Is there such a thing as randomness?

[CSMR]

Originally posted by: f95toli

Originally posted by: CSMR
How can you call a probability correct? What would it mean for "the particle will be seen here with a 60% chance" to be right?

First of all, in the majority of cases where we use QM to calculate a quantity the expectation values are very sharply peaked, i.e. the part of the variance that is due to "quantum noise" is very small; ususally much smaller than the noise that is due to e.g. thermal fluctuations. Hence, from a practical point of view the uncertainty is rarely a real issue. Even in the field of quantum computing there are single-shot methods that can measure the state of a qubit with nearly unit visibility.

Secondly, the time evolution is a QM system can be complettely deterministic (i.e. there is no "randomness" in the propagator of the density matrix); it is only when we try to MEASURE a observable of that system that the probabilistic nature of QM becomes apparant. Hence, there are lots of systems where we have in a sense "full control" as long as we leave it alone; a trivial example is measurements of Rabi oscillations or Ramsey fringes; there is no way to observe them directly since we can only measure one point at a time (since the measurements itself causes the system to collapse) . However, by repeating our measurement many times and measuring at various times we can "build up" the curves. The expectation values of the level populations are therefore ina sense deterministic observables.

Finally, and this is the most important point.
While we can't e.g measure anything "exactly" there are still many things you CAN measure. The mean and the variance of an observable are the most obvious example.
However, in many systems we can also measure higher order moments meaning it is possible to study various correlation functions. A good example is a Hanbury-Brown and Twiss-type experiments which are not only interesting for fundamental studies but also has some practical applications when it comes to e.g. characterizing single photon sources.

A QM system can result in predictions which have a very high probability (your point 1 and points 2 and 3 which are examples of it - take a lot of measurements and you can say quite a lot about the distribution with a high probability). But there is a fundamental difference between saying "E will happen with probability 99%" and "E will happen". The second is an objective statement whereas the first lacks a definition as an objective fact - unless you or anyone else would care to attempt one.

The same type of experiments can be used to demonstrate violations of Bell-type inequalities, these violations are important since they give importat support for the idea that QM is fundamentaly correct.

Yes some of these high-probability predictions are observed and are unexpected in other theories. I do agree it is a very good theory but calling it correct is going too far.

 

[Blouge]

>>QM is fundamentaly correct.
>calling it correct is going too far.

Just like Newton's F=ma, QM is 100% correct - within its limited domain. By "correct", I mean that none of its predictions (usually statistical ones) have been falsified. Not a single one.

Unfortunately, QM's domain is limited in that QM is incompatible with general relativity, another "correct" theory. That QM requires dirty tricks like renormalization also means that it's mathematically and logically bankrupt. QM is also incomplete. For example, it offers no known way to compute the fine structure constant.

My position is that randomness can logically never be distinguished from plain ignorance or distinguished from deterministic ideas such as Everett's many worlds. All three of these ideas are scientificially unfalsifiable and indistinguishable. "randomness" is a matter of faith, not science. As a matter of faith, "randomness" cannot be defended, but only accepted or rejected per your whim.

So the attempt ot use QM to back up the idea of randomness is an attempt to do the impossible. Furthermore, since QM is compatible with many ontological interpretations, including both randomness and determinism (Bohm or Everett), to arbitrarily pick out the non-deterministic interpretations is an error. Finally, an attempt to back up "randomness" with QM is of little worth because QM doesn't actually match reality: the shortcomings listed above show that, as it stands, QM is only an approximation to reality.

 

[CSMR]

Originally posted by: BlougeJust like Newton's F=ma, QM is 100% correct - within its limited domain. By "correct", I mean that none of its predictions (usually statistical ones) have been falsified. Not a single one.

That is a strange use of words because if you go by it the theories "the world will end in 2050" and "the world will not end in 2050" are both correct.

Unfortunately, QM's domain is limited in that QM is incompatible with general relativity, another "correct" theory.

GR and newtonian mechanics have been falsified so are not correct by your definition.

My position is that randomness can logically never be distinguished from plain ignorance or distinguished from deterministic ideas such as Everett's many worlds. All three of these ideas are scientificially unfalsifiable and indistinguishable. So the attempt ot use QM to back up the idea of randomness is an attempt to do the impossible. Furthermore, since QM is compatible with many ontological interpretations, including both randomness and determinism (Bohm or Everett), to arbitrarily pick out the non-deterministic interpretations is an error. Finally, an attempt to back up "randomness" with QM is of little worth because QM doesn't actually match reality: the shortcomings listed above show that, as it stands, QM is only an approximation to reality.

I think you're taking a philosophy of falsification beyond its domain but yes many world theories and random ones can be interchangeable and so your argument looks right.

