Pointers? Bayesian interpretation of General Relativity

[fredrik]

I have resumed working on going through the logic of modern physics as part of the quest for a more systematic and philosophically consistent approach that may help solve some of the fundamental issues and unify not only the mathematics but also the logic and philosophy which has been severly lagging for too long.

 

Intuitively general relativity can be given a very natural and plausible abstracted bayesian interpretation, where the geometry identified with probability priors and geodesics are simply the most basic a priori rule for evolution.

 

Anyway, I have not spent alot of time reading every possible treatise on the subject so I wonder if any of you guys on here happens to know of any papers where this has already been elaborated and perhaps some logical implications has been worked out explicitly? Considering the universe of ideas that people are working on all over the place it seems reasonable that there are at least some papers. I suspect though that they might have been published in philosophy, logic journals or possibly math rather than physics journals. Or maybe it's in somebodys private waste basket. Who knows what is important these days

 

I did a quick google and didn't find anything obviously interesting.

 

Any pointers would be much appreciated.

 

/Fredrik

[fredrik]

A note: I managed to find some related stuff under the label "information geometry" which seems to be the common name.

 

http://en.wikipedia.org/wiki/Information_geometry

 

( The plausible part is that one can choose to define a metric such that the distance between two points represents the difference in information. And in the general case the geometry is dynamical and it can be "trained". )

 

So now I found some partly related papers to this wich I'll scan to see what has been done.

 

Still if anyone has some suggestions for particulary brilliant papers to read I'm interested.

 

/Fredrik

[fredrik]

I found by searching that there is a US professor at the University of Albany - Ariel Caticha that has written a number of at first glance very relevant papers to what seems to focus on the right direction of investigation. I'm happy to see that he seems to have an unusually appropriate focus IMO.

 

I'm going to print and get an overview of some of his work and see how far he takes it. They seem well worth their time reading in more detail.

 

/Fredrik

[fredrik]

The idea (which I beleive in), would be that this is the line of reasoning that will systematically lead us to a first proper step towards a consistent, and even highly plausible, philsophically appealling and logically consistent framework for quantum gravity that from first principles easily integrates with the quantum philosophy.

 

Moreoever the model would also simultanesouly present a easily extendable framework for artificial intelligence.

 

I think there are still a few missing steps here, but I think this is the least implausible method we have at hand.

 

/Fredrik

[fredrik]

I did a quick overview and from the papers I can find from him he is in a very early stage of elaborating the ideas, and it's fairly basic(early in the process), though still important. But he is definitely asking questions right in my taste. I definitely acknowledge some of his posed questions. I hope he keeps the focus up. The last paper I found from him on his are dated 2003, and it was on relative entropy and inductive inference.

 

If anyone knows of any other people that work along these lines, I'm interested to hear about it.

 

Many questions are still missing but since it seems that not that many people are working along these lines (??) as compared to other ideas it may not be that surprising?

 

I have not yet had time to read the papers in detail, but it seems he does consider the consistency of the wavefunction(probability amplitue) formalism, however I am not sure he ever considered it's completeness. And if not, exactly what options are excluded in the formalism. This is important, but maybe I just missed it. I was working on verifying this myself but then perhaps he already did it.

 

Other than that I'm not sure he has yet fully considered dynamical sample spaces and dimensionality, but OTOH that might be the next extension, in a yet young approach. I'll take my time to read his papers more properly and see.

 

/Fredrik