Hacker News Reader

Quantum gravity gets the headlines, but only a pretty tiny minority of physicists actually work on stuff like that. Most people are researching problems much closer to hand. Basically the utopia you're looking for is already here.

> But I also wish that physics were less focused on understanding extremes

To defend this approach: Extrema are a good way to get a handle on a problem, which can then be extended.

In programming we almost always have to handle the null case ("what if the graph is unpopulated?") and often an extreme case as well ("What if we end up with 10 million users? How to we quickly respond to XXX").

In physics, like so many other domains, the poles are often the most enlightening region.

However, as you say, work on those extrema rarely translates into more common situations until a breakthrough happens.

>In programming we almost always have to handle the null case ("what if the graph is unpopulated?") and often an extreme case as well ("What if we end up with 10 million users? How to we quickly respond to XXX").

Even the extreme case has to be somewhat realistic to be worth considering. Physics has gone way past that point once they started building accelerators larger than cities to detect subatomic reactions. Not saying it's not worth investigating but it's not at all comparable to any practical domain, at this point it's l'art pour l'art.

It's realistic, they are learning about the particles that make up everything on Earth. But not necessarily pragmatic, in that there aren't likely new uses for that knowledge in the foreseeable future. Still, imagine what we can do once we've achieved a model of how it all really works.

I don't really know how the model enables much if it requires that kind of effort to test the predictions. If the predictions implied something more practical you could test that. I think the time where this kind of physics breakthroughs lead to anything useful is gone.

At the same time I think there's stuff that's much more practical with still a lot of discoveries to be made - like superconductivity.

The same was said about quantum mechanics (of which these accelerators are part of the investigation) in the 1900s, yet without that work there would have been no transistor .. or superconductors.

Umm how could the same be said about quantum mechanics ? Double slit experiment can be setup in a classroom.

If the predictions of your hypothesis require a country sized accelerator to test what are the real world applications ?

General relativity is maybe a better example: It's necessary for GPS but it was initially very difficult to validate experimentally.

There's a lot more to QM than the double slit experiment (and a lot more experimentation before predicting superconductors).

But if you consider basic research worthless, physics is hardly the only "offender".

I don't consider basic research worthless - just saying the likelihood of anything practical comming out of fundamental physics seems very low when even theory can't predict measurable effects outside of TeV particle collisions.

>the likelihood of anything practical comming out of fundamental physics seems very low when even theory can't predict measurable effects outside of TeV particle collisions.

You may be interested in the work being done on the connections between QGP and strange metals. Some progress is being made on extending the laws discovered in simple scenarios to complicated situations.

I can relate to this. It might seem stupid but understanding this made me change the way I tied my shoes, and my laces come undone way less frequently now.

https://news.berkeley.edu/2017/04/11/shoe-string-theory-scie...

(Also thanks to everyone for their responses, a lot of good stuff to check out.)

Mandatory link to Ian's Shoelace Site: https://www.fieggen.com/shoelace/

This site helped me to realize I had been tying one of my shoes wrong my entire life.

Let's be real. There are more physicists than fitting research problems. That's why your average physics PhD graduate is typically employed outside of the field they learned and earns less than an average React programmer.

That said, theoretical physics has the goal of understanding reality from a reductionist point of view. At the scales that goes from the nucleus to the Solar system there are few questions left. We know the Standard Model and GR and they fit the data perfectly. Of course some questions remain for how things interact when there are many of them (e.g. warm superconductivity) but there are few questions about fundamental laws

You can think about it like bootstrapping an open source system (it was in the homepage today). There are still many technical hurdles downstream but we are interested in reducing the binary blob from which all starts to its perfect minimal form. And the only places we still have not figured out well are things at the limit of our instrument capacity. Black holes (GR and relativity, we still cannot figure out that and proving black hole seems experimentally challenging), exotic particles (what are quarks composed of?), dark matter (why far away galaxies seem to rotate so quickly?), dark energy, inflation, that stuff.

I think physics should be smaller, it has too many graduates. But actually these problems should be researched. They are the fundamental questions that remain and there is a reason that the layman considers this stuff to be "real physics" and not origami folding

> That's why your average physics PhD graduate is typically employed outside of the field they learned and earns less than an average React programmer.

Ha, I used to think that a big chunk of physicists by education are employed outside of the field they learned and earn top 10% salary of the industry doing software in financial companies.

Heard more than once that computer science is easier than physics. Physical background helps hugely with software.

How would physics help with software though? My impression from trying physics is that its just that if you are smart enough to do physics you are smart enough that software engineering is not really hard but at most merely complicated.

Physics teaches people how to work with emergent properties of simple systems, something with applications in software wherever the typical industry case analysis + ontology* paradigm is not the right way to work.

(*This used to be case analysis + object oriented ontology, but now there are other language features to map ontologies on to.)

It’s quite likely that the oversupply of physicists is paradoxically shrinking the space of what can be researched.

