Here you’re making unreasonable demands of the substrate the simulation runs on, though. Could be that in this upper-level world, those few trillion years it takes to come up with some convincing conversation-tree are about as relevant as five minutes are to us; could be, depending on how computation works up there (how fast the implementation is, that is – I don’t think it could really work fundamentally different from how it works in here), that it takes all of five minutes to actually pull it off.
I’m not sure about that, but somebody totally should steal the title for a paper.
As I said, I’m not sure if this is sufficient. Nothing truly exists in isolation – something that, I should think, remains true whether or not we live in a simulation --, so, the need to cover up any incoherency here is liable to have effects all the way over there, effects that, for any sufficiently complex system, will depend chaotically on the cover-up efforts. I suppose you could, in theory, cover up the original incoherency, and cover up all effects of the cover-up, and cover up all effects of that in turn, but could you do it arbitrarily long, arbitrarily well, even if you had total control over the simulation? Everybody seems to take it as a given that yes, of course you could, but to my mind it could well be one of those intuitively plausible assumptions that all too often turn out to be wrong after all. For instance, as I’ve hinted at, it might be that hiding the cover-up’s cover-ups, and so on and on, is a bona fide supertask, and that the execution of such is impossible no matter what paradigm of computing you operate under.
In that case, it would be impossible to ensure that the simulation at all times appears coherent without actually being coherent – in other words, each fakery would in principle be detectable from inside the simulation (not to mention the obvious question of why you would want to simulate anything when you’re not gonna let it play by the rules anyway – the motives of whatever the higher-ups might be is something best not speculated upon).
Think about a relatively simple simulation, such as the global weather forecast. There are certain weather conditions that are impossible, like for instance a non-zero wind strength at all points, via the ever-popular hairy ball theorem. If you let your simulation pan out according to fixed rules, you’re guaranteed that such impossible conditions never arise; however, once you go in and change something, it might be that you lose this guarantee. So then, you’d have to watch out for such anomalies – but, and here’s the cracker, you only see them when they arise; the system being chaotic in nature, there’s no way for you to reliably predict them in advance.
Of course, you can then erase and rewind, introducing another manipulation in order to avert this inconsistency from manifesting itself, but then, you have to watch out for the consequences of two manipulations. The question then is whether any finite entity can keep abreast of all such unforeseeable consequences all of the time – and I think there’s a very real possibility that that’s not the case.
But that’s just you, imposing limitations on god. Remember, I assumed that, against logic, he actually can answer this question. I don’t actually believe that an entity with this ability is even conceivable, but it’s also pretty inconceivable for me to spend all of my days henceforth in a moonlit pool made of solid gold, skinny dipping with Megan Fox, yet this doesn’t appear to hinder my imagination at all!
Eh, my delusions are far more garden-variety than that…
It doesn’t need to be, but sometimes, things are done just to show that they can be.
But it’s true that I’ve gotten over-enthusiastic about what’s perhaps a half-baked idea at best. I’ll concede that, much as I may not like it, it seems enormously more plausible to me that those in total control of a simulation that I am part of are indeed able to keep me in the dark about any sort of trickery they decide to pull on me – but I still think there’s a tiny spark of a possibility that that’s not so.
Anyway, I have to dash and skip the proofreading – any mistakes, I’ll just blame on Agent Smith fucking with the datastream.