting off as robots, then receiving the curse of flesh and becoming similar to how dwarves are now. Etc, etc.
But, of course, all that backstory is a load. In reality World of Warcraft went online a few years ago, kingdoms and all, programmed into existence. They didn't even decide on the backstory/origins for these guys until years down the line, most likely. The archaeologists that the dwarves have can find evidence after evidence about their true origins (being rockmen, the curse of flesh, etc), but that just isn't how it went down.
You probably see where I'm going with this.
A technologically advanced civilization with tremendous processing power at its fingertips, capable of simulating fully conscious beings, could damn well make a world look like it's 14 billion years old. The sims' scientists, their every measure, would find results consistent with this. And yet, it wouldn't necessarily be true for them. Maybe the programmer wanted to "accurately" simulate a given period of time. Maybe the programmer is being imaginative. Maybe a million things, many of which entail that the history virtual scientists are discovering is bupkis. Just as, when a scientist creates a simulation of this or that phenomena, he doesn't necessarily import the entire history since the Big Bang into the simulation. It's just not his interest.
But that means, if we have reason to suspect we live in a simulation, it seems we have tremendous reason to view our evolutionary history as fictional. Last Thursdayism is in play, as is 6000 years-ago-ism.
Here's the part I particularly love. Bostrom, along with Chalmers and others, have dug in their heels and insisted that the simulation hypothesis is a naturalistic one. Which I suppose they'd have to - if the idea of our living in a simulated universe was supernatural, then confirmation for the supernatural exists all over the place, right here and now. In fact, we're supernatural beings, even if it's only in relation to our simulations.
I think the ramifications of this are far reaching, radically so. For one thing, it makes Young Earth Creationism conceivable as a naturalistic hypothesis. YECs, in fact practically all religious claims, can almost entirely be reconceived as naturalistic.
So why did Bostrom focus on 'ancestor simulations' in his simulation papers? I can't help but wonder if it's because he wanted to avoid arguably the single most grievous scientific sin in the modern world: Doubt of evolution. But if it's possible to simulate universes complete with conscious inhabitants, that - and other kinds - of doubt are introduced. I agree with Chalmers that this doesn't entail that simulation theories are necessarily skeptical hypotheses in the sense of forcing us to doubt that there exists an 'external world', or doubting the possibility of knowledge, etc. But, it introduces skepticism enough.