"randomness" is a matter of faith, not science. As a matter of faith, "randomness" cannot be defended, but only accepted or rejected per your whim.

First you have to define it. Only then can you accept or reject it. And perhaps then you can defend or attack it? I think some definitions certainly can be defended.

 

[imported_iNGEN]

Originally posted by: PhatoseAlpha
AFAIK, as it stands nuclear decay is currently understood to be dependent upon vacuum energy. Vacuum Energy is a quantam effect, therefore subject to Heisenberg uncertainty, therefore unknowable, therefore random.

No, no! Not random, but having the appearance of random. Heisenberg's principle of uncertainty is only applicable to mechanism that are deterministic. If an event actually is random, then the impact of observation is irrelevant.

 

[f95toli]

Originally posted by: Blouge
My position is that randomness can logically never be distinguished from plain ignorance or distinguished from deterministic ideas such as Everett's many worlds. All three of these ideas are scientificially unfalsifiable and indistinguishable. "randomness" is a matter of faith, not science. As a matter of faith, "randomness" cannot be defended, but only accepted or rejected per your whim.

I think we are talking about different things here. I always draw a sharp line between physics (and science in general) and philosophy and I generally don't "trust" the latter.
The "scientific" answer to the question "is there such a thing as randomness" is yes, to the best of our knowledge there are truly random phenomena.

The "philosophical" answer is that there is no way to know for certain. HOWEVER, just as in the case of EPR it is possible that we someday might come up with some ways of testing this. Einstein would probably have been very surprised to learn that there was a way to experimentally test the "paradox" (but I do think he would have acceptated that he was wrong, as he did with the expanding universe).
Remember that between the EPR paper and the first Bell type equalities (and subsequent experimental tests) a lot of ink was wasted trying to argue "for or against"; I doubt that the participants in THAT discussion ever expected to get a definitive answer (since most philosophical debates are never settled).
Don't get me wrong; there is nothing wrong with philosophy but don't confuse it with science. In many cases (and I believe this to be one, at least for now) the discussion is meaningless if the aim is to reach a final answer.

Btw, I might be wrong; but isn't Bohm interpretation in some trouble due to the fact that it doesn't really fit with quantum field theory? If so it just goes to show that what I have written above is correct, perhaps we will someday be able to rigorously test which interpretation is correct.

CSMR: I think we have different definitions of "correct". I simply mean that QM can correctly predict the outcome of my experiments; that is all that is important if you ask me (and Popper).

Also, "E will happen with probability 99%" IS an "objective fact" according to me (and most other scientists); there are uncertainties in all experiments and even in "classical" physics you would get results that can strictly speaking ONLY be interpreted in a probabilistic fashion; the variances are sometimes very small but they are never zero in a real experiments; there is no such thing as an exact measurement which is why statistics is so intimately linked to metrology.

(There are a few things that are exact BY DEFINITION, e.g. the speed of light but that is not really relevant for this discussion)

 

[Blouge]

and philosophy and I generally don't "trust" the latter.

I'm not sure what you mean by philosophy, but I think you are saying that you don't trust logic. That's quite normal and understandable; people have operated on the basis of irrationality, superstition, and ignorance for millenia.

>Just like Newton's F=ma, QM is 100% correct - within its limited domain. By "correct", I mean that none of its predictions (usually statistical ones) have been falsified. Not a single one.

That is a strange use of words because if you go by it the theories "the world will end in 2050" and "the world will not end in 2050" are both correct.

My comment is not strange at all; F=ma is taught all of the time at universities and entire industries rely on it (e.g. everything from civil engineering to space exploration and video game simulations). QM is useful in other, typically microscopic, situations. They both have their theoretical and practical limitations, and both fail to achieve a complete description of reality.

>"scientific" answer to the question "is there such a thing as randomness" is yes,

If by "randomness", you mean ignorance and an ability to predict only statistically, then yes I agree randomness exists. This is analagous to how Brownian motion was understood before.

If you mean that the universe is inherently random, then I disagree. Our knowledge of physics is far beyond what was known 100 years ago, and we are retarded compared to what will be known 100 years from now. Did you know that an entirely new particle, the axion, was discovered only about 60 days ago? It's hubris to think that a passing knowledge of a contemporary, flawed, limited theory is sufficient to establish whether or not the universe is deterministic - a task that's logically impossible anyway.

 

[f95toli]

Originally posted by: Blouge

and philosophy and I generally don't "trust" the latter.

I'm not sure what you mean by philosophy, but I think you are saying that you don't trust logic. That's quite normal and understandable; people have operated on the basis of irrationality, superstition, and ignorance for millenia.

