Lately, probably due to Elon Musk being in the media spotlight 24/7, the issue of whether or not we live in a simulation seems to be everywhere. I think the reasons for the virality of this view is because it’s considered outlandish, not because people believe it has merit and may be a useful way in which to look at the world. I’m not saying this is a bad thing, the more new ideas people encounter the better, even if they are quickly dismissed. This post is therefore about the merits of such an idea.
Technology in general has been advancing at an exponential rate. Coinciding with an advance in technology is an advance in computer simulations. These range from simulating rocket launches, cancer treatments, the spread of disease based on sociological data and the list goes on. If we can simulate all of this, what’s stopping us from simulating all the neurones in 10 billion human brains and everything else in our universe?
It may still sound far-fetched, but the reasoning is the same as the “artificial intelligence will surpass us” argument. For AI, humans only need be able to build a machine as smart as a human, that way it has the capacity to improve itself and the process can iterate indefinitely from there. In the same, we only need to build robots that can build computers as well as we can so that they can accelerate the building of bigger and faster computers.
The size of a computer must be massive to allow the simulation of such a universe. So big, there is not enough resources on Earth, or even our solar system, to be able to build one. Nick Bostrom was the man who has predicted this limitation will be overcome by using computer building, self-replicating robots. Robots do not need oxygen to survive, they can be made to be robust (more than can be said of humans) and thus can be sent away to other planets, solar systems and even galaxies. With enough advancement, these robots would be able to turn planets into computers. All for the purpose of simulation.
Let’s grant that the above is possible; what does that even leave us with? Is this a good future? With this sort of computing power at our hands, it would be inevitable that AI would advance far beyond our control and comprehension. Assuming the AI is benign, what does this computational power mean for us? It means we have the ability to simulate universes of our own. Which may in turn simulate universes and so on. We may even choose to move our consciousness into one of these worlds. If that becomes possible.
Also, if we were to enter a simulation that we built, we would already know how to build another layer of simulation inside it. Given that we don’t yet have another layer of simulation ourselves, we may be simulated by some higher species as an experiment. Perhaps we do have a layer below us, from our universe. It’s just that we’re not the species to have found it. There are limitless possibilities.
So, if it’s possible for us to create a simulation of a universe, we must grant the possibility of ourselves being in a simulation. The exponential nature of the formation of deeper and deeper levels of simulations would therefore mean it’s almost certain that we are in a simulation.
How does this impact our philosophy then? If we are part of simulation, realism cannot describe reality. You do not perceive the noumena of binary digits e.g. binary describing a soccer ball, you perceive the phenomena, the soccer ball as you see and interact with it.
The simulation theory also supports Berkeley’s subjective idealism. Taking God as those beings responsible for our universe’s creation, subjective idealism almost becomes a perfect description of a simulated reality. The computer is the thing that keeps you and I aligned in a consistent reality. Whether the consistency is real, that is, it’s consistent at all times, is another thing. For example, if I look at an apple I see its skin, imperfectly shaded and lined mostly red with some white. If I put that apple under a microscope, I see that there are cells that make up the skin. There is no evidence, however, that the cells are always there. Perhaps the computer only renders the cells if you look into a microscope. It’s a bit like if you open a door in an open world video game and you have to wait for the new room to load, because it hadn’t been loaded from the beginning. In fact, a smart computer would do it this way, just with undetectable loading times.
Personally I don’t think it matters too much if we’re in a simulation or not. I think we should try to simulate universes ourselves, regardless. We could create a universe free of suffering!
So keep a look out for any glitches or lag. Don’t get your hopes up though, we are a much higher chance of being in a simulation billions of levels down from the first than we are the second. An AI smarter than any AI that you could imagine could imagine probably built this.
