If the universe is completely non-deterministic with infinite random events happening, shouldn't the odds of my living in the specific sub-universe that appears fully deterministic be almost indistinguishable from zero?

As I said, I want to argue that the sizes of ordered and chaotic regions are of the same cardinality.

That's what happens with decision-making under uncertainty: you aren't sure of something, yet you have to lawfully choose your actions. You don't choose your actions based on what you know is certain, you choose your actions depending on the specific state of uncertainty you're in. If you saw that X happened 100 times, you choose action P, and if you say Y happen 100 times, you choose action Q, even though your state of uncertainty permits the future to go identically in both cases, so that choosing different actions won't do any good. And maybe even the possibilities open for the future are exactly the same, but the fact that the past was different weights on the decisions just as well. That is what we are, cogs in the engine of possibilities, determining what happens even if we don't know what it is.

Ugh. Agreed. I think this post is a reference to Loschmidt's paradox, i.e. that there should not be a "thermodynamic arrow of time" in an otherwise time-symmetric universe. It has been discussed before on OvercomingBias, but I can't find the link since it isn't mentioned by name.

In an infinite universe of the right kind, order can locally emerge out of random events. Our universe is of the right kind.

So, we can account for the order in our observed (local) part of the universe.

Here, Spluck seems to be using one of the common attempts to resolve it, which is that the universe is time-reversible, but we are in a region where entropy happens to only increase.

The (oft-cited) problem with this position is that local irregularity should not imply the further "irregularity" we see. That is, if the universe is truly random and time-symmetric, and you find a locally-time-asymmetric part, you should not expect to see this further asymmetry again and again. It's like believing that because you won the lottery, you're likely to win it again.

Firstly, this kind of multiverse is esentially the same as a parallel worlds one; the only difference is which dimension you take the multiplicity to occur in. I prefer parallel worlds as it implies a logical branching structure, a cladistic tree which provides an overlying system to the otherwise arbitrary worlds.

Er, I suppose there is a weak sense in which you can "prefer" many worlds to inflation but they're basically unrelated theories. Its not like we're choosing between them. Both could well be true. Modal realism could be true too... thats when it gets crazy.

Now we're getting to the heart of it. Upvoted. What does it mean to live in a hume world? For example, we would have to accept the existence of non-reducible mental states (everybody here granted the consistency of the theory until now) and take everything on faith. But indeed we cannnot take anything on faith, since we cannot think, if thinking is a causal notion!?

Suppose for the sake of argument we're not living in a hume world, but had massive, perhaps infinite computing power. we could simulate so many hume worlds that there are some with order and inhabitants in them. They would then quasi-think, quasi-feel and make quasi-experiences. Everything happens as if there were necessitating laws governing it, but there aren't. But still, the universe quasi-looks ordered to them.

This theory and solipsism have something in common, but they are distinguishible. solipsism surely is consistent but higly implausible compared to the standard model. But there could be evidence for it. But it is of another sort than the evidence for a hume world. If pigs start to fly, only hume world-theory (HWT from now on) can explain this easily.

Another point not enough discussed so far are evidences for HWT: causal gaps and anomalies in the fabric of the world as we already know it: In a Causal world, how do we properly deal with mental causation, qualia, time travel paradoxes and in general, indeterministic processes? I'm not saying there are no other solutions, but a lot of people think we did not and possibly cannot make progress in these questions, at least in the current framework. But HWT delivers here.

I've added links, but it has no summary.

Okay... I wasn't going to downvote the post because this is the sort of thing I thought was marvelously clever when I was thirteen (so there's hope) and you'd already been downvoted below posting threshold (so it'd be just punitive), but really?

You have provided a way the universe could be (the word my undergrad prof used for a universe with no causation was "Humeiform", as opposed to "uniform"): there could be a lot of sub-universes arranged spatially in a single possible world, in which infinitely many universes instantiate every possible sequence of events.

You don't have any way to explain why we should find ourselves in a sub-universe that seems to fall into miraculously well-regulated patterns, when the overwhelming likelihood if you pick a sub-universe at random is that wacky, irregular stuff will happen all the time. The anthropic principle won't help you, because there's no reason in this system for the observers to mostly find themselves in sub-universes with apparent regularity.

Less mysterious? "There is no reason anything should be as it is; it just is, at random" doesn't eliminate the mystery or even push it back another level or two - it wallows in mystery, it rests its head on a pillow of mystery at night, it has a middle-management job at the mystery factory, and it watches the mystery channel on TV every day. It's just so saturated in mystery that, having never found this much mystery in one place before, you have chosen to ignore it.

