Have advocates of the simulation argument actually argued for the possibility of ancestor simulations? It is a very counterintuitive idea, yet it seems to be invoked as though it is obviously possible. Aside from whatever probability we want to assign to the possibility that the future human race will discover strange previously-unknown laws of physics that make it more feasible, doesn't the idea of an ancestor simulation (a simulation of "the entire mental history of humankind") depend on having access to a huge amount of information that has presumably been permanently lost to entropy? Where is the future civilization expected to get all the mental structures needed to simulate the entire mental history of humankind (or a model of the early Earth implausibly precise enough that simulating it causes things to play out exactly as they really did)?
If things don't play out exactly as they really did, does the simulation argument lose any force?
This thread is for the discussion of Less Wrong topics that have not appeared in recent posts. If a discussion gets unwieldy, celebrate by turning it into a top-level post.