No, I am saying that there are plenty of examples where "logic" (as you put it) has been used to reach conclusions that were later shown to be wrong. Again, EPR probably appeared very "logical" when it was published but it Einstein&co were neverthless wrong.


>"scientific" answer to the question "is there such a thing as randomness" is yes,

Originally posted by: Blouge
If by "randomness", you mean ignorance and an ability to predict only statistically, then yes I agree randomness exists. This is analagous to how Brownian motion was understood before. FYI, our knowledge of physics is far beyond what was known 100 years ago, and we are retarded compared to what will be known 100 years from now. Did you know that an entirely new particle, the axion, was discovered only about 60 days ago? I'm amazed at such hubris - that a passing knowledge of a contemporary, flawed, limited theory is sufficient to establish whether or not the universe is deterministic - a task that's logically impossible anyway.

I am not sure what you mean by your "Brownian motion" example. Brownian motion is more or less the quintessential statistical theory. I you work through the theory of Brownian motion (starting with the hamiltionian for a classical particle copupled to a heath bath of harmonic oscillators) you end ut with a set of expressions for e.g the MEAN distance from the starting point, the variance etc. Brownian motion is perhaps the LEAST deterministic of all classical theories. The mere fact that the results depends on temperature explicitly means that it is a probabilistic theory.

It is incidentally possible to derive similar expressions for Brownian motion using the full QM theory (staring with a Caldeira-Legget Hamiltionian). In the high temperature limit these experssions agree with the classical results (since kbT>> hbar w). At low temperatures they DO give different results and it turns out (not surprisingly) that the full QM theory agress with the experiments.

Also, I am fully aware that we are learning more about nature every day.
If I thought we had learned everything we could I would probably stop working as an experimental physicist. Learning new things is what drives me and most of my collegues.

 

[Blouge]

>I am not sure what you mean by your "Brownian motion" example
>Brownian motion is perhaps the LEAST deterministic of all classical theories.

Einstein made a model for Brownian motion, based on random collisions. Using the model, he could make statistical predictions that agreed with experiment. Yes, the model contains randomness. But Einstein didn't then proclaim that he had proved that the universe itself is inherently random. I guess I can see how a lesser scientist, ashamed of their own ignorance, would resort to adopting a philosophical concept like indeterminism. Belief in The Gods has fallen out of fashion and might not be politically-correct nowadays.

>I am saying that there are plenty of examples where "logic" (as you put it) has been used to reach conclusions that were later shown to be wrong. Again, EPR probably appeared very "logical" when it was published but it Einstein&co were neverthless wrong.

I would say that they their approach was basically correct, but needed some adjustment (i.e. allowing the variables to be non-local).

Edit: Sorry... I think I should be using the term "indeterminism" (a belief in randomness as a fundamental property of the universe) more often. I have no problem accepting that dice rolls are random from a practical point of view; I've been trying to argue against belief in indeterminism. I've been a fan of Einstein and a determinst for a long time, but my own arguments are slowly convincing me to abandon that belief since it is equally unscientific.

 

[CycloWizard]

Originally posted by: f95toli
First of all, in the majority of cases where we use QM to calculate a quantity the expectation values are very sharply peaked, i.e. the part of the variance that is due to "quantum noise" is very small; ususally much smaller than the noise that is due to e.g. thermal fluctuations. Hence, from a practical point of view the uncertainty is rarely a real issue. Even in the field of quantum computing there are single-shot methods that can measure the state of a qubit with nearly unit visibility.

Secondly, the time evolution is a QM system can be complettely deterministic (i.e. there is no "randomness" in the propagator of the density matrix); it is only when we try to MEASURE a observable of that system that the probabilistic nature of QM becomes apparant. Hence, there are lots of systems where we have in a sense "full control" as long as we leave it alone; a trivial example is measurements of Rabi oscillations or Ramsey fringes; there is no way to observe them directly since we can only measure one point at a time (since the measurements itself causes the system to collapse) . However, by repeating our measurement many times and measuring at various times we can "build up" the curves. The expectation values of the level populations are therefore ina sense deterministic observables.
Hence, while we can't "observe" it continuosly we can still know where the system will be at a given time. This is why it is possible to e.g. use shaped microwave pulses to manipulate the state of an atom or molecule in NMR in a determinstic fashion.

Finally, and this is the most important point.
While we can't e.g measure anything "exactly" there are still many things you CAN measure. The mean and the variance of an observable are the most obvious example.
However, in many systems we can also measure higher order moments meaning it is possible to study various correlation functions. A good example is a Hanbury-Brown and Twiss-type experiments which are not only interesting for fundamental studies but also has some practical applications when it comes to e.g. characterizing single photon sources.
The same type of experiments can be used to demonstrate violations of Bell-type inequalities, these violations are important since they give importat support for the idea that QM is fundamentaly correct.

