eln1

The goal is to find e.g. Lagrangian, which consequences are in agreement with nature ... but calculating these consequences is quite tough - AI could generate required simulations, and it is already starting.

The question is where to search for such e.g. Lagrangian? There are already many people having own models developed for decades - AI could help to objectively verify with simulations - building kind of arena for models, to select the ones in the best agreement with nature, identify their best features.

After checking the already available physics models from humans, maybe succeeding would be also AI generated - e.g. combining what was the most successful in the tested models.

Waiting for enabled by default in Firefox (already in nightly 153) and Chrome (hopefully this year).

Sure, most, see e.g. https://en.wikipedia.org/wiki/JPEG_XL#Official_software_supp...

There are lots of materials about ANS, e.g. gathered here: https://encode.su/threads/2078-List-of-Asymmetric-Numeral-Sy...

One can get grants in physics nearly only if being mainstream ...

Consa brings concrete arguments regarding g-factor, I still haven't seen any concrete explanation, only saying that it is fringe science because of criticism of mainstream ... but he quotes mainstream papers.

Mainstream censors non-mainstream, hence for half a century we are staying in place e.g. quantum gravity, there are appearing new problems ...

https://phys.org/news/2022-05-standard-particle-physics-brok...

The paper claims there are problems already with g-factor, could anybody explain why its objections are not valid?

Thanks, I will read, but generally beside e.g. the gravity problem, with increased accuracy there appear inaccuracies all over the standard model, so maybe it is worth revisiting QED? Are you saying the g-factor inaccuracies are not a problem?

https://phys.org/news/2022-05-standard-particle-physics-brok...

Indeed, e.g. of Dirac:

    “I must say that I am very dissatisfied with the situation because this so-called ’good theory’ does involve neglecting infinities which appear in its equations, ignoring them in an arbitrary way. This is just not sensible mathematics. Sensible mathematics involves disregarding a quantity when it is small – not neglecting it just because it is infinitely great and you do not want it!. ”

and Feynman:

    "The shell game that we play is technically called ’renormalization’. But no matter how clever the word, it is still what I would call a dippy process! Having to resort to such hocus-pocus has prevented us from proving that the theory of quantum electrodynamics is mathematically self-consistent. It’s surprising that the theory still hasn’t been proved self-consistent one way or the other by now; I suspect that renormalization is not mathematically legitimate.”

Recently granted patent: https://patents.google.com/patent/US11234023B2/en

Describing rANS used e.g. in JPEG XL: https://en.wikipedia.org/wiki/Asymmetric_numeral_systems

And many others: https://encode.su/threads/2078-List-of-Asymmetric-Numeral-Sy...

or replacing synonyms with single representation

Your data is now written with ANS if using e.g. Apple, Facebook, Google, Linux, soon JPEG XL: https://en.wikipedia.org/wiki/Asymmetric_numeral_systems

This patent covers rANS variant which is used for example in https://en.wikipedia.org/wiki/JPEG_XL - if granted, only Microsoft will be able to make its hardware encoders/decoders.

Lots of materials about ANS: https://encode.su/threads/2078-List-of-Asymmetric-Numeral-Sy...

The Google patent story: https://arstechnica.com/features/2018/06/inventor-says-googl...

Also got this ANS coding: https://en.wikipedia.org/wiki/Asymmetric_numeral_systems

Jon Sneyers is currently working on JPEG XL, which uses ANS: https://www.spiedigitallibrary.org/conference-proceedings-of...

Here are 400+ comments on this topic: https://www.reddit.com/r/programming/comments/6h08z5/google_...

I've meant special relativity.

If you restrict the speed of propagation of interactions (of massless waves), you nearly automatically get STR ... like in sine-Gordon model - speed of massive kinks becomes limited, and kinks are being contracted to zero while approaching this limit.

Lorentz invariance appears naturally in practically all theories with waves - its violation would be a huge surprise.

