The Problem of Evolution in a Simulated Universe

I have been spending a considerable amount of time lately (2018) thinking about various aspects of different simulated universe hypotheses. One topic that arose in a recent post piqued my interest, how to account for evolution by natural selection in a simulated universe. I wish to address the issue from the most common simulation hypothesis perspective, what I call the simulation hypothesis 1 (SH1) type universe. In SH1 scenarios both the universe and ourselves are 100% simulations. In other words we are 100% simulated beings living in a 100% simulated universe. Immediately it seems a concern reveals itself for evolution by natural selection implies that there be a ‘natural’ world for the selection by fitness to occur in. In SH1 there is nothing natural, everything is simulated so how could evolution actually function in such a universe?

There are at least two and no doubt many more ways out of this seeming contradiction. The ease at which one can imagine ways around the problem highlight my main issue with all SH1 hypotheses, the too good to be true problem (tgtbT). Basically tgtbT says that SH1 is extremely unlikely because in SH1 it is way too easy to account for any aspect of how or why the universe is the way it is and how or why we are the way we are. Essentially it is too strong of a hypothesis. In the case of evolution by natural selection all one need do is say that the simulators have ‘programmed’ the simulation in such a way so that the beings and worlds within the simulation behave in ways consistent with evolution by natural selection. They have set the appropriate inputs (physical laws and behavioral characteristics) which result in an output that is consistent with evolution by natural selection. Alternatively one could argue that the simulators ‘programmed’ us so that we only think things evolve by natural selection when in fact it is a totally random process or there is no evolution at all, it is just the simulation spitting out new life forms as it churns through its vast database of program options. We see it as evolution of species because that is what we must see, what we have been ‘programmed’ to believe.

I want to set aside these, and any other easy answers and ask the question under the assumption that they are not available to us. Let us imagine that the simulators really only act as ‘first movers’. They set up the initial conditions and hit ‘run program’ but beyond having complete control over the initial conditions and physical laws they have limited or no control over what happens next and no special insight into how the ‘program’ may unfold. Let us also assume that it is possible for a race of such power and intelligence to actually do such a thing and that they would be willing to risk doing so. (i.e. smart and powerful enough to set up a simulated universe’s initial conditions and physical laws but not smart and powerful enough to know what the outcome will be, and daring enough to risk the potential consequences if things were to ‘get out of hand’). Incidentally I find this option highly unlikely given the high level of risk inherent in such an endeavor. Moreover what the purpose of such a risky gambit might be is less than obvious. Certainly entertainment or curiosity might be posited as possibilities, however, to risk so much, basically because you are bored seems highly unlikely for beings with such intelligence and power.

Let’s set that argument aside for now and simply accept my setup as a given. Simulators as first movers who then are totally hands off from that point on. Let us also assume that they have set the initial conditions in a manner that is consistent with the eventual formation of a universe, a galaxy, a solar system, and a planet, very much like our own. On this simulated planet, virtually identical to earth in every way, or we can ask the same of our very own earth, evolution by natural selection is said to have occurred and it continues to this day. In another universe their exists a virtually identical planet. This universe sprung into being completely naturally, perhaps via a ‘big bang’ followed by inflationary expansion, etc. or perhaps via some other mechanism(s) currently beyond our understanding. The point is the creation of the ‘natural’ universe was not influenced or directed in any way by powerful, ultra intelligent simulators. This planet also has beings that have (supposedly) evolved by natural selection. We then ask the question is their any difference between the beings on these two planets, in these two fundamentally different, but virtually identical in every other aspect worlds. Did the beings on both worlds really evolve by natural selection? Going a step further let us say we find the answer is no, does it even matter? If the end results is the same why should we care?

Typically at this point I would give a long winded spiel outlining my answers to the questions I have posed. Unfortunately at the moment I must say I am stumped. My heart is telling me there has to be some difference, the natural world really should be favored and that it should matter, but my head is not able to posit any reasons that withstand even a minimum level of scrutiny. I am going to leave this one be for now, publish it and let anyone who has some ideas take a whack at it, or not. It really will not bother me one way or the other.

I wonder if they have Arby’s on this simulated earth and if so do they have the beef Gyro? Certainly if there is an Arby’s they must at least have curly fries and I think it goes without saying that Mountain Dew will be widely available. Perhaps the creation of the beef Gyro was the purpose all along and we have already reached peak simulation.