It seems increasingly obvious that we hilariously live in a simulation. I put together a loose proof.
If these are true
- Consciousness is substrate-independent: carbon-based neurons and silicon-based semiconductors can both have consciousness as an emergent property
- The marginal cost of memory and computation will trend to 0
- Any intelligent society can simulate reality at some degree of fidelity.
- The rate of improvement of the fidelity of the simulation will improve every year at a non-zero rate
- The post-simulation intelligent society (An intelligent society once they can fully simulate) will have the desire to simulate reality
Then there is some date in the future in which we will reach a 100%-fidelity simulation of the world. This seems to be not too far away for our civilization. If this is true of every intelligent society, then it is more likely than not that we are simulated.
There are a few potential scenarios where we may not be in a simulation:
- Every intelligent society on the way to being post-simulation goes extinct via an asteroid, nuclear war, etc., so no simulations exist (P(DOOM) = 1.0). This seems unlikely. This could be true if the base reality doesn't want its simulated realities to simulate so the reality gets nuked before they simulate anything. Adjacent to the Tower of Babel.
- Post-simulation intelligent societies will not desire to simulate societies. If doing a high fidelity simulation gives society some pleasure and quenches a curiosity thirst, and society realizes it can get the same—if not more—pleasure from thinking -> triggering a BCI -> shooting dopamine out of the hypothalamus, then there would be no reason for society to simulate. If all post-simulation intelligent societies are like this, then would be no simulations. But I suspect (I hope) there would be a few people that wouldn't be dopamine sheep.
If a society can simulate, there would probably a inducted chain of simulated worlds since each simulated society would presumably also simulate, but there must be a start reality. To each simulated society, their simulator society would be considered God. At our current degree of understanding, it would be impossible to know if we are the start reality. If I simulated a society, I make it impossible for them to figure out if they are simulated. But if I had a dark humor, I would make small miracles and tiny signs appear to make people believe that there is something out there, always out of reach. If being simulated is highly likely, the most probable way that we aren't in a simulation is if we are the start reality.
Selective simulations are the most interesting, where you are the only one simulated with a "free mind" while every other person in the world is a NPC designed to seem real but actually are zombies. No one is real but you, slightly schizophrenic. Maybe the schizophrenic individuals are the only ones that realize this, yet we treat them as outcasts. But if all worlds are predetermined with no chaos, then your "free mind" wouldn't really be free, and advanced simulator societies probably wouldn't care about doing this sort of simulation. The only interesting simulations would be where there is novelty.
Some people may point to how memory and computationally expensive this would be. If I, as the base reality, simulated 100 societies (would probably be insanely larger), and they simulated 100 societies, and so on, then at generation n, the number of societies would be 100^n. Assuming strong advances in quantum computing, machine learning, etc. this would be super doable. And if I get concerned with server load, I can always extinguish trillions of realities with the run of a command (rm -rf earth_0000000000000010000000031201).
Every process — all the way down to the quantum level or lower — doesn't need to be rendered and can be rendered ad hoc. For example, our satellites can only see so far into space. We literally have a render distance (Minecraft type shit), a function of our best sensing technology. Information degradation of celestial objects far away is the same as quality lowering for objects far away in video games. No point in rendering it at 100%. Information dimensionality could be a function of your sensing distance to any object, with the farthest objects not even being represented (if beyond the render distance) or materializing as highly reduced representations.
If all information began in the big bang, then today's information is the butterfly effected version of that original reality. Each {insert the smallest unit of time} is a derivative of the previous one, with the delta between them being the mapping function. This would form a long multi-dimensional spatiotemporal Markov chain. if you know every past "event" from now, you can trace back today's world in this exact second to the big bang atom by atom. Some very interesting pulse and press experiments could be done to measure the long-term impacts of small things. Like the impact of rm -f "ant_00000000000000420" in the Indus Valley in 10,200 B.C. to multi-stellar democracy and liberalism in 1,233,274 A.D.. I'd have so much fun as simulator. And if we treat our simulated worlds the same way we treat ChatGPT or our AI girlfriends, then we would have no problem of committing Thanos-esque multi-world genocide with the click of "Tab".
This wouldn't really change the way we live. But rather than caring about everything, it seems best to act as if you would act in a video game. Accumulate all the influence, generally via accumulating all the capital. This is true in our game. Manipulate the world according to your internal worldview. Die a martyr after saving the world from existential destruction. Figure out the meaning of life and find better questions to ask.