More parsimonious? Perhaps you can say "it's all random" in fewer words than you can recite a basic physics textbook, but three words that have no evidence backing them, no way to obtain such evidence, and no connection with the world in which we find ourselves have a much lower usefulness density than any verbose, overexplained thing you care to pick out from science. Parsimony is about getting your theory as small as you can without losing information. Your theory, since it contains no information, could be expressed by chirping like a cricket.

The fact that it predicts nothing makes it practically useless and uninteresting. If you think it's fun, it's no skin off my nose if you believe it, but believing things because they are fun and not because they inform your behavior in an instrumentally useful way or because you have evidence for them is not what we as a community are getting at, so it should not come as a surprise to you that (given the causal regularities of this website) you were poorly received.

Judea Pearl shakes his head sadly at you.

Hey, we're giving your ideas serious discussion.

By your standards, we don't have a "good explanation" for gravity, electricity, formal logic or indeed anything at all. However deep we go, you can always point your finger and say nyah nyah, you haven't "explained" this base level! But your request would be hard to satisfy: pray tell, what conclusion could possibly follow from no premises at all? What concept is so obvious that it could be explained to an "ideal philosophy student of perfect emptiness"? None! So you're forcing us into bottomless recursive descent instead of, you know, building more precise models and actually predicting stuff.

The sciences do not try to explain, they hardly even try to interpret, they mainly make models. By a model is meant a mathematical construct which, with the addition of certain verbal interpretations, describes observed phenomena. The justification of such a mathematical construct is solely and precisely that it is expected to work. - John Von Neumann

If you want a mathematical model of causality, read the links to Judea Pearl. But if you want a "philosophical explanation", sorry, I can't help you here and I doubt that anyone can.

It isn't counterintuitive at all. I just think you have the physics/implications wrong. I think it probably does predict something. The anthropological effect only justifies sensing order to the extent that order was necessary for our existence. So if the only environment that mattered for the purposes of our existence was the planet Earth then we should only expect to find order on Earth. If you're right, as we gather information about more and more remote locations we should see less and less constant conjunction. For example, I don't think the neighboring galaxies perfectly obeying the laws of gravity is necessary for our continued existence-- so that order cannot be explained away with an appeal to the anthropic principle.

More importantly though, inflation does not predict that in an infinite universe there will be a finite region that contains every modally possible configuration. The possible configurations are constrained by the laws of physics. Indeed, you can't predict inflation unless you assume some (all?) of those laws are working throughout the universe. I think accepting inflation requires you to accept some universal laws of physics. As such, your theory is implicitly denied by inflation.

the fact that it predicts nothing can be seen as a downside, but should not disqualify it, since the theory can explain that very fact.

If that doesn't disqualify it, then I propose the following "Divergent Theory", defined as follows:

• Everything is Divergent Theory.
• Divergent Theory predicts Divergent Theory, thus Divergent Theory predicts everything.

The fact that Divergent Theory predicts everything means it predicts nothing. This can be seen as a downside, but should not disqualify it, since the theory can explain that very fact.

Where do I say or imply that? did you read it at all?

the model is ontologically more parsimonous [sic] because it posits all standard entities minus causal relations.

That isn't parsimony, that's ontological promiscuity of the worst sort.

Say we have two universes. One of them looks like what you describe: infinity sub-universes going about their random business in space. The other has no further division into sub-universes, and it has tight, deterministic causality.

You can find out anything you might care to learn about universe 2 by fully describing one instant and a list of the laws of nature. Then, you can (without any further premises) extrapolate forwards and backwards in time indefinitely and get the right answer every time, at least if you are a superintelligence.

In order to know as much about the first universe, though, you can't do that. The instant you pick to describe is more complicated than the one of the second universe to begin with, but let's ignore that - if we ignore that, then at first it looks like universe 1 wins, because you just describe the instant and no rules. But what if you want to know what happens at instant+1? You can't extrapolate: you have to do more positing. In fact, you have to individually posit every object, event, and area of empty space in universe 2, because they don't follow from anything and have to stand alone, random pillars of acausality in a vast sea of acausality.

The list of things you need to know about universe 1 in order to be able to derive everything there is to know about universe 1 just is the list of everything there is to know about universe 1. The list of things you need to know about universe 2 in order to be able to derive everything there is to know about universe 2 is much more parsimonious. Universe 1, says, "here is a WHOLE LOT OF STUFF." Universe 2 says, "here is some stuff, and the mechanisms by which it will change."

