Fake Randomness?

[sanctus]

This is a very simple question, answering to a post in where does intelligence come from thread, I was wondering if a deterministic process could be seen as random if it so complicated that we are far away of a possible analytical solution. Hope you understand what I mean.

[C1ay]

I recall reading a book one time that was simply entitled "Chaos". The author wrote of a scenario that described a resulting hurricane around the globe that had been initiated by the breeze of a butterflies wing on the other side of the world. This suggested a deterministic outcome that we would not be able to predict with our ability in analysis. At the same time he pointed out that even a slightest difference in temperature causing a draft, might have offset the initial breeze and nulled the resulting hurricane. This suggests a randomness to the very survival of that breeze to cause a hurricane in the end. These seems to imply to me that the deterministic outcome still relied on a coincidental set of random, supporting events. One could argue though that those very events were themselves the result of a deterministic system. I believe this could be an example of that type of randomness that you are seeking.

[sanctus]

Yes exactly, I was wondering if its correct to call that random?

[Qfwfq]

I was wondering if its correct to call that random?

Whether or not it's correct depends on what you mean by 'random'.

 

I say it is random. Roll a die and try to calculate which of the six numbers you will get, based on the initial motion. Would you bet your savings on your calculation? Hardly anyone would call the rolling dice fake randomness.

 

What C1ay mentions is usually called the Lorenz butterfly, Lorenz proved that the weather is an example of a dynamic system with a chaotic attractor, often called a strange attractor.

 

Although people often phrase it that way for simplicity, Lorenz doesn't say that the butterfly causes the hurricane, the hurricane is caused by a hell of a lot more than the butterfly. What he shows is that, if you try to predict the weather at a distance of a month or so, outcomes differing by that much may evolve from initial conditions having a difference comparable to the flap of a butterfly's wing.

 

Another example that has been calculated is when the triangle of pool balls is broken up by the first shot. No hope of prediction.

[UncleAl]

It's called "encryption." Computer progams are deterministic. Any software that encrypts text is deterministic - because the text can be recovered by software. One strives for long period repeats and no apparent pattern to output.

 

Consider a B&W GIF of text. One uses a seeded pseudorandom number generator to specify streaks across the screen of varying edge origin, angle, and width. All pixels under the streak reverse black and white. Wrap the screen if you like, add wrap perturbations, etc.

 

You can appreciate how after a very short time the screen is reduced to a random jumble of B&W noise. There is no way to look for letter frequencies or other text regularities because there is no text - only a massaged image. Reverse the process and the text reappears unaltered.

 

The B&W scramble is obviously not random. What possible diagnostic would detect this irrefutable fact vs. a screen of random B&W noise?

[Fishteacher73]

I say it is random. Roll a die and try to calculate which of the six numbers you will get, based on the initial motion. Would you bet your savings on your calculation? Hardly anyone would call the rolling dice fake randomness.

 

The problem is that you are not calculating all the variables (Which may be a lost cause itself, but lets put that aside for just a moment). I you know the mass off the die, the velovity, angle of impact, the compression of the table, etc, etc. the out come is not a debatable occurance. Just as any other physics problem would be. We calculate events whith even higher complexity to a very accurate degree (moon shots for example). There is nothing random about the outcome of a die roll.

[infamous]

The problem is that you are not calculating all the variables (Which may be a lost cause itself, but lets put that aside for just a moment). I you know the mass off the die, the velovity, angle of impact, the compression of the table, etc, etc. the out come is not a debatable occurance. Just as any other physics problem would be. We calculate events whith even higher complexity to a very accurate degree (moon shots for example). There is nothing random about the outcome of a die roll.

 

Very true Fishteacher; What ever select position one prefers to view their earthly existence as, our lives are nothing more than the fall of a domino in the domino row.

[bumab]

Thats a great naturalistic (or materialistic) position, and certainly very compelling, I think.

 

So, how would you tackle our perception of ourselves, the famous "I think, therefore I am" argument?

 

I will grant that something could become self aware. But when does it become beneficial to have the illusion of no longer being subject to causality, as we all do.

[Biochemist]

Yes exactly, I was wondering if its correct to call that random?

I believe the correct answer to this question is "No".

