Why the Simulation Argument is Untenable (Part 1)
There has been a lot of talk lately about the idea that we and the world we are living in are, in fact, simulations of some highly advanced civilizations, possibly our future descendants. The argument was proposed by Nick Bostrom in 2003 and has gained momentum in recent years, even receiving endorsements by Silicon Valley investors and entrepreneurs, including Elon Musk. (For those who are unfamiliar with the argument, hit here or here.)
I think this argument is questionable on a few grounds. Here are the reasons :
1) A reason that the argument caught so much attention is perhaps that people take one of the possibilities of this argument —the simulation hypothesis (that we are simulations), seriously. However, a probabilistic argument is just that : it expresses our degree of belief in some statement, given our prior belief/background knowledge. Whether the statement is true cannot be decided through such probabilistic arguments. In fact, our probability assignments are always susceptible to update when we obtain new data, via Baye’s theorem.
For example, given weather condition of the past few weeks and certain prior belief/background knowledge, we assign a certain probability to the belief that it will rain tomorrow. If it does not rain the next day, we should just revise our prior belief/background knowledge.
In the case of simulation argument, we should just find out if we are really in simulation and then update the probability assignment.
2) Let’s say the simulation hypothesis is true - that we are simulations. This implies that our descendants — the posthumans are simulations too. But this would mean that the ancestors simulations created by them are simulations inside a simulation. We are simulations in a simulations (simulation^2) ! This will, in turn, imply that the posthumans themselves are, same as us, simulation^2, ad infinitum.
Since this logic goes on forever, we found ourselves to be simulation^∞!
However, there is a problem, as Bostrom’s analysis seems to rely on the assumption that universe is finite. But an infinitude of simulations will probably require an infinite amount of energy from the underlying reality, as each level of simulation has to maintain the sense of realness for its inhabitants. No matter how small is the energy required for each level of simulation, the total energy is very likely to be infinity, contradicting the finite universe assumption.
But even if the universe is infinite, it would be impossible to have a region of space (or even the entire universe) to possess an infinite amount of energy, as it would end up as a black hole.
In short, the (3) option in the simulation argument is untenable.