This weekend, Sabine Hossenfelder did a video and post about the simulation hypothesis, the idea that we might be living in a computer simulation. She dismisses the notion that consciousness can’t be a computational simulation, which I think is correct, but then settles on the idea that physics itself can’t be simulated, because we have no idea how to do it, apparently not noticing that her objections about physics are very similar to the objections many make about consciousness.
To be clear, I’m not a fan of the simulation hypothesis. However I don’t think it’s as easy to dismiss as Hossenfelder implies. But when thinking about it, I think we have to distinguish between two varieties. One involves simulating an entire universe that we just happen to be part of. The other involves simulating us and our reactions within a version of the existing universe.
The first kind requires that the simulation be running within an outer universe that is more complex than the one we’re in. You can’t simulate a universe that is as complex or more complex than the reality the simulation is running in, at least not without the simulation executing far more slowly than the outer universe. That means the reality we’re in is either very different from that outer reality, or is a more coarse grained version of it. But that doesn’t mean that outer universe isn’t there, just that it can’t work exactly as our universe does. So dismissing the idea because we can’t simulate our own universe with fidelity from within our own universe is unjustified.
I do agree with Hossenfelder that this type of simulation has religious overtones. If we are in a simulated universe, then the aliens running the simulation are effectively our God, or gods. But a simulation proponent would likely argue that there’s nothing about the hypothesis that implies those beings are even aware of our existence, or if they are, that they care about us one way or another. We might be an unforeseen side effect, one that could eventually gum up their desired results.
The second type of simulation is very different. It’s not an attempt to simulate an entire universe, only an attempt to simulate us. In this type of simulation, reproducing the physics of the universe is not necessary. The only thing that is necessary is giving us the conscious impression of those physics. In other words, there’s no need to simulate quantum physics or general relativity, only to simulate our reaction to those physics. Indeed, there’s no reason to assume the entire human race is being simulated. It might only include you, or me, or you, I, and a few other people.
Why would someone create a simulation like this? Who can say. They might want to simulate our reaction to a particular situation. Maybe we’re prisoners being probed. Or maybe we went to Rekall to pay for a simulated vacation. (If so, customer support needs to drop in right now so I can demand an exit and refund.) Or maybe our AI overlords are testing our loyalty.
The key thing here is none of Hossenfelder’s objections apply to this more limited type of simulation.
My own issue with the simulation hypothesis is that there’s no particular reason to think it’s true. We might live in a reality, or we might live in a simulated reality. Straight reality seems like a simpler model than simulated reality, at least given the evidence we have so far. But this is a far less polemical stance than Hossenfelder’s position that it’s pseudoscience.
Personally, I’m not a fan of flinging the pseudoscience label at something unless it’s a clear case of fake science. The simulation hypothesis is speculative metaphysical guessing, but I don’t think it rises (falls?) to that level. The increasing trend of people calling anything they find implausible “pseudoscience” cheapens the bite of that label, which I think should be reserved for things like astrology, creationism, homeopathy, and similar notions that are clearly contradicted by science. Using it against ideas we just find implausible is simply pushing our own views in a sanctimonious manner.
Unless of course I’m missing something?