Bostrom’s Simulation argument can be summarised as follows:
(1) If highly advanced civilisations can construct computer simulations of reality populated with simulated but fully conscious inhabitants; and
(2) If a significant fraction of civilisations will actually get to the level of technological advancement required by (1) before they become extinct; then
(3) The probability that we are, in fact, simulated conscious beings inhabiting a computer simulation of reality run by highly advanced superbeings will approach 1.
There seem to be a number of unspoken assumptions in this argument.
(1.a) A significant fraction of highly advanced superbeings will devote their time and computational resources to behaving like the bad guys in a Keanu Reeves movie.
(1.b) It is technically feasible to expect a computer simulation of physical reality to have fault-free uptimes on the order of billions of (subjective) years.
It should also be possible to search for empirical evidence of (3) in the form of programmer comments. There are rumours that the pattern of quark-antiquark jets in early LHC results spell out the following message
/* Grrr, have to set FSC to 1/137 so that this will work on UniverseOS releases earlier than 1.9. What a fucking mess. Backward compatibility yay! */