I knew you were just waiting to point that out.
Daniel DeMarco

On Simulation Hypothesis as an Omnigod

A tangent about the point(lessness) of game-like theoretical physics


It is true that a maximally broad definition of the hypothesis that this world or we (or a combination of the two) are a simulation can explain absolutely anything in some way. Or in other way of looking at it, explain nothing at all. It could amount to saying something about as useful as “god did it”. However, when I talk about trying to test whether this world might be some kind of a game, or that a proof of astrology may be a proof of this reality being simulated and vice versa, I’m intentionally proposing narrowed down, concretized versions of the simulation hypothesis.

It may take some time spent thinking about the possibilities before a tangible experiment shows itself, but that doesn’t make this line of inquiry very different from string theory or any other kind of theoretical physics. Unlike string theory, I believe that simulation hypothesis is actually much closer to devising some sort of practical test. It may not work out for a number of reasons once tried, but if the focus isn’t too broad, there should be something to try that hasn’t really been tried before.

In regard to astrology in particular, the useful way of thinking about it is asking what kind of algorithm would it be. On the quantitative level, what kind of algorithms that we use does it resemble? If this was, say, a game, what purpose would such algorithm serve on the qualitative level? The answer to both of these questions is thankfully pretty straightforward in the case of astrology. At it’s core, it would be a randomizer. Which, if you know how random works in software, is perhaps more aptly named in the opposite way — an organizer, making sure things happen in the proper order and proportion, in a synchronized manner.

In simulations, and especially games, there’s no true randomness. All kinds of things happen in very complex patterns, but the point of those is mainly to create an illusion of there being no particular pattern. Qualitatively, the purpose of the application of “randomness” in a game-like simulation is to create experiences like freedom, variety, or novelty, while still maintaining a systemic balance (or specific imbalance) of the game to ensure things like proper difficulty or particular progress trajectory. Overall, it can be thought of as a meaning machine.

In fact, I intend to use astrology inspired mechanics precisely to introduce meaning and balancing thereof with notable mathematical features into a game universe. As such, it’s actually more advanced than what has been used in games so far (bipolar choices with a neutral or average third option). Significantly more advanced. Astrology allows static balancing of twelve unique types of meaning along six connected polar axes in four qualitative different categories, or dynamically in a cyclical triad or a repeating or evolving twelve-step progression. It can also be scaled up to meaningful systems numbering twelve to any power of members, which is something I call meaning resolution.

What I’m driving at is that if this is a simulated universe of which astrology is a function, it’s not a hypothesis infinitely broad in its implications. A piece of software that has such very specific organizer as part of itself narrows down its possible purposes or types significantly, and even if one allows for the possibility of unintended consequences or flaws in the execution, their nature is also narrowed down to how an organizer can fail technically or in terms of fulfilling one of the purposes that an organizer can have.

If astrology is a real feature of this universe, then qualitative experience of sentients is an important feature of the simulation. Otherwise, perfect order or unrestrcited randomness make more sense (if efficiency is the objective or there’s no meaning-related objective). Fortunately, astrology really is quite testable. It hasn’t been tested in very interesting or comprehensive ways so far, but that’s not because it can’t be done. As long as one respects the astrological theory in the research design.

One big thing I failed to mention last time is that statistical studies of astrology need to approach, among other things, the sampling differently than how it’s normally done in behavioral social sciences. The more average and “random” the sample, the less likely it is that a statistical imbalance between people of different signs or planetary aspects will show itself. All other things being equal (“random”), a sample made up entirely of for example geminis measured against a sample of for example scorpios, ideally uncommon extreme geminis and uncommon extreme scorpios, logically has much higher chance of differing significantly in their behavior, or “luck”, based on astrological theory.

Astrology also supports the idea of generations, which should be taken into account, and that various times are better or worse for particular kinds of tests involving particular astrological types of people. There could be a time that’s predictably equally great or bad, lucky or unlucky, or pleasant or unpleasant for a whole element (three signs at a time) or a whole quality (four signs at a time). There are multiple variables in complex relationships to factor in, but a finite number of variables in strictly defined relationships. Come to think of it, one can first try to figure out experimental procedures that can identify or reverse engineer similar algorithms in actual games, to test out the test designs on something where we’re sure to find a pattern.

There’s no need to get into any more detail right now, but believe you me that there is much more detail to go into. Which is why the Barnum/Forer effect in particular is such an egregious insult — astrology can be criticized for many things, but vagueness or lack of unique testable propositions is really not one of them. More apt criticism, though not tested at all, is that the real smokescreen is the complexity. When so many aspects can mean so many specific things, it can make it possible to always have something to fall back on that applies to the situation at hand (make enough guesses and you’ll be guaranteed to get it right).

On the other hand, it’s clearly unreasonable to expect that something like this would be anything else but complex. If a machine designed to give meaning to the whole universe was very simple, it would probably not be able to produce very convincing experiences. Still, whether a complex system like astrology works or not cannot be proven by assumptions. Push it to all of its logical extremes, involve a lot of test subjects, and then we’ll have something to talk about.

As for the interference of the simulation runners if we get close to figuring out that this reality is simulated, I’d personally wager that this system looks a lot like one where there’s not a whole lot of crude and obvious outside interference. For starters, this universe is clearly very bottom-up physical in its architecture. Also, we’re already guessing it might be the case that this is a simulation, and we’re still here. Besides, one wouldn’t interrupt an experiment or a game that’s running within its normal parameters, perhaps not even one where something highly unusual is happening, to see it play out.

Given how fleeting species or civilizations seem to be and how vast universe is, the problem of someone figuring out they’re living in a simulation is also likely to solve itself pretty fast by natural means, in the grand scheme of things, and as Fermi paradox shows us, remain quite locally contained. It could be the whole point of the thing to figure it out, actually. Technically, anything is possible and we know nothing for sure yet, but that’s like saying that using math to figure out the correct physical model of the universe is pointless because it allows for all sorts of currently untestable wacky models.

While I would concede to conventional physicists that a simulated universe may be created in such a way that a material proof of it being artificial is impossible to obtain from the inside, I don’t think you can create a simulation that can lack or hide all of the key functional components of software. It only requires looking for abstract mathematical patterns that violate true randomness and conventional causality in connection with experience of “players”, and understanding material particle physics as something the point of which is to provide a particular experience for the “players”. It may turn out to be an incorrect hypothesis, but so may anything else, and you have to actually do some science into the matter first to know anything.

Like what you read? Subscribe to my publication, heart, follow, or…

Make me happy and throw something into my tip jar