Dude, I understand the impatience/frustration due to the slow progress of physics in past 40 or so years, but if I understand you correctly, you're seriously suggesting to sit on our asses and do nothing just waiting/praying for the next Einstein to turn up ? I'm also on the opinion that the Standard Model is a bit like curve-fitting experimental results, and it does (obviously) work with observation around the range it was designed for (and we already know it fails outside of that, but still definitely useful in practice), but the best thing we can do IMO is to push experiments further and further to try to find discrepancies which will make it easier to come up with new theories, which is exactly what is going on at the LHC and many other places around the world. The next revolution might very well come out of a totally different field ~ personally I'm hoping that exascale computing will make it easier to test new theories (being able to accurate simulate entire organs at the molecular level could revolutionise medicine, advances in quantum chemistry could also [in]validate some theories and have plenty of practical implications).

The anser to most of the questions above would be "No, but I can easily learn it" The ability to learn is the key feature of the brain, not what you learn specifically. Different encironment -> different problems to solve. I would also add that contemporary science problems that we're trying to solve are more difficult than the problems that have been solved in the past.

does anyone really care ?

The real problem is a lack of a common API for encoding regardless of GPU/CPU, which leads to vendor-specific implementations with varying degrees of quality. The most efficient way to pretty much do anything is a dedicated HW block (from both perf and power point of view), so there is no question that there is value in encoding using dedicated hardware, but the software has to catch up.

The key part of the article is "according to vaccine researchers"...

One man's USB treasure is another man's garbage...

I'm no mathematician, but the proof in the paper looks very solid (certainly makes sense from a coding point of view). Though the proof is more general, it also pretty much demonstrates that factorization can't be achieved in polynomial time (thus that RSA is indeed secure).

Many image-enhancement techniques exist that do just this, and this is not really new. In fact this proves that square pixels work just fine to transmit the information, but the image can be enhanced to a larger resolution by non-linear techniques that work better than simple [traditional] upsampling.

nice.