True random number generator

[Charla]

Khamsin could have meant the question either way, so we can't be sure which interpretation is "misreading it". The way he phrased it, the
return value++;
answer is a valid counterexample, because he didn't specify the meaning of "run", so it's perfectly valid to read it as "a function that will return a different value on each execution".

Now, the far more significant question (and the one that would have to do with true randomness as opposed to pseudorandomness) would be "can a program return a value that is not deterministically generated from its inputs, if its inputs includes the hardware on which it is executed, the exact time of its initiation, and the number of cycles it takes between execution and return." As far as I know, the answer to that is no, but as I said earlier, a system with hardware based on quantum mechanics would be intrinsically nondeterministic, and therefore would need no "external" inputs to generate nondeterminism as current technology would.

[aurik]

Khamsin could have meant the question either way, so we can't be sure which interpretation is "misreading it". The way he phrased it, the
return value++;
answer is a valid counterexample, because he didn't specify the meaning of "run", so it's perfectly valid to read it as "a function that will return a different value on each execution".

No. He was very clear that everything it reads in from outside sources is constant. A function which depends on non-constant static data to generate varying output invalidates that precondition.

[Devek]

And input is input.. it doesn't matter if it is quantum mechanics, the clock on the computer, or random "words" gleaned from myspace pages.

[Khamsin]

Aurik and Devek got what I meant.

On a side note, here's another question:

If you could save and load states like a video game console emulator, but you could do it for the universe such that the location, momentum, spin, charge, potential energy, kinetic energy, etc. of every single subatomic particle were preserved and re-loaded at will to the exact state everything was at (including chemicals in the brain, electric pulses in the nervous system, etc.) would it be possible to have a different outcome everytime you load the same state? Is it possible for *anything* to be different? Is there such a thing as random?

[Devek]

The meaning of "random" is subjective. If you can't predict the outcome it is random... if you knew EVERYTHING nothing would be random, in theory anyway.

To answer your question, I'm going to have to defer that to someone who knows more about quantum mechanics than me.

[aurik]

The answer is that the question is moot because you can't know everything. http://en.wikipedia.org/wiki/Uncertainty_principle

[Khamsin]

Except this isn't a question asked from an anthropocentric point of view. It doesn't matter if anyone knows anything. It just matters whether there could be a different outcome each time the state is reloaded. Nobody ever has to have any knowledge of it.

[aurik]

Except this isn't a question asked from an anthropocentric point of view. It doesn't matter if anyone knows anything. It just matters whether there could be a different outcome each time the state is reloaded. Nobody ever has to have any knowledge of it.

As I said, it's moot because even if it would be possible to "reload" it would be impossible to verify that the "reload" was correct.

[Charla]

According to current theories, even if the entire "current" state was reproduced somehow, some things can exist simultaneously in multiple states. That's what Schroedinger was really trying to say with his famous cat analogy. The cat really was both alive and dead at the same time, because it was an analogy for a quantum particle, which can have both states at once and only ceases to do so when it is observed. In other words, as far as we can tell, the universe itself is truly nondeterministic, so that state save could be different the second time. Of course, relatively recent scientific theories are sometimes overturned by newer ones, so we might not understand it all in the first place.

nerding it up even more:

In quantum mechanics it is the amplitudes of wavefunctions, or state vectors, that add. It occurs when an object simultaneously "possesses" two or more values for an observable quantity (e.g. the position or energy of a particle).

[Khamsin]

It was my impression that Schroedinger wasn't saying something actually was in multiple states at once, only that something could be in any one of a number of states, and we just don't know which one exactly until we observe it. The cat is either dead or alive, not both. We just don't know which it is until we look (which would interfere with the experiment). As such, the best we can do is use probabilities to make an educated guess as to what state something is in.

Likewise, with the Heisenberg Uncertainty Principle, we can't know everything about a system at once, but that doesn't mean those parameters aren't defined. A particle simultaneously has an exact momentum and position. We just don't know them, and attempting to know them would change them. Again, probabilities are used to make educated guesses since we can't know for sure.

Probabilities are only in reference to human understanding and observation. They're acknowledgements to the limitations of our ability to accurately observe without interfering with the results.

All of those theories and principles are anthropocentric in nature. They deal with what we can know and observe, not with what is. The entire universe is precisely defined at any moment in time. It's just that nobody knows anything about the vast majority of it. Just because we close our eyes, doesn't mean the rest of the universe disappears. Electrons are going to be where they're going to be. Us looking at them will interefere with them and change the results from what they should have been, but before we looked, they were doing exactly whatever they were doing. Will anyone ever know what that was? Does anybody need to ever have known what that was for it to have actually happened?

So in that sense, no, I don't believe anything in the world is literally in multiple states at once other than merely as an idea for the purposes of describing what we can and cannot observe. And if something were literally in multiple states at once, assume that, too, was preserved in the save state loading of the universe.

[aurik]

As I said, even if you could reconstruct a situation exactly, you couldn't verify that you did, so any variation of result could be "randomness" or it could be a flaw in reconstructing. Therefore you can't ever know the answer to the question.

[Charla]

And if something were literally in multiple states at once, assume that, too, was preserved in the save state loading of the universe.

That's what I said. If it is in multiple states at once, that constitutes nondeterminism. Trying to decide whether superposition is really an artifact of perception or a reflection of an unobservable objective reality is beyond our capabilities, and therefore the question could be construed as unanswerable. Of course, there are also some very strange theories that perception itself creates reality, and therefore the entire concept of a "non-anthropocentric" state would be contradictory.

[Khamsin]

And if something were literally in multiple states at once, assume that, too, was preserved in the save state loading of the universe.

That's what I said. If it is in multiple states at once, that constitutes nondeterminism. Trying to decide whether superposition is really an artifact of perception or a reflection of an unobservable objective reality is beyond our capabilities, and therefore the question could be construed as unanswerable. Of course, there are also some very strange theories that perception itself creates reality, and therefore the entire concept of a "non-anthropocentric" state would be contradictory.

Well, it is unanswerable, since you yourself would be reloaded to your previous state, so you wouldn't even know you were supposed to be checking the outcome, let alone be able to tell that it was any different than before.

I guess a better question to ask would be, are the outcomes in the universe simply defined by the action and reaction of all the components within it, in such a way that there are no truly random events? For example, flipping a coin should be 50/50, but is the side it lands on determined merely by the force and angle at which it was flipped, the medium it was flipped through, and all the millions of various other factors that influenced it during its flipping? Now take that concept to the quantum level, and we can ask if the state of anything is a result of the influences of everything else (observable and unobservable), and would have been predictable if we had some way of knowing everything about the universe at once without interfering with it.

It may be a moot question, since all we care about are the things we can actually know and observe, and how it affects us. But I don't believe the universe is anthropocentric, because I'm pretty sure that if there were nothing responsive to stimuli in the entire universe (let alone humans), and it was all just lifeless barren rock, it'd still exist and carry on doing its thing. It's just that, from a human point of view, which we will always have (being humans and all), the universe will be anthropocentric in perception. But that's not what it "is".

[Charla]

Whether or not it's a moot question, it's beyond our current knowledge, meaning we've just left physics and entered philosophy.

[Devek]

It is all a computer simulation anyway and our bodies are somewhere else being harvested for energy.

[Enkidu]

It was my impression that Schroedinger wasn't saying something actually was in multiple states at once, only that something could be in any one of a number of states, and we just don't know which one exactly until we observe it. The cat is either dead or alive, not both. We just don't know which it is until we look (which would interfere with the experiment). As such, the best we can do is use probabilities to make an educated guess as to what state something is in.

He was saying something actually was in multiple states. The fact is until the wave function collapse (or appearance of collapse) there are multiple states. Without decoherence (impossible with a cat, but Schrodinger just used it as an example), you could really put something into "two states". We've done it before with rubidium atoms and it can even be measured with quantum nondemolition, the uncertainty principle surely still exists, but like all things in the quantum world you can avoid it if you're clever .

[Charla]

It is all a computer simulation anyway and our bodies are somewhere else being harvested for energy.

http://www.machall.com/comics/20030514.jpg

(anyone else still remember machall?)

[EvilSid]

This could happen in ffxi too if you knew which random number generator it used. Collect a long enough string random numbers and you'll be able to figure out the seed and position and predict the next numbers. Practically this is impossible because so many random things are going on (monster movement, NM spawn times, etc.

EPIC POST!

I always wondered how a computer can generate a random number, and now i know it simply cannot, because it's a deterministic machine.

So now I wonder what's the seed used by FFXI's server in the algorythm used for generating the numbers with /random command (i guess the same numbers are used by system for NM spawns, drop lottery, weather pops and such random things). I thought it could be the clock, but theorically 2 /randoms calculated at the exact same time should give the same number. This can't be proven because of connection lag btw...

[Mizango]

. . . .

[aurik]

necro