In which I falsify the Level IV Multiverse

[Toontown]

I won't argue that a rudimentary simulated universe is not possible. I will argue that a comprehensive simulation of a universe would require a computer bigger than the universe, using more energy than the universe. And I will argue that such a comprehensive model would be required to produce sentient beings.

http://discovermagazine.com/2013/dec/09-do-we-live-in-the-matrix#.UyN2Y_ldVV0

I don't think it follows that daily miracles, observed by sentient beings, would likely be happening in these rudimentary simulations. Take immutable laws of physics out of the simulation, and what do you have? Certainly not a simulated universe. A cartoon, at best.

Are sentient beings possible in one of these rudimentary toonverses? I doubt it.

 

[drelda]

I won't argue that a rudimentary simulated universe is not possible. I will argue that a comprehensive simulation of a universe would require a computer bigger than the universe, using more energy than the universe. And I will argue that such a comprehensive model would be required to produce sentient beings.

But would the simulation even have to be run for the sentient beings to have their moments of awareness?

Imagine you are using a fractal browser to zoom into a Mandelbrot set. Imagine that somewhere in the depths of that you come across complex structures and work out that you are looking at a network of structures equivalent to a brain - and a sentient brain too. Deep in the pure maths of the Mandelbrot set you discover what looks like a definition of a moment of consciousness.

The MUH claims that this moment of consciousness would actually be experienced by the inhabitant of the Mandelbrot set. However the fact that you browsed down to look at it (equivalent to running the simulation) isn't necessary to make this happen - the contents of the Mandelbrot set exist whether you calculate them or not.

Now the Mandelbrot clearly doesn't really have this kind of complexity - but replace its equations with the equations of a Grand Unified Theory of physics and follow the same logic. If you simulate those GUT equations you would see the structures of physical reality unfold - including us eventually.

But maybe you didn't need to run the simulation for us to 'exist' - maybe the fact that the outcome of the simulation exists mathematically is enough.

If so then Max Tegmark's MUH is true - and we live in the Level IV multiverse.

- Drelda

 

[Toontown]

But would the simulation even have to be run for the sentient beings to have their moments of awareness?

Well, my argument is that a comprehensive simulation of a universe complex enough to contain sentient beings cannot be run, because running it would require a computer bigger than the universe, using more energy than there is in the universe. IOW, an actual universe is the computer required to run itself. But that doesn't mean a universe cannot fundamentally be a mathematical structure.

And I really don't think the way out of that is as simple as you propose. There are programs on my hard drive that I can access and look at. However, I notice that looking at their structures is not the same as watching them run, or participating in their output. And the notion that a program capable of accurately simulating my entire computer could fit on it's hard drive seems to me to defy logic.

Perhaps universes exist in which such simulations are possible, due to radically different laws of physics. However, Robin has done nothing to show that such universes would be typical at all, so I don't see why Robin expects to be seeing the laws of physics broken on a daily basis if the MUH is true.

 

[Robin]

Your argument from incredulity is a logical fallacy.

Where exactly do I give an argument from incredulity???

 

[Robin]

I won't argue that a rudimentary simulated universe is not possible. I will argue that a comprehensive simulation of a universe would require a computer bigger than the universe, using more energy than the universe. And I will argue that such a comprehensive model would be required to produce sentient beings.

http://discovermagazine.com/2013/dec/09-do-we-live-in-the-matrix#.UyN2Y_ldVV0

I don't think it follows that daily miracles, observed by sentient beings, would likely be happening in these rudimentary simulations. Take immutable laws of physics out of the simulation, and what do you have? Certainly not a simulated universe. A cartoon, at best.

Are sentient beings possible in one of these rudimentary toonverses? I doubt it.

But according to the MUH every possible run of every possible algorithm exists. It is not a matter of these being in any physical universe, they just are - part of some sort of Platonic realm. Universes are just one of the objects in this Platonic realm.

I can't see how a full universe simulation would be necessary to produce sentient beings, but even if this was the case, this would still be no problem.

 

[Robin]

Perhaps universes exist in which such simulations are possible, due to radically different laws of physics. However, Robin has done nothing to show that such universes would be typical at all, so I don't see why Robin expects to be seeing the laws of physics broken on a daily basis if the MUH is true.