 

The book that C1ay mention in post #2 (Chaos, by James Gleick) is a great treatise on the origins of Chaos Theory. Fundamentally, the fact that we cannot solve for the underlying mathematics of complex systems (like weather, or ecological systems) does not mean that their changes in state are not driven by deterministic factors. Specifcally, if a system is driven by more than two non-linear processes, it is unlikely that we will be able to tease them apart and solve for the source relationships.

 

Ergo, chaotic systems are deterministic but unsolvable. It has been suggested (and I personally believe it is true) that most environments that appear random are actually chaotic. We just can't prove it because we can't solve the mathematics for causality.

[TINNY]

It's called "encryption." Computer progams are deterministic. Any software that encrypts text is deterministic - because the text can be recovered by software. One strives for long period repeats and no apparent pattern to output.

how about quantum encryption?

[Qfwfq]

The problem is that you are not calculating all the variables (Which may be a lost cause itself, but lets put that aside for just a moment). I you know the mass off the die, the velovity, angle of impact, the compression of the table, etc, etc. the out come is not a debatable occurance. Just as any other physics problem would be. We calculate events whith even higher complexity to a very accurate degree (moon shots for example). There is nothing random about the outcome of a die roll.

No Fishteacher, the problem isn't that of not calculating all the variables but that of knowing the boundary conditions with perfect precision.

 

W dynamic system having a chaotic attractor means that, however small the uncertainty on the boundary conditions, the integral of the differential equations will have uncertainties that bring to qualitatively differing behaviour.

 

This is how deterministic dynamics can yet describe random events. If you don't consider that 'chaotic' and 'stochastic' mean 'random' or 'casual' that's a semantic issue that would have to be sorted out.

 

According to Bohm and similar interpretations, even the casuality of QM outcomes would have the same type of basis but this is something we can't determine as far as is now known.

[Qfwfq]

The B&W scramble is obviously not random. What possible diagnostic would detect this irrefutable fact vs. a screen of random B&W noise?

If your meaning is that the CIA would be unable to decrypt the message, encrypted by this method, then you'd better hope they don't come to know you've published the method across the Internet. I heard of a guy that's in jail for this type of reason.

 

Actually, the main point is the ratio of message length over key length. If you're happy with a key that's as long as the message then even the simple x-or method is good enough. I believe it was used for key diplomatic comunications between the White House and the Cremlin up to recent times.

[sanctus]

The conclusion I draw therefore by all of your answers is that the procesuss' aren't random, but they appear to us as being random as we can't solve them mathematically. Therefore, as you said Qfwfq, it depends on how what we mean by randomness.

[Qfwfq]

Looking up the definitions of random, chaos and chance the most relevant meanings are along the lines of "without previous calculation", "disorder", "confusion" and similar things. The only hint by which you might consider fakeness is "without rule" in a definition or two, this would imply specifying what that definition means by 'rule'. The Euler-Lagrange equations for the given system? Or some imperial decree?

 

Not only have we not previously calculated the rolling die, but even if we tried we couldn't predict the outcome. I don't even say that we can't solve the problems mathematically, I say we can't have enough precision, quantitavely, in knowing the boundary conditions. Mathematically means knowing the equations and their integrals. From this pov even chaotic systems are solvable, the solutions are chaotic.

[paultrr]

Looking up the definitions of random, chaos and chance the most relevant meanings are along the lines of "without previous calculation", "disorder", "confusion" and similar things. The only hint by which you might consider fakeness is "without rule" in a definition or two, this would imply specifying what that definition means by 'rule'. The Euler-Lagrange equations for the given system? Or some imperial decree?

 

Not only have we not previously calculated the rolling die, but even if we tried we couldn't predict the outcome. I don't even say that we can't solve the problems mathematically, I say we can't have enough precision, quantitavely, in knowing the boundary conditions. Mathematically means knowing the equations and their integrals. From this pov even chaotic systems are solvable, the solutions are chaotic.

 

Good point and rather fits something some of us have been trying to get across when it comes to pure determinism. If there are unknows(Randomness) then saying everything is pure based upon determnism is not a proper position to take. In this case one can solve out chaotic equations into solutions that are chaotic themselves. But one cannot predict everything with certanity. If anything the chaotic seems to be able out of the radomness to generate order on higher scales would be a proper conclusion.