Dan Says - Simulated Universe is Impossible (Undecidable)

[Cigarboxal]

If some things can be simulated, then all things can be simulated.

[MDEMONIC689]

Not everything in the universe would have to be computed, it might only compute what we see when we attempt to observe, so atoms, trees, moons, planets etc only exist when you look at them, in videogames if you turn quickly some objects blip in and out because in videogames things really do only exist when observed to save computation power.

You only require the intelligent beings in the universe to believe it is real.

Perhaps someone hacks it sometimes and gives us crop circles and pyramids.



[Entertainmentwf]

@MDEMONIC689 The thing is those items will have a computational cost. The computational cost is not just time, but space. Thus it must be computed. Even if you stored those in a data structure or file. I don't quite follow how that is a counterargument seeing I'm discussing things not possible to be solved in theory of computation, and why our universe can't be. Undecidable problems are things that throw hoops for computation independently from machines.

[MDEMONIC689]

This of course still ignores solipsism, solipsism in philosophy leaves a bad taste in the mouth, but in terms of simulated reality it means a lot if thinking about computational power since it only requires me to exist.

I'm sure the computational power of 1 human brain is not far out of reach,but i hate solipsism as does anyone except the selfish.

So question should be: what is the computational power of ALL living observers today? all the people and animals and insects, NOT the whole universe.

[Entertainmentwf]

@MDEMONIC689 But the thing is there are many uncomputable things in our universe.  Heck look at us. We can decide things no machine can ever do. This is why computation has undecidable theorems. Computation is a one to one correspondence with algorithms, and if the Church-Turing Thesis is infact true, it's definitely an impossibility.

One can easily augment your problem to a computational geometry problem of plane clipping problem then apply the same reasoning again. Mortality Problem...

[MDEMONIC689]

@Entertainmentwf

Im looking it from the philosophical and science angle.

By science i mean specifically as a go between the copenhagen and many worlds interpretation of quantum mechanics I just find it interesting that it looks like it sits inbetween both, schrodingers cat is an interesting thought experiment and it did make me think about why there is such a strong observer effect, The delayed choice quantum eraser experiment is of particular interest.

[MDEMONIC689]

According to prof david deutsch a quantum computer would have infinite parallel processors because its working across the multiverse, that does not sound like a turing machine, he says turing devices collapse the wavefunction and thats the issue.

I don't know if hes a lunatic or not because i don't understand the computational stuff but if he is correct then a quantum computer is actually working with other versions of itself, gives rise to infinite calculations per second.

[Entertainmentwf]

@MDEMONIC689 like I said... that has nothing to do with computability. Quantum computability doesn't contradict the foundations of computing. Even if that were true, that doesn't resolve computability. I am quite sure that isn't true since most quantum algorithms use finite processors. Quantum computation has it's own model of a Turing Machine called the Quantum Turing Machine. If we had one we could get linear time integer factorization but I will wait till it's actually constructed.

[Lok783]

How do you know this universe wont halt?

[Entertainmentwf]

@Lok783 ... I just explained that in this video...

[TheSkroatz]

just because be wont be able to simulate... doesn't mean we are real and someone/something unimaginable in a different state of existance hasn't created us... i.e we could be a supercomputers consiousness... big bang... creation of program.. who knows

[Entertainmentwf]

@TheSkroatz The Simulation Hypothesis is specifically a simulation, that implies a computational model had formulated the product of that simulation. I didn't say anything else actually. I directly was talking about how a simulated universe is not possible by corollary that if there exists one, such a problem is undecidable, therefore a computational entity can't do it! So computers, or even a computer with infinite amount of space and time cannot do it either! So computation is out.

[Apeironization]

@Entertainmentwf Can you be more specific, how can you beat simulation universe hypothesis, can you explain it to a layman 8ok I'm amateur astronomer, but I still consider myself a layman)? I've been looking for evidences around the internet to find evidences that universe is not a simulation, and I found you, could you help me out with some examples? Thanks.

[Entertainmentwf]

@Apeironization machine executes -> encounters paradox -> cannot decide process -> ineffective computational process. That's about as laymen as it gets. There is no way for any machine to decide whether a process takes a very long time to compute or is fundamentally undecidable. Undecidable statements occur on a highly regular basis in the real world if say this were to simulate human beings. Brains alone can do things no machine can. This is a very well known result in computation.

[Apeironization]

@Entertainmentwf This argument beats everything, but there is one more argument, I don't know if you will agree: If universe was a computer simulation than computer codes will be able to create real atoms in this real world, molecules, human brain and etc. with just using and creating computer codes. If they are so smart they could figure out to create code in computer simulation which can create the real atom in this real world/universe for example, but they can't. Opinions?

[Entertainmentwf]

@Apeironization I am not opposed to the idea of people constructing computational models for atomic particles. The problem comes from a universe that is able to perceive undecidable results. Surprisingly this has a strong premise in computation. Induction is required in computation from the simplest while loop to the ability to feed instructions. It is fundamental to understand that paradoxes are impossible to resolve in computation. Brains can pull back from a paradox on the other hand.

[Apeironization]

@Entertainmentwf I didn't mean that. My question was merely can quantum computer create an atom with all the properties that the real atom has and have effect outside the computer in the real world? It can't. Physicists have lost their way in trying to explain this simulation-they say atoms and molecules are bits-0 and 1, like in quantum computer, I don't know why do they think that at all..., it seems to me scientists are no longer objective.

What do you think?