As I said before, these simulations do not need to be in any sort of Universe at all - according to the MUH they simply are, since they are mathematically possible.

In this case a fantastical universe is just as probable as a seemingly orderly one, but there are more ways for a fantastical universe to happen than there are for a non-fantastical universe to happen - since non-fantastical universes must accord to a set of laws and fantastical universes don't have to.

 

[Robin]

I think your argument assumes a particular 'measure' - ie a way of ordering an infinite set of universes. There was some discussion of this recently here.

I don't think you falsified the MUH (or any of the other levels of multiverse which would all have the same issue) - but I think you have falsified that this particular measure is the one that drives what we observe in reality.

To illustrate the problem - imagine you ask me what percentage of the integers are even. I answer 2/3 and prove it by showing that when I order the integers in this way :

1, 2, 4, 3, 6, 8, 5, 10, 12, 7, 14, 16, 9, 18, 20...

This list will contain every integer - but it looks like there are twice as many even numbers as odd numbers. In fact for every odd number in the list you can see that there are always 2 even numbers.

This is what you did essentially when you argued that for every 'normal' universe there would be 2 or more 'godzilla' universes - you constructed an ordering (measure) that did this - but that result was entirely dependent on the measure used.

For integers we have a natural ordering - so its easy to say there is a right answer. However for infinite sets of universes we don't know which measure to use.

This is the 'measure problem' that Max Tegmark thinks is the "greatest crisis in physics today" - because without the correct measure we can't predict anything.

- Drelda

Only I didn't order the universes at all. Show me where I did that.

I said that there are more ways for it occur if it doesn't have to accord to a set of rules than if it had to accord to a set of rules. I argued that for each and every scenario according to those rules there are many others which don't. That is an entirely different argument.

Similarly I can say that there are as many ways for a number to be even as there are for it to be odd.

On the other hand if I could show that there are more ways for a number to be non-prime than there are for it to be prime (for example if I could show that for each prime there are infinitely many non-primes whose only prime factors are that number) then I could argue that any number picked randomly from any sufficiently large subset of the set of integers is more likely to be non-prime than it is to be prime.

 

[Robin]

Strangely enough, I just got an email that accidentally came through from another of the multiple universes. Clearly it's Obama's fault. Anyway, it's from someone called Nibor, who is, remarkably enough, having the same discussion, and taking the same position as Robin. It says, in part:

I am not sure what your point is. Are you saying that these occurrences break the laws of physics in this alternate world?

I don't see how - is there, in this alternate world, some physical impossibility involved in injecting bodily fluids into another body, or watching a ball being thrown around?

Also, what world would they be talking about where Miley Cyrus is considered a singer?

 

[theprestige]

Where exactly do I give an argument from incredulity???

We've been over this. Your OP. You start by saying that X is extremely unlikely, then conclude by saying that you don't believe something so unlikely can be true, and therefore X is false.

That's an argument from incredulity. It was an argument from incredulity when you tried it six months ago, and it'll be an argument from incredulity when you try again six months from now.

"Extremely unlikely" is rational. "Extremely unlikely and therefore false" is... not.

 

[ehcks]

Miracles are not mathematically possible, therefore universes in which miracles occur are not mathematically possible, therefore computers which model universes in which miracles occur are either not mathematically possible, or are producing fiction.

 

[andyandy]

Interesting post Robin - a melding of Tegemark's views and Nick Bostrom's computer simulation argument.....

Something that I was wondering, not sure I know the answer - if you have 2 different sized infinite groups, and then choose 1 element from one of the groups at random relative to all members, what is the probability of choosing an element from either group?

Obviously we can have different sizes of infinity - say Natural and Real. So if I choose a number at random from all numbers, is it more likely to be only real or natural (and real)?

is it 50:50? Is it more likely to be only real? Does the question even make sense?

This seems related to the OP - where we have different sized infinities, which are used to calculate probabilities....

 

[tableplay]

contraints

That is true Robin, laws generate constraints on the number of universe possibilities, therefore there would have to be more universes that have no laws since the degrees of freedom are not imposed.

Only I didn't order the universes at all. Show me where I did that.

