I have two main problems with the idea of the end of physics: 1. Physics has an obfuscation problem - the terminology is too thick so we can't see meta connect

I have two main problems with the idea of the end of physics: 1. Physics has an obfuscation problem - the terminology is too thick so we can't see meta connections between concepts or form new insights at other levels of abstraction. 2. Tragically, the more advanced physics gets (as in the more money is directed at things like high-energy particle colliders), the fewer real-world applications are found for new discoveries. The problem with (1) is that, the deeper we look into the edge cases, the less we are sure we know what we are doing. For example, I looked up "nucleon self-wave" because I couldn't remember the terminology around self-interaction, and stumbled onto this article: https://www.osti.gov/biblio/6862283-imaginary-part-nucleon-self-energy-relativistic-field-theory https://www.osti.gov/biblio/6862283-imaginary-part-nucleon-s... Not to pick on it too much, but even if it wasn't behind a paywall, it would probably take 10 years of study to even begin to read it. Why is this important? Because when we're thinking about the very large or small, second-order effects usually overwhelm any first-order effects that we're used to. So we don't have the abstractions to properly think about what happens in, say, the nucleus of an atom. The vast majority of weekend warriors brainstorming about physics will never get anywhere, because their initial assumptions have little bearing on reality. In other words, today we're unlikely to stumble onto a new idea along the lines of relativity or quantum mechanics, because we can't see the forest for the trees and simply map the evidence to formulas which describe it. A simple example of this is the recent hubbub about black hole information escape. As a layman, I can totally understand how a particle could teleport across the event horizon. But as a layman, I also look at different edge cases. If we imagine space as flowing into a black hole at faster than the speed of light past the event horizon, then I find it extremely unlikely that matter will ever teleport out. It's probably all packed into the center, squeezed smaller than a neutron (since pressure there is higher than neutron degeneracy pressure) and maybe even traveling backwards through time, while the event horizon is miles away: https://jila.colorado.edu/~ajsh/insidebh/rn.html https://jila.colorado.edu/~ajsh/insidebh/rn.html Can someone prove to me that the new equations account for all that, and more importantly, fit the explanation on a napkin? Probably not. I have to take their "proof" on faith. Not good. The problem with (2) has more to do with attention than failures of execution. We spend $10 billion here on a particle accelerator, $10 billion there on a fusion reactor, but absolutely nothing on grants to hackers. The next Einstein is likely working for a startup somewhere optimizing ad revenue. He or she will simply never have the time to daydream and see the forest for the trees. My prediction for the 21st century is that this problem will become so widespread that capitalism will reach maximum underemployment. Nearly everyone will spend the entirety of their lives paying the fruits of their labor to wealthy individuals and organizations to innovate in areas that are profitable but aren't material to human progress. A specific example of this is an organization like Google (again, not to pick on them) sinking vast sums into proprietary/niche machine learning hardware and software, rather than providing general-purpose parallelized computing to the masses. We should be able to buy 1000+ core chips for around $100 and program them with something open source like GNU Octave or run a Docker swarm utilizing every core so we can run our Python programs 1000+ times faster. Instead we have to adapt our experiments to subsets of computer science like DSP and SIMD. So we focus on neural nets or genetic algorithms and completely miss the abstract portions of machine learning that don't care about implementation details. And why that's important is that stuff with seemingly emergent behavior (like fluid dynamics) can become tractable when we're not compute-limited. But we can't build simple simulations from first principles and scale them, because we're forced to rely on cookie cutter tools for NVIDIA video cards that others have put together because most languages and frameworks aren't up to the challenge of modeling whatever we're trying to do. Put all this together, and the real innovation is either hopelessly out of reach (who has a compute cluster?) or is happening on a shoestring on YouTube instead of at CERN. I don't know whether to laugh or cry about that. Sorry this got long, but I've been discussing physics like crazy during the first vacation the world's had in half a century, courtesy of COVID-19. Happy Thanksgiving everyone.