As for mystery: you must have an atypical understanding of the word. Mystery is what happens when something cannot be explained. There is no explanation for why everything might happen at random: there is only one step between the question and the mystery. ("Why is this stick of gum yellow?" "It is yellow at random." "Why is the color of the gum random?" "It is a mystery.") A causal model of the universe does not eliminate mystery, but it postpones it and makes it less conspicuous. ("Why is this stick of gum yellow?" "It is yellow because it refracts light at a wavelength that the human eye sends to the brain, which interprets it as a color which is called yellow in English because the word is derived from Old English geolu." Then you can follow the chain backwards pretty damn far in any number of directions: why does the substance reflect that wavelength? Why did light fall on the stick of gum? Why does the human organism interpret that wavelength as a color and not a temperature or a sound? Why is yellow its own color in English and not in some other languages? What's the etymology of geolu? The mystery takes a long time to turn up. There is less of it.)

the cardinality of ordered regions is the same as that of the chaotic ones in this sort of infinite universe

Accepting this statement for the moment, I should point out that cardinality is probably not the best measure of the size of the sets for this purpose. For a simpler case, the natural numbers, see natural density.

The model is less complex, possibly, but this is not the sole criterion on which it should be judged. Occam's Razor discourages complexity, but it encourages precision, and a Humeiform universe avoids the first only at the cost of utterly failing the second.

To use a metaphor: you have to pay for everything you put in your model, but your return on that investment is related to how much accuracy the model gives. More expensive (complex) models can still win if they give a bigger enough payoff (sufficiently complete prediction of the observations).

Why are you replying to us in top-level batches? Are you trying to limit how many total comments you make as a form of karma damage control, or just determined to make the threads disparate and hard to follow?

Anyway:

i talked about ontological parsimony. you're talking about something else. epistemological parsimony, perhaps?

Do you mean you think you're proposing fewer basic kinds of entities (i.e. you think objects exist, and we think objects and causes exist)? That seems to me very suspiciously like a feature of how things are worded - you could just as easily say that we're proposing quarks and space and time, and you're proposing quarks with different properties (e.g. the property of appearing, disappearing, and moving at random instead of the property of interacting with other quarks causally) and also space and time.

same for mystery. that you can prolong it doesn't mean there's less of it.

Most of those mysteries boil down to a fairly small number of things we do not yet know. Yours, granted, boils down to just one thing you don't know: why the hell do all these random things happen? But it is a very BIG and very CONSPICUOUS mystery, and there is no good way to get rid of it or push it farther away or shrink it. This is not true of our small host of mysteries, which we regularly shrink and push and can even hope to eventually do away with.

Side point: if you click on the "Reply" link below a specific comment, your reply will be threaded with that comment and the person you address will be notified of your response - I recommend it.

As for your actual points: I don't care about ontoepistemoparsiwhatever. I care about being right. And your proposal - I guess we're calling it the Humeiform theory - isn't supported by any conceivable block of evidence, including that which actually holds true. If there weren't any other theories supported by the evidence of the universe, it might win by default, being simplest, but this universe is regular enough to be accurately fitted by many not-overly-complicated theories.

If I want my beliefs to correspond to reality, I'm better off favoring simple theories over complex ceteris paribus - this is Occam's insight. But I'm also better off favoring accuracy over vacuity. And when I shut up and multiply, the latter dominates the former for this special case.

jack: yes, I saw that problem too. That's why I said the theory might be self-defeating. My idea was that even if inflation as a theory is strictly speaking forbidden, it can phenomenologically point in the right direction. I mean, we might be still able to say something like: the "quasi"-observation" or the "quasi"-theory is true.

But what I said isn't the same as saying the theory is self-defeating. The theory is just based on a false premise (that inflation allows for regions of finite space that violate our recorded laws of physics). Inflation says: "Any given of configuration of a region of finite space that does not violate the laws of physics exists infinitely many times." You say: "There are some reasons of finite space where the laws of physics are violated!"

This does not follow!!!

And as I said before, the size of the ordered region, and the amount of order in our region is too great to be justified by the anthropic principle.

alicorn&robinZ: i talked about ontological parsimony.

In the sense of subtracting an angel (causality) from the head of a pin (our surfboard)? :)

The natural density is for natural numbers. The point is that cardinality is probably not the right thing to look at -- there are more representative notions of the size of a subset (even in the natural numbers).

And a procatalepsis: if P(O|T) is not small--perhaps because orderly areas are more prominent than chaotic areas--then tell us how probability is determined, such that this orderly area is so likely. If you can't do that, then you know less than just about anybody.