The simulation hypothesis doesn't necessarily imply anything religious, any external observers or simulator makers. The universe, if it is a simulation, might be a result of a short code in John Neumann's (1958) sense, running on another universe. A short code is just a higher level programming language, so while it requires programming the underlying computer, all the instructions that follow are shorter, simpler, compressed.
The computer might be simulating another but both might be randomly occurring codes that self propagate and form groups. In that case, the universe we observe can be a simulation as a result of code formed randomly on an underlying universe and self propagating, without any programmer. Stephen Wolfram (2002), indeed, shows how ergodic systems can generate all possible rules for themselves to follow and one of those rules might be a simple program, meaning it can be a result of randomness at a lower level, but the simple program codes for a complex simulated universe.
If so, then without any programmer or simulator, or external observer, there are not necessarily any applications to real life.
The string theory gentlemen use moduli spaces, which are all possible local least energy states, also called vacuum states. Some might belong to systems that underly our universe in the sense that n =(M) m if exists f ∈ M such that n = f(m). If we can separate a physically meaningful group according to vacuum states into independent layers M(1), M(2), . . . , taking these as sets of generators, and we observe only some of these, but all appear physically meaningful, we might suspect our reality can be generated in the sense of a simulation, our reality being a short code, not a complete code, compiled and simulated on an underlying system that meshes with it but exists below it.
(Just added this reply to your question to the main text.)