Nice post. I think I learned more about QM from your post than from my physical chemistry class way back in the day. At least, it made more sense this time around. :beer:

 

[CycloWizard]

Originally posted by: Blouge
>I am not sure what you mean by your "Brownian motion" example
>Brownian motion is perhaps the LEAST deterministic of all classical theories.

Einstein made a model for Brownian motion, based on random collisions. Using the model, he could make statistical predictions that agreed with experiment. Yes, the model contains randomness. But Einstein didn't then proclaim that he had proved that the universe itself is inherently random. I guess I can see how a lesser scientist, ashamed of their own ignorance, would resort to adopting a philosophical concept like indeterminism. Belief in The Gods has fallen out of fashion and might not be politically-correct nowadays.

Einstein believed in determinism. Einstein believed in a lot of things. Einstein also knew that he didn't know everything and would admit when he was wrong. I don't know if the natural universe abides by determinism or indeterminism. The bottom line is that you don't either, nor did Einstein. Attacking someone for arguing one way or the other is plain ridiculous. Someone should introduce you to the vacation stick until you can learn some manners. f95toli is probably the most helpful and least argumentative person in these forums, yet here you are waging personal attacks against him when it's pretty clear, at least to me, that he knows a bit more about this than yourself.

 

[StormRider]

Originally posted by: alimoalem
to those that believe in God, no, there is no randomness. God is all knowing and knows where every one of His atoms are and what they're doing.

to those that don't believe in God, then yes, there is such a thing as randomness. there are some things that we (as humans) will never understand or know

But what about Free Will? Didn't God give us Free Will? With Free Will, God doesn't really know how we'll act either.

To me, Free Will is another face of randomness... when we write computer games, we can either have all NPCs follow a script (meaning they do the same thing in the same situation every time) or insert some randomness to simulate "free will" in these NPCs (flip a "weighted" coin, if it comes up heads then do this, otherwise do that).

 

[CSMR]

Originally posted by: f95toli
CSMR: I think we have different definitions of "correct". I simply mean that QM can correctly predict the outcome of my experiments; that is all that is important if you ask me (and Popper).

To determine whether that is true we need to know what QM "predicts". At best a QM model associates particular probabilities with various outcomes of your experiment. Now it has to be said what its "prediction" is.

Also, "E will happen with probability 99%" IS an "objective fact" according to me (and most other scientists);

You and some other scientists may make statements like this but I think you should explain what it means as an objective fact if you are to make a case for it being an objective fact.

(I am not sure whether 99% is specially chosen to be close to one. Of course you can't take the predictions that QM makes with 99% certainty as "the predictions of QM" or else you will come up with a lot of false predictions.)

there are uncertainties in all experiments and even in "classical" physics you would get results that can strictly speaking ONLY be interpreted in a probabilistic fashion; the variances are sometimes very small but they are never zero in a real experiments; there is no such thing as an exact measurement which is why statistics is so intimately linked to metrology.

Yes probability is used and is useful in the practice of science. That is not the question but whether the theories produced with it are "objective fact".
Science produces theories of all sorts and our sorts of theories take place in a mathematical framework that includes probability. For some theories there is a tangible correspondence between the model and the "real world" of observed phenomena. Such theories can be objectively right or wrong. For most theories the relation is less straightforward and more ambiguous. Probabilistic theories fall in this category - until someone is able to provide a definite correspondence.

Since you bring up statistics I don't know a statestical model that even claims to be "the objective truth" about a subject. Not only would you have to believe in objective probability to assert this but that your model whether linear or more complex is just right and the error terms are given just the right (generally normal) distributions.

 

[CycloWizard]

Originally posted by: CSMR
You and some other scientists may make statements like this but I think you should explain what it means as an objective fact if you are to make a case for it being an objective fact.

(I am not sure whether 99% is specially chosen to be close to one. Of course you can't take the predictions that QM makes with 99% certainty as "the predictions of QM" or else you will come up with a lot of false predictions.)

I don't do anything with QM. My experiments right now are determining the mechanical properties of various parts of the eye. I'm validating my experiments using pig eyes, which are all killed at exactly 6.5 months of age and 243 pounds. All of them are also the same breed ('GenetiPork®' ). I can't predict the mechanical properties with greater than 85% probability by any method, simply because there is that much inter-sample variation. Still, I can publish this data and it will be accepted. Is the source of this error truly random? Does it have an environmental basis? There's no way to tell. The point is that, simply because the R^2 value of my correlation is not identically 1.000000 doesn't imply that the model is wrong or that it doesn't tell us things. It can be used to make predictions (e.g. as an input to mechanical models of eye behavior), like any good model.

