Tuesday, May 31, 2016

Compression, Randomness, a theory of everything, and a 200TB proof

If the universe is describable (and I think it is if we allow neologisms), will its description fit in 1,000 pages? There is a computer that in two days apparently churned out a 200 TB proof yet my assumption is that the code is probably less than one GB. 200000GB to 1 GB is a mighty high compression ratio. Lossless compression at that. That 200 TB proof contains enough structure to be a proof. One could call the code that produced it a description of the proof.

In fact, 200,000:1 is exceptional levels of structure.

I tried using winRAR to compress various types of things. I attempted to compress white noise, speech, and a perfect 440Hz. I also compressed the output of a random number generator, plainext, and ciphertext. Finally images.

None of them is as high a compression ratio 200000:1 but that pales in comparison to infinite compression:

A sequence of all ones.

That description is a lossless compression of an infinite string to a finite string. So infinite to finite compression is attainable.

The question is will we ever find it?

Imagine the totality of all descriptions. This is included in the so called hyperwebster. Since the universe is describable, the hyperwebster contains that description.

Is that close?

Friday, May 20, 2016

Free Will, Omniscience, whether God plays dice, and a Theory of Everything

I'm having trouble accepting that random phenomena occur in the material universe but even less willing to accept that I don't have free will.  If truly random phenomena occur, that would imply that determinism is false.  Laplace had an idea called Laplace's demon which if feed the demon all the info about all the matter and energy in the material universe it will be able to use physics (or something) to predict what the state of the universe will be.  It seems that if there can be Laplace's demon (which remains to be seen), that would imply determinism, no free-will, and the non-existence of truly random phenomena.  If we assume that either determinism is false, or we do have free will, or truly random phenomena exist, then Laplace's demon does not exist.  

Laplace's demon would seem to be equivalent to an omniscient agent of some kind.  Basically the formulas in the TOE can predict the future and recall the past if fed absolutely every iota of information about the initial state of material such as every quark's momentum and position.  In a sense, the TOE itself is the omniscient agent.  It's just that the omniscience is not revealed until that TOE is programmed into a computer.

So, omniscience exists if and only if Laplace's demon (or something like it) exists.  Also, if we take the stance that we have free will then that strikes down the possibility of Laplace's demon, meaning that the existence of an omniscent being would mean we have no free will.  

And if we do have free will, that would suggest that there are no omniscient beings.

But maybe the universe is closer to a poem than a physics textbook (like a textbook from the future containing a TOE).  Maybe this dance we play with symbols and truth tables with a T and an F, depending on the truth values of the two propositional variable, has no inherent bearing on truth (what is actually true).  Maybe there is no such thing as simply black and white, false and true, maybe there is just relative grammatical correctness.

If the material universe is made of vibrating strings which exist not in the usual 3+1 dimensional space-time, then the set of states of the material universe is like a symphony.  But a symphony has sheet music and whether or not that sheet music has been written down yet, it might still exist.  This sheet music would be a TOE.

So we have a couple of options:
A. Laplace's Demon exists (or can exist)
B. There exists an omniscient agent
C. Determinism
D. Randomness does not exist
E. There is a TOE

OR

1. Laplace's Demon cannot exist (at least in the material universe)
2. Omniscience is impossible (though there may be a "maximally-scient" agent)
3. Non-determinism
3. God does play dice
4. There is no TOE

A scientist at NASA named David Wolpert seems to have proved that Laplace's Demon cannot exist.  He used purely mathematical arguments; there was no reference or appeals made to quantum mechanics.  I forget the details but something about having two omniscient agents try to emulate each other (i.e., copy what the other knows) and that leading to some kind of contradiction.

But argument by contradiction is not tautological in many-valued logics.