More physicists mean more competition, more physicists mean more people to convince that a radical idea is worth pursuing, and even more so more people to convince that a radical result is true.

If the field was smaller and those in it had more freedom, we might see more interest in exploring new areas. Right now it’s hard to see many people in the field with enough freedom to do anything that isn’t the prevailing orthodox view.

Really interesting take. Now I wonder myself if the smaller sizes of fields in the 1800s to late 1930s enabled the great discoveries then.

Also the cost of experiments. You can't build a gravitational wave detector in your backyard. You can't do the data processing on your abacus.

Everything is physics. We break apart chemistry as it’s own subject not because there is a clean line separating protean folding etc, but rather for historical reasons. Stick a lead bar in a particle accelerator and it’s Physics stick a hand in and it’s Radiology.

Material science, Chemistry, and Astronomy really cover most of the obvious areas physics could expand into.

There’s a lot of relatively unexplored ground regarding macro phenomena such as fluids, and molecular dynamics, as well as a lot of open ground in alternate approaches to most common theoretical frontiers.

We’ve barely touched alternatives to tokamaks and string theory. Blending classical and quantum molecular dynamics simulations is under explored. Improving the efficiency of simulating large scale dynamical systems is under explored.

These are the areas I know offhand, I’m sure every working physicist has ideas they think are worthwhile that they seemingly don’t have time for/can’t get funding for.

I blame Gödel's platonism. He showed us that our theories will always have limits, which is true. But there are two ways to confront this problem:

- Incrementally improve the existing theory, knowing that some of your goals are impossible and hoping that you don't get stuck on one of these.

- Develop as many inconsistent-but-locally-useful theories as possible along a method for selecting the right one in the right situation.

The latter didn't sit right with him or his contemporaries--more for gut-feel reasons than anything practical--so many of us are stuck in this rut where we just compete for opportunities to participate in the former.

If there's a method for selecting the right one for the situation, it's all one self-consistent theory

I don't see why the two should be connected. Consistency and utility are somewhat orthogonal.

Here's a counterexample:

Einstein chose hyperbolic geometry for arguments about space ships traveling near light speed, but we use spherical geometry for arguments about the shortest path an airplane should take. Those theories disagree about the playfair postulate, so they're inconsistent. Yet we can pretty reliably pick the right one for the job.

Would you recommend using Newton’s laws to model the behaviour of colliding toy trains at a few feet per second, or Relativity?

Newton's laws are a part of relativity, in the way a map of California is a part of a map of the US.

But .. but ... sure, science as a whole has that particular moby dick, but even when it comes to particle physics there there are many theories (models) side-by-side. Some useful for this, some for that. Grand unified theories (and any theory of everything) will only cover the fundamentals. It's not a coincidence that structural engineers don't have to whip out the Einstein Field Equations when they want to check for resonance, and so on.

And very likely we'll always have competing theories/models at the extremes and they might be inconsistent but they might simply turn out to apply in different regimes, etc.

See also https://en.wikipedia.org/wiki/Model-dependent_realism (coined by Hawking and Leonard Mlodinow)

While the goal of theoretical physics is to have a reductionist approach to understanding reality, it looks like most of this is being spent on sub-nuclear particle physics and extreme astronomical situations. There are many real life situations that can be explained and understood better with branches of theoretical physics such as dynamical systems, condensed matter, statistical physics and others. There needs more funding so more people can solve problems that are of direct societal interest which in turn can attract more funding and students.

>>> I suspect that we could have much richer conceptual tools for thinking about the physical world actually around us if only more resources went into looking at it

What I hear in this statement is more towards the domains of research engineering as opposed to theoretical physics. Sand piles are of course the bridge between the two ;)

>Sand piles are of course the bridge between the two ;)

Bridges are an engineering topic, most coastal areas lie on sedimentary rocks, and wormholes, a theoretical physics topic, are also called "Einstein–Rosen bridges."

In other words, while bridges are bridges between sand piles and sand piles, sand piles are bridges between bridges and bridges.

Yes. You took the entropy right out of my mouth.

I was thinking more along the lines of complex system research, self-organizing criticality etc. E.g. https://en.wikipedia.org/wiki/Abelian_sandpile_model

This can be a scientific rather than an engineering discipline, insofar as it's concerned with understanding the behavior of systems, rather than creating a solution which makes a system behave according to our wishes.

We have a limited understanding of "order" or patterns arising out of systems with many aggregating interactions, and often focusing on relatively idealized and isolated systems.

wine legs on applied science: https://www.youtube.com/watch?v=s6w0tSg-msk !

And let me randomly throw in metallic water too: https://www.youtube.com/watch?v=Vdz18ibX7rE

Are these videos working for you guys?

For me Firefox blocks the YT embeds:

https://www.youtube.com/embed/a6ANMKRBjA8?version=3&modestbr...