I think this is the point he was making - science doesn't tell us anything with 100% certainty. You find the simplest mechanistic model that fits the data and predicts the results of experiments. If the logic used to arrive at the model is sound, then it gives insight into the mechanism of the workings of the system. A correlation, on the other hand, will simply give us a relationship between a few variables with no mechanistic basis, though perhaps with less error (as I can fit an arbitrary number of data points exactly with enough parameters). /late night rambling

 

[Matthias99]

Originally posted by: CSMR

Also, "E will happen with probability 99%" IS an "objective fact" according to me (and most other scientists);

You and some other scientists may make statements like this but I think you should explain what it means as an objective fact if you are to make a case for it being an objective fact.

I think he might be referring to things such as radioactive decay; QM allows you to establish probability distributions for when atoms will decay, but explicitly states that the process is really random and so you can only predict it stochastically. If QM predicts that element X will decay at rate Y, and this correlates better to reality than any other theory, you might say that is a "fact" (the actual observed decay rate is an "objective fact", but if you want to get technical any prediction is always theoretical).

Of course, it is impossible to "prove" that the distribution does always match reality, or that some other model doesn't fit it better, or that there isn't some way to get around the uncertainty that QM describes (since you can't prove a negative). It could be that QM is wrong in some cases and we just haven't found those cases yet (much like GR, and Newtonian mechanics before that).

Also, you can't "prove" anything definitively in regards to the real world, because any such prediction is based on inference from observed data (and, usually, an implicit assumption that future events will be like past events). If you get technical enough with your definitions, you can't make any objective determinations about how future actions will behave. But that's more philosophy than anything else, since most people assume in practice that the 'laws of nature' aren't going to be pulled out from underneath us all of a sudden.

So, in some sense -- there is no way to truly know that something is unknowable. QM asserts that some things are truly random and unknowable, and there is no obvious way to test this assertion other than to make observations and see if reality correlates with the theory. If everything we can observe correlates with what the theory predicts, then the theory may also be correct for the things we cannot directly observe. On the other hand, you can't ever be absolutely certain that this is true (among other reasons, you have probably not personally observed everything that is possible).

QM says that some information about the state of the universe simply cannot be obtained, and therefore leads to randomness. You can find that certain ways of trying to get at the information are impossible, but you can never rule out that there might be some way around it (since, again, you can't prove a negative). But this is far more philosophy than science. In practice, we work with the best theories we have and try to prove or disprove their correlation with reality.

 

[Blouge]

> QM...explicitly states that the process is really random

Matthias99, if that is stated in your QM textbook, then that textbook should belongs in the fiction section of the library. QM consists of PRECISE, DETERMINISTIC formulas and yields PRECISE statistical distributions that can be verified experimentally. A statement that something is "random" can never be verified experimentally. It's not science. A statement that some quantity satisfies a statistical distribution is experimentally verifiable (or falsifiable, if the theory in question fails). That IS science.

For example, "humans are born and grow to 5.5 plus or minus 1.5 feet" is a verifiable/falsifiable statement. Saying QM is based on randomness is like analagous to the statement "this particular human is born and then grows to 5.5 plus or minus 1.5 * rand() feet". If his parents are giants, and he ends up at 7 feet, did the rand() function take that into account for us? How can you check this experimentally?? When I tried to duplicate the experiment, the rand() function gave me a different result from yours!

Computers have pseudorandom number generators. For example, you can use the C++ rand() function (random number generator) to simulate QM's predicted statistical distributions, that's completely acceptable and will generate results that are almost indistinguishable from real experiments. One could submit a paper with experimental results based on this PC's output and probably no one would ever discover the deception. But note that the PC's rand() function is completely deterministic, it is ONLY PSEUDORANDOM, not random.

Maybe QM operates based on a deterministic, pseudorandom function? It might even be an exact copy of the function that's in your own PC! This is an example of completely deterministic QM. There are MANY others. Occam's razor suggests discarding this because it's more complex than the "random" model. But Occam's razor just helps us simplify our models. It doesn't guarantee that the universe is anything like what we are modeling. Our models have failed time and time again through the centuries and QM is no exception. It has plenty of its own failures including GR incompatibility.

Furthermore, Occam's razor suggests droping the randomness part and simply admitting, WE DON'T KNOW. Extending the model called QM by gluing the part called "randomness" onto it, in order to explain things you don't understand, makes your model more complicated, violates Occam's razor, and is simply a dumb and arrogant thing to do.

>Einstein believed in determinism.

Yes, CycloWizard, I posted that already.

>The bottom line is that you don't either

Correct, I don't know whether the universe is deterministic or random. And I never will know, because it cannot be known through science. If you have ways other than science to gather knowledge, e.g communication with dieties or ESP, please let us know.