To see it, understand STR, the perfect model is sine-Gordon: just many coupled pendula - we get particles ("kinks") with rest mass, which are created/annihilated in pairs, the mass grows exactly like in STR and is released while annihilation ... while moving these particles undergo Lorentz contraction (speed is limited by speed of massless waves) and oscillating particles ("breathers") slow down (time dilation) - exactly like in STR.

https://en.wikipedia.org/wiki/Sine-Gordon_equation

"Universe model with a drill" ;) https://www.youtube.com/watch?v=nl5Qq5kUbEE

Animation of kink-antikink annihilation: https://en.wikipedia.org/wiki/Topological_defect#Images

As the main hypothesis in this paper is that the momentum is transferred by pilot waves, it is worth to emphasize the (cited there) most known recent approaches - Couder's experiments which use classical object with wave-particle duality (droplet coupled with waves it creates), getting e.g. interference in double-slit, orbit quantization and many other quantum-like phenomena:

https://en.wikipedia.org/wiki/Hydrodynamic_quantum_analogs

https://www.youtube.com/watch?v=nmC0ygr08tE

Golomb-Rice with base M is prefix code optimal for approximately geometric probability distribution Pr(x) ~ sqrt(2)^(-Mx). Arithmetic coding or FSE/tANS would allow to use the actual probability distribution. The question is how large the gain could be - how far from Shannon is Golomb-Rice for this specific type of data? If this probability distribution varies, maybe it's worth thinking about adaptive rANS, like in Oodle LZNA and BitKnit: https://fgiesen.wordpress.com/2015/12/21/rans-in-practice/ ps. Is M fixed or adapting?

It depends if you have static or adaptive coder.

Static is much cheaper, uses the same probabilities for the entire data block (e.g. 30 kB), probabilities are stored in the header - practically all Huffman and tANS compressors (however, there are considered exceptions: https://en.wikipedia.org/wiki/Adaptive_Huffman_coding ).

Adaptive can start with e.g. uniform probability (no need to store in header) and learns on the way - it is more costly but gives better compression, used with arithmetic coding or rANS. See https://fgiesen.wordpress.com/2015/05/26/models-for-adaptive... https://fgiesen.wordpress.com/2015/12/21/rans-in-practice/

Before Yann, this ANS variant was implemented by Andrew Polar in 2008: http://www.ezcodesample.com/abs/abs_article.html Here is a list of implementations: http://encode.ru/threads/2078-List-of-Asymmetric-Numeral-Sys...

Code: https://aomedia.googlesource.com/aom/+/master

WebP is switching to rANS: https://chromium-review.googlesource.com/#/c/338781/

Here is a superfast implementation of rANS: https://github.com/jkbonfield/rans_static

Here are explanations of various people:

http://www.ezcodesample.com/abs/abs_article.html

http://cbloomrants.blogspot.com/2014/02/02-18-14-understandi...

http://fastcompression.blogspot.fr/2013/12/finite-state-entr... (Yann Collet, he has many posts)

https://fgiesen.wordpress.com/2014/02/02/rans-notes/

interactive simulator: http://demonstrations.wolfram.com/DataCompressionUsingAsymme...

It points to even older paper for ANS: http://arxiv.org/abs/0710.3861

The more recent one is: http://arxiv.org/abs/1311.2540

The first one is about kind of multidimensional analogue of Fibonacci coding ( https://en.wikipedia.org/wiki/Fibonacci_coding )- on a lattice, such that there cannot be two neighboring '1's. It also uses an interesting concept of Maximal Entropy Random Walk.

The difference is size probably come from suboptimalities which were repaired after open-sourcing: http://encode.ru/threads/2221-LZFSE-New-Apple-Data-Compressi... The difference in speed is more surprising, it could come from using a different compiler - there can be really large differences.

It is poorly optimized for 32bits - some benchmarks: http://encode.ru/threads/2221-LZFSE-New-Apple-Data-Compressi...

Arithmetic coding is very expensive - zstd uses new entropy coding with compression ratio like arithmetic coding, but with Huffman-like speed: https://github.com/Cyan4973/zstd https://github.com/Cyan4973/FiniteStateEntropy http://encode.ru/threads/2078-List-of-Asymmetric-Numeral-Sys...

There is an attempt to patent something ANS-related: http://cbloomrants.blogspot.com/2015/05/05-21-15-software-pa...

Generally, a symbol of probability p contains lg(1/p) bits of information: 1 bit if p=1/2, 2 bits if p=1/4 etc. In Huffman you directly assign a concrete bit sequence to every symbol - it is perfect if their probabilities are powers of 1/2, but generally is not true: requires approximations, leading to a suboptimal compression ratio.

Accurate entropy coders like arithmetic/range coding or ANS family can directly work on symbols of general probabilities: containing a non-integer number of bits. It has to finally produce complete bits - their fractional number is handled by the state of the coder - kind of a buffer containing a fractional number of bits. Complete bits are produced as soon as they accumulate.