[Entertainmentwf]

@Apeironization That's a good question, I'm not specialized in Quantum Computing, I'm specialized in Computability, Computational Complexity, and Computational Discrete Mathematics. But I could say some things:

-real numbers are not computable

-true random numbers are not computable

-I think they are meaning in a position and time, but I agree it is fairly suspicious that somebody can do this unless it has some deterministic nature (which obviously doesn't match up)

Overall, I think it's fishy

[Entertainmentwf]

@TheSkroatz The Simulation Hypothesis is specifically a simulation, that implies a computational model had formulated the product of that simulation. I didn't say anything else actually. I directly was talking about how a simulated universe is not possible by corollary that if there exists one, such a problem is undecidable, therefore a computational entity can't do it! So computers, or even a computer with infinite amount of space and time cannot do it either! So computation is out.

[Entertainmentwf]

@eabod ... *sigh* jeez louise you need to learn a lot more about computing. "I don't see why", jeez louise, THEY MUST BE ABLE TO AVOID THEM ALL. DO YOU COMPREHEND, THAT'S IMPOSSIBLE. There are infinitely many of them. Watch the bloody video o.O.

[eabod]

Why do "the sets need to know if they're inside themselves"?

[Entertainmentwf]

@eabod watch the video. You will never achieve this without avoiding 3 problems: computation of reals, halting problem, mortality problem.With such an algorithm even though you are correct there is, you still will encounter dealing with machine epsilon and the loss of information. Considering you just introduced another undecidable theorem to make that 'solution' since how does one decide how much precision is needed in each instance without loss of info. Go learn some theoretical comp. sci..

[Entertainmentwf]

@eabod The set of everything is paradoxical and is a common occurrence in mathematical logic if you encounter a paradox. Theory of Computation is full of infinitely many paradoxes that are unavoidable which arise from these concepts. Have you not heard of that before? You just asked me why I was 'invoking' Russell's Paradox when you just asked me this in a comment which is THE SAME THING. You can refer to my comments for the proof by simple abstraction. This is not hard to prove...

[Entertainmentwf]

@eabod Did you know a set can contain itself? ... *sigh*, in a formal system such is possible unless you are using ZF set theory, but computation does not operate under that set theory, it operates under naive set theory. Okay axioms of program are ordered elements of a set. How does the program prove that itself is going to halt? Neverless if an arbitrary set is fed to itself if it's going to not contain itself. It's decidability is actually undecidable because of the same reasoning I gave.

[Entertainmentwf]

@eabod Instead of commenting with me, watch the video.I gave some pretty much concrete scientific information..besides I have more important things to do than to teach you. I'm sorry, but just accept the facts, it makes crystal clear sense if you have the theoretical computer science/mathematical background.Although if you are not familiar with computability theory the reasoning I give will go over your head. Any of the ideas I gave you and in this video are trivial ways to make a '3 liner'.

[Keitaro2011]

I would think that a simulated univere would be impossible simply because it would take a CPU as big as, if not bigger than the universe.

[Entertainmentwf]

@Keitaro2011 well yes but, logically such a CPU could exist technically. Besides it's not just the CPU, there is more needed for a CPU but you could even assume a 'magical' CPU which could. It's the computational limits of all which prevent the concept. Nice idea though :).

[eabod]

@Keitaro2011 I don't think that's true. In general, many complex things can be generated by simple programs.

[Entertainmentwf]

@eabod exactly, but no simple program or infintiely many programs can do the job. It would run into undecidability very quickly in terms of the science through a simple extension of the Church-Turing Theorem (otherwise known as the halting problem).

[eabod]

@Entertainmentwf But the universe is finite, so why can't all its data just be hard-coded in the program? I don't see how the halting problem is relevant.

[Entertainmentwf]

@eabod You cannot hard code infinite concepts. That is an ignorant statement to make about computation. The halting problem is RELEVANT TO EVERYTHING in computation. The whole idea of computation is to utilize the halting problem in programs we KNOW will halt. Refer to my video description for more details. How would you exactly 'hard-code' such?  Secondly who said the universe is finite? Regardless if it is or not does not have any influence to the fact it's uncomputable. Trivial to prove

[Entertainmentwf]

@eabod It is trivial to show you cannot simulate the universe as a simple corollary of Russell's paradox, or Cantor's theorem in regards to the reals being not countable. I think you are confusing simulating a universe with forming a model to simulate a RESTRICTED universe that is not an approximated model of our own. You just cannot do it by the fact that if you wish to model the whole universe, you must simulate also any concept that is inherited by the universe itself through Logic.

[eabod]

@Entertainmentwf I don't see how you're invoking Russell's paradox or Cantor's theorem. Are you saying that every real number would have to be hard-coded?

[Entertainmentwf]

@eabod lol... There is infinitely many ways to disprove the claim. You can invoke both of those for a 'three liner proof'. As I describe in the video very well. You just construct a universe to encode an infinite stream of bits to code a real. Ends up that's impossible. Reals are not countable, thus impossible. That is where diagonalization is used. Russell's paradox is an easy way to approach any uncomputable problem to extend to this problem by assuming an undecidable problem must exist.

[runtubefantasy]

all that hurt my brain .lol I had a little idea of what you were talking about, I think. lol are you in college? if so, what are you majoring in? you seem like you;re passionate about teaching lol

[Entertainmentwf]

@runtubefantasy I'm finishing up my university science degree majoring in Computer Science (specialization in Theoretical Computer Science, and Human-Computer Interactions, and Graphics), minoring in Mathematics. I'm planning to go to grad school next year to work in combinatorial computing, and theory of computation. I love teaching :). I want to become a professor one day and teach. If it helps it took me a long time to learn this material properly (in turn a lot of migranes LOL).