>Attacking someone for arguing one way or the other is plain ridiculous.

Agreed. One should attack the message, not the messenger. But I don't think mentioning the other debater automatically constitutes an ad hominem attack. Maybe you are being too sensitive.

>Someone should introduce you to the vacation stick

I don't know what a "vacation stick" is, nor does Google, but note that threatening others with physical harm constitutes a crime in most places.

>f95toli is probably.. least argumentative person in these forums

That's too bad. I was hoping for an exciting and productive debate. Not off-topic whining and defending other posters when you should be defending ideas.

>yet here you are waging personal attacks against him

Not at all. Where did you get that idea? Were you bothered by my "lesser scientist" remark? If you think that remark was directed at him, then you are wrong. Even if it were directed at me, I wouldn't consider it an insult. I don't consider myself to be a greater scientist than Einstein.

 

[blackllotus]

Originally posted by: Blouge
Computers have pseudorandom number generators. For example, you can use the C++ rand() function (random number generator) to simulate QM's predicted statistical distributions, that's completely acceptable and will generate results that are almost indistinguishable from real experiments. One could submit a paper with experimental results based on this PC's output and probably no one would ever discover the deception. But note that the PC's rand() function is completely deterministic, it is ONLY PSEUDORANDOM, not random.

C/C++ rand() is not an acceptable pseudorandom number generator to use in many simulations because it is generally implemented as an LCG. LCG's have serial correlation which is very bad for simulations that need good random numbers.

 

[Matthias99]

Originally posted by: Blouge
> QM...explicitly states that the process is really random

Matthias99, if that is stated in your QM textbook, then that textbook should belongs in the fiction section of the library. QM consists of PRECISE, DETERMINISTIC formulas and yields PRECISE statistical distributions that can be verified experimentally. A statement that something is "random" can never be verified experimentally. It's not science. A statement that some quantity satisfies a statistical distribution is experimentally verifiable (or falsifiable, if the theory in question fails). That IS science.

Yeesh. That's what I get for trying to simplify.

A "statistical distribution", no matter how precisely defined, is a stochastic process (one that most people would call "random"). That is, you can say that there is a certain probability of an event happening or not happening -- and you may even be able to determine that probability very, very precisely -- but you cannot say for a given trial or sample what the result is going to be.

For example, trying to determine exactly how long it will be until an atom of a radioactive element decays. According to QM, this is impossible to determine. Not "hard", not "we don't have the tools to determine it", but "not physically possible". You can calculate statistical distributions that describe how a large sample of such atoms will behave, but you cannot predict the behavior of the individual atoms.

You can "verify" that a set of experiments produced results that appear to follow some sort of statistical pattern. But, if you want to get into the epistemological side of things, you cannot know that your conclusion is truly correct, or will continue to be correct. You can't truly know that anything you predict or conclude about the outside world is correct.

Saying QM is based on randomness is like analagous to the statement "this particular human is born and then grows to 5.5 plus or minus 1.5 * rand() feet". If his parents are giants, and he ends up at 7 feet, did the rand() function take that into account for us? How can you check this experimentally?? When I tried to duplicate the experiment, the rand() function gave me a different result from yours!

I didn't say "QM is based on randomness", I said that some of the conclusions that come from QM (and the results of a number of experiments) imply that there are truly random events and not just ones that 'appear' random due to a lack of knowledge about initial system states. In fairness, I could have probably been clearer.

snip -- long rant about PRNGs

Thanks. My CS degree covered that in far more depth. I know what a pseudorandom number generator is, and how it differs from a true source of randomness.

Furthermore, Occam's razor suggests droping the randomness part and simply admitting, WE DON'T KNOW. Extending the model called QM by gluing the part called "randomness" onto it, in order to explain things you don't understand, makes your model more complicated, violates Occam's razor, and is simply a dumb and arrogant thing to do.

You can't prove that there are truly random processes out there, but this is more of a philosophical point than anything else. QM suggests that at a certain scale, there are nondeterministic phenomenon (although I will admit that this is a gross oversimplification and I am not a physicist).

>Someone should introduce you to the vacation stick

I don't know what a "vacation stick" is, nor does Google, but note that threatening others with physical harm constitutes a crime in most places.

Lighten up. He's implying that you should be temporarily banned for your combative attitude. Play nice, please.

Not at all. Where did you get that idea? Were you bothered by my "lesser scientist" remark? If you think that remark was directed at him, then you are wrong. Even if it were directed at me, I wouldn't consider it an insult. I don't consider myself to be a greater scientist than Einstein.