I said that there are more ways for it occur if it doesn't have to accord to a set of rules than if it had to accord to a set of rules. I argued that for each and every scenario according to those rules there are many others which don't. That is an entirely different argument.

Similarly I can say that there are as many ways for a number to be even as there are for it to be odd.

On the other hand if I could show that there are more ways for a number to be non-prime than there are for it to be prime (for example if I could show that for each prime there are infinitely many non-primes whose only prime factors are that number) then I could argue that any number picked randomly from any sufficiently large subset of the set of integers is more likely to be non-prime than it is to be prime.

 

[Robin]

Miracles are not mathematically possible, therefore universes in which miracles occur are not mathematically possible, therefore computers which model universes in which miracles occur are either not mathematically possible, or are producing fiction.

Well, yes, a simulation is, by definition, producing fiction. That would only be a problem if it were not possible for us to be conscious in a fictional world.

Did you ever have a dream? Did you ever have a dream where you could, say, fly, or fall a long distance and not get hurt? If so then you have experienced a simulated (fictional) world in which miracles occurred.

In a simulation the rules can be changed at any time. An inconsistency in the simulated world would not produce an inconsistency in the simulation or the simulator.

 

[Robin]

Interesting post Robin - a melding of Tegemark's views and Nick Bostrom's computer simulation argument.....

Something that I was wondering, not sure I know the answer - if you have 2 different sized infinite groups, and then choose 1 element from one of the groups at random relative to all members, what is the probability of choosing an element from either group?

Obviously we can have different sizes of infinity - say Natural and Real. So if I choose a number at random from all numbers, is it more likely to be only real or natural (and real)?

is it 50:50? Is it more likely to be only real? Does the question even make sense?

This seems related to the OP - where we have different sized infinities, which are used to calculate probabilities....

You are possibly right. Although Chaitin does have a proof that the probability of any number in the interval [0..1] being computable is zero. And computer numbers and non-computable numbers are different sized infinities, so I don't think it is necessarily meaningless.

But in the end I have to admit that the concept of the MUH is not nearly well enough defined for me to make any reasonable estimate of the relative probabilities.

 

[ehcks]

Well, yes, a simulation is, by definition, producing fiction. That would only be a problem if it were not possible for us to be conscious in a fictional world.

Did you ever have a dream? Did you ever have a dream where you could, say, fly, or fall a long distance and not get hurt? If so then you have experienced a simulated (fictional) world in which miracles occurred.

In a simulation the rules can be changed at any time. An inconsistency in the simulated world would not produce an inconsistency in the simulation or the simulator.

A dream is no different from imagination. It is a creation of my own brain that only exists as electrochemical reactions inside my skull. The universe is not inside my skull, nor am I a fictional being inside someone else's. I think, therefore I am.

An inconsistency in experimental results on the state of the universe would be measurable, and would lead to either knowing our understanding of the universe is wrong, or that the universe is wrong.

As for the rest, solipsism is the most useless of philosophical drivel. It should be the very first thing you pass off as pointless.

 

[drelda]

Only I didn't order the universes at all. Show me where I did that.

Here :

I argued that for each and every scenario according to those rules there are many others which don't.

This is an ordering.

You have an infinite number of 'weird' universes and an infinite number of 'normal' universes. How can you say there is 'more' of one than the other?

The only way to do that is to define a measure - an ordering of them so that you can calculate how many are of each type in the first n of that ordering.

Similarly I can say that there are as many ways for a number to be even as there are for it to be odd.

Only because you have a natural ordering of numbers (1, 2, 3 etc). For universes you don't have a natural ordering. This (I think) is the key flaw in your argument - it relies on an particular measure - we could construct an alternative measure that reaches the opposite conclusion.

If however you can justify that your particular measure is the correct way to calculate probabilities in an infinite set of universes - then I would agree with your argument. And a bunch of physicists would be very happy for you to do so - its a huge problem for them that some theories predict infinite sets of universes - but they don't have a clue what measure to use to calculate anything.

- Drelda

 

[Hercules Rockefeller]

Here's an interesting lecture with Nima Arkani-Hamed.

Space-Time, Quantum Mechanics and the Multiverse: