This is a special post for quick takes by Warty. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
You could think, Solomonoff induction, something something, therefore the universe is simple. This doesn't work.
There is a very simple program that simulates all programs (run program i at 1/2^i speed), including our universe. But that doesn't specify your observations.
The description length cost for Solomonoff has to have the universe + locating you within it. So complex worlds where locating you is easier are more likely.
It could be fun to calculate how big the components are %wise but idk how to do that. Also, idk why I'm still expecting the ultimate physical laws to be very simple 🤷♂️
well to me that's a neat thing, in a way occam's razor is mathematically necessary because there's just too many complex universes so you will always have some level of complexity beyond which you assign arbitrarily low prior.
Though Solomonoff prior takes that beyond whats mathematically necessary, because you could still assign 99.9% to your favorite giga complex model, and Solomonoff says derive prior just from simplicity.
2Spective
I agree, but if you have already convinced yourself of Occam's razor to your satisfaction, don't we get the expectation that universe is "simple" just from that? In the weak sense of all-else-being equal we put more on weight simpler theories and penalize for complexity. Perhaps there was some stronger sense of simplicity that you had in mind.
So Solomonoff induction provides sort of formalizaton of Occam's razor. But it works on universe+location. And location seems very big, so it sounds like universe is not necessarily super small. But I still have a guess that it is super small. But my argument in OP proves that's nonsense. So maybe I would like if there was a better cool mathematical philosophy thing that would reflect my intuitions. Or if they can finally figure out the turing machine for the universe so I don't need intuitions.