I guess I can see how a lesser scientist, ashamed of their own ignorance, would resort to adopting a philosophical concept like indeterminism. Belief in The Gods has fallen out of fashion and might not be politically-correct nowadays.

Thrown into a direct response to his post, that's flaming. Stick to talking ideas and not implying that people who disagree with you are ignorant and "lesser".

 

[Blouge]

>According to QM, this is impossible to determine.

Yes, QM says it's impossible to determine in the model called QM. I think the Copenhagen interpretation says it's impossible to determine precisely by using a measuring device. (I'm not sure what kinds of particle collections constitiute a measuring device, and which do not.) Proving the non-existence of a way of determining something also seems non-scientific and impossible at first glance, except perhaps in pure mathematics. Anyway, this is kind of diverging from the topic of randomness.

>not "we don't have the tools to determine it", but "not physically possible".

There are more advanced physics than QM in which more things are possible. I base this opinion on progress made in the past century. I don't think we've finally reached the pinnacle of ultimate knowledge quite yet.

I am not sure why this thread has diverged and the focus has shifted to QM. QM is a definitely flawed theory and therefore cannot fully explain the universe. No theory can - science merely makes models that are better and better approximations. Neither QM nor any successor can establish the existence of ultimate indeterminism, as it is a non-scientific, religious concept.

>I know what a pseudorandom number generator is, and how it differs from a true source of randomness.

Great! You KNOW the difference between a deterministic pseudorandom number generator, and true indeterministic randomness. But can you MEASURE it? Can you experimentally distinguish between them? If I give you a computer printout with a stream of seemingly random numbers, can you conclusively rule out that its pseudorandom? No, you can't. You will struggle, checking countless pseudorandom number generators, and then you'll give up and say either "I don't know how these were generated". Or you'll say "Wow! This must be true randomness!" or "Wow, your pseudorandom number generator must be really complex!". Only the first of the three answers is acceptable; the others involve jumping to a conclusion based no evidence.

>You can't prove that there are truly random processes out there

Thanks, that's what I'm trying to say.

>Thrown into a direct response to his post, that's flaming.
>Stick to talking ideas and not implying that people who disagree with you are ignorant and "lesser".

You couldn't be more wrong. I said that Einstein rejected a certain idea (indeterminism), but I can see how some lesser scientist might adopt it (and I explaned the reason why). Einstein is usually considered the greatest of all scientists, and therefore any other scientist must be lesser than him. If the person I said this to is in fact a scientist, AND thinks I was referring directly to him/her (which I was NOT), AND thinks I called him/her less than Einstein, AND that person is in fact better than Einstein, then perhaps they might take offense. But hardly enough to warrant other posters joining up to start a flame war. And I never said I was a scientist (I'm not actually), so how could "a lesser scientist" possibly imply that the person was lesser than me?? It's quite possible that I'm the "lesser scientist", because I'm not a scientist at all.

 

[Matthias99]

(side note: please use the [ q ] and [ /q ] tags, or use the "quote" function in the forum. It's much easier to follow than you sticking in lines prefaced with '>'.)

Originally posted by: Blouge
Proving the non-existence of a way of determining something also seems non-scientific and impossible at first glance, except perhaps in pure mathematics.

Yeah, it's kind of a problem. One of the reasons, I'm sure, why Quantum Mechanics has been (and continues to be) so controversial.

>not "we don't have the tools to determine it", but "not physically possible".

There are more advanced physics than QM in which more things are possible.

...well, there's a vague statement. Care to elaborate?

QM is a definitely flawed theory and therefore cannot fully explain the universe. No theory can - science merely makes models that are better and better approximations. Neither QM nor any successor can establish the existence of ultimate indeterminism, as it is a non-scientific, religious concept.

This is true. It is impossible to know if something is truly indeterminate, or merely impossible for us to distinguish from indeterminism because of a lack of knowledge.

>I know what a pseudorandom number generator is, and how it differs from a true source of randomness.

Great! You KNOW the difference between a deterministic pseudorandom number generator, and true indeterministic randomness. But can you MEASURE it? Can you experimentally distinguish between them? If I give you a computer printout with a stream of seemingly random numbers, can you conclusively rule out that its pseudorandom? No, you can't. You will struggle, checking countless pseudorandom number generators, and then you'll give up and say either "I don't know how these were generated". Or you'll say "Wow! This must be true randomness!" or "Wow, your pseudorandom number generator must be really complex!". Only the first of the three answers is acceptable; the others involve jumping to a conclusion based no evidence.

There are a number of mathematical tests you can do. Essentially, the stream is not random if you are able to predict its later outputs based on earlier ones. If I am able to predict future outputs significantly better than chance, the numbers you are giving me are not random.

Well, at least they are probably not random. The more numbers you are given, the more confidence you can have in any statistical tests or conclusions about it. It is, however, impossible to be certain -- no matter how much data you are given or how well your conclusions seem to match up. But if an overwhelming volume of data seemed to indicate that the sequence is truly random or not, that would be a reasonable conclusion to draw.

You can't really be certain that a seemingly stochastic process you are observing is 'really' going to keep following the patterns you have seen it follow so far, either. This is a philosophical problem, not a scientific one. At some point you have to decide that you have enough evidence to make a conclusion, or else you will never be able to draw any conclusions about anything. And, of course, you should always keep an open mind about alternative explanations if some kind of evidence to the contrary shows up...

>You can't prove that there are truly random processes out there

Thanks, that's what I'm trying to say.

...and what I'm trying to say is that you can't dismiss theories you don't like on those grounds, because you can't really "prove" anything definitively through observation and prediction. While you can't definitively prove that QM is right (at least with our current knowledge) -- it is possible to prove that QM is wrong, and nobody has been able to do that. Although nobody has reconciled it with GR either, so it's hardly a closed case.

From a practical perspective -- if there are phenomenon that behave in what appears to be a nondeterministic fashion, and the theory that best predicts that behavior concludes that this behavior is nondeterministic, it seems reasonable (lacking evidence to the contrary) to at least consider that there could be truly nondeterministic behavior at work.

You couldn't be more wrong.

Maybe you didn't mean anything by your comment. But it's VERY easy to read what you wrote as insulting/denigrating, especially given how you just went right into it in the middle of directly responding to the previous poster's argument.

Saying "Einstein (who is implied to be brilliant) said X, but someone who was 'ignorant' might have come to a different conclusion" implies that anyone who disagrees with Einstein (and, by extension, you) is ignorant. You didn't use those exact words, but it's not hard to read what you wrote that way. Please be more careful with your choice of words and phrasing if you don't want people to misunderstand you.

 

[CycloWizard]

Originally posted by: Blouge
Not at all. Where did you get that idea? Were you bothered by my "lesser scientist" remark? If you think that remark was directed at him, then you are wrong. Even if it were directed at me, I wouldn't consider it an insult. I don't consider myself to be a greater scientist than Einstein.

Your entire premise in this thread is that Einstein is right and true randomness/QM is incorrect simply because you consider him to be the greatest scientist of all time. This is, I believe, what is known in philosophical circles as fallacy of composition, with the form "Einstein said many great things pertaining to science, so everything he said pertaining to science must also be true." To borrow a quote from the linked site (simply because it's humorous):

This is fallacious for the same reason that the following is fallacious: "You like eggs, icecream, pizza, cake, fish, jello, chicken, taco sauce, soda, oranges, milk, egg rolls, and yogurt so you must like this yummy dish made out of all of them."

One might also construe your arguments as begging the question, but I think the argument for the composition fallacy is better suited to your statements thus far.

 

[PooBeetle]

but cyclowiz, you are using an argument called reductio ad absurdum, which is inconsistent, and proved fallacious by people living in the dark ages, and THEN basing it on an analogy, which as bertrand russel once proved in ONE step is inconsistent, no analogy can ever be remotely akin to logic or math.

also;
many worlds is the only quantum mechanical theory that is consistent, ie pure in it's math.
these theories rest on the outside observer.
ie the watched pot phenomena, goes basically, an atom refuses point blank to opt for a state when you are watching it, no matter how much energy you pump into him, or how hard you hit him, he steadfastly will not be coerced into doing anything.
but as soon as you turn away for an instant, the atom will decide.

the implications are left to the individual.
(but it's all really obvious.)

 

[CycloWizard]

Originally posted by: PooBeetle
but cyclowiz, you are using an argument called reductio ad absurdum, which is inconsistent, and proved fallacious by people living in the dark ages, and THEN basing it on an analogy, which as bertrand russel once proved in ONE step is inconsistent, no analogy can ever be remotely akin to logic or math.

Wrong. I never said that Einstein is wrong. I simply said that his being a great scientist doesn't mean that everything he said about science is correct. I didn't draw any conclusions about whether or not he was wrong. I merely stated that his being a great scientist does not in any way impact the 'correctness' of whether truly random phenomena exist. If you're questioning my summary of Blouge's comments in this thread, then I suggest re-reading his posts, as this argument is the essence of everything that he has said on the subject.

I'd also point you to Wiki, which clearly states:

There is a fairly common misconception that reductio ad absurdum simply denotes "a silly argument" and is itself a logical fallacy. However, this is not correct; a properly constructed reductio constitutes a correct argument.

Of course, this is fairly obvious, since proof by contradiction is a commonly applied and accepted technique in mathematics.