Hacker News new | past | comments | ask | show | jobs | submit login
Physics is stuck (salon.com)
99 points by mellosouls 15 days ago | hide | past | favorite | 133 comments



Whenever I see headlines like this, I'm reminded that physics has a communication problem where the public thinks that high energy physics is the only type (or the only worthy type) of physics out there. There's a lot more to the field than the next theory of everything, and those other fields aren't stuck

For instance, in just this decade human kind turned on the first x-ray free electron laser and increased the state of the art in the brightness of x-ray sources by 10 ORDERS OF MAGNITUDE!!! I'm not sure I can fully explain how transformational that level of improvement was because I work on the accelerator side, not the user side. But, at least understand that there are entire subfields that now exist which didn't before. Not just physics, but in biology, chemistry, materials science, and numerous other fields. It's an instrument so powerful that every other scientifically advanced nation is now building their own. That instrument was developed by physicists and not high energy physicists.

I feel that this also gets at a sort of toxic notion that I've noticed some high energy physicists in my own department have (not all, but you probably know the people that I'm talking about). The notion is that if you aren't working on 12 dimensional quantum theories of gravity or the holographic principle and black hole physics, then you aren't a real physicist. Somehow they think that if your work has a real world impact, then it's tainted in some way. I'm not sure how it's gotten this way, but I've actually been told as much by another grad student that my field should be moved to the engineering department because we aren't doing real physics.


> I feel that this also gets at a sort of toxic notion that I've noticed some high energy physicists in my own department have (not all, but you probably know the people that I'm talking about). The notion is that if you aren't working on 12 dimensional quantum theories of gravity or the holographic principle and black hole physics, then you aren't a real physicist. Somehow they think that if your work has a real world impact, then it's tainted in some way. I'm not sure how it's gotten this way, but I've actually been told as much by another grad student that my field should be moved to the engineering department because we aren't doing real physics.

I got told that by a professor when I was looking at PhD options, I ended up in the maths department physics group because it was much nicer.

It's an odd kind of self selection. People who aren't there to feel smart don't sign up for the course so high energy theoretical physics gets ever more filled with insufferable assholes.

Not to say that the whole group was, but while every other group had the one token dickhead everyone avoided it seemed that the high energy theoretical group was run by them and the non-dickheads in the group were barely tolerated and mainly there to boost citations.


I like to ask the following question: What do we know about reality that we didn't know before the work of these HEP theorists?

This is not rhetorical. There are some works for which there is a concrete positive answer (which might not be simple). But there is much more for which the answer is simply nothing.

I left the field years ago because it was clear that most of the work being done is not physics.


What work were they doing?


My understanding as an x-ray crystallographer and electron microscopist is that XFEL has been nothing short of a massive disappointment. Yes, it works in theory, but in practice it's entirely inefficient and doesn't address the current bottlenecks of structural biology at all.

I'm not saying this to downplay the project (I know very little about it) but the reception among those in the field it's supposedly applicable for has been underwhelming.


I'm not really sure about crystallography (not my field), but I think it's maybe an issue with expectation management. Everyone is trying to sell the new technology as fix for every field. In reality XFEL experiments do have a lot of overhead associated with them and stuff like synchrotron light sources are probably better suited for "everyday" type measurements.

The real breakthroughs that I was talking about are for time resolved and "diffract before destroy" experiments for delicate samples. Those ones can't really be done at a synchrotron with long bunches.

I'm especially excited for femtosecond time resolved measurements of chemical reactions. It's going to be so cool to watch chemical reactions occur at an atomic scale.

I guess the other cool application is that there is so much light that you can perform diffraction off of single molecules when you want your sample in solution/gas phase or can't form crystals. Correct me if I'm wrong, but my impression was that the intensity of XFELs also enabled that.


I didn't know about the time resolved stuff, and thats the type of thing that big expensive projects seem much more economical for. Presumably we will learn foundational (previously theoretical) information which will have wide, lasting effects. The project is so staggeringly complicated that when I've talked to people involved, they knew very little outside of their own minor field (for instance, microfluidics to time droplets with pulses).

The way that the project is being sold from the structural side is that you won't need a crystal, which is obviously huge. However, there's the issue of scale, and also the classic issue of... We can't refocus diffracted X rays so we're still limited by protein we can express in a selenium doped media, since you need the selenium signal to determine phase data from the diffractions. The list of proteins that are soluble but won't crystallize/can't be resolved by Cryo-EM is pretty small, and so it's an incremental upgrade no one will have access to, at best. At least, from the structural side.


I get the impression that cryo electron microscopy is now the big new(ish) technique for structure determination.


Yes, advances in the detectors (essentially CMOS electron detectors) have absolutely revolutionized the field and solved many of the major problems. However, there are still limitations, especially with smaller molecules.


> The notion is that if you aren't working on 12 dimensional quantum theories of gravity or the holographic principle and black hole physics, then you aren't a real physicist. Somehow they think that if your work has a real world impact, then it's tainted in some way. I'm not sure how it's gotten this way, but I've actually been told as much by another grad student that my field should be moved to the engineering department because we aren't doing real physics

Wow, so The Big Bang Theory is actually pretty good documentary about how modern physic departments operate?


Been out of physics for 23 years, but, yes. Without the studio laugh track, and generally humor. Also add in lots of politics.


So... why is the XFEL not “just” an achievement in engineering? Is there a (new) physics somewhere in it? (Genuinely curious.)


I would call it new physics. Getting a relativistic electron bunch to coherently radiate is pretty novel.


I forget who said it, but there is a quote in physics along the lines of "we already have a theory of practically everything, but it is so complicated that we can use it to predict almost nothing."

All of basic principles that underlie our everyday experiences don't need quarks and gluons. They're governed by standard quantum mechanics which was fleshed out in the 20's. However, just because we have Newton's laws of the subatomic world doesn't mean we can answer interesting scientific questions with them. I'd even say that people that call something like the standard model a "theory of everything" are a little naive since you can never use to it make predictions outside of the most simple systems.

That area (pushing ahead away from model systems) is the complexity frontier of physics and is where I'm most excited for advancements. It's where we can write out the equation that describes high-TC superconductors, but it has so many dimensions that we can't solve it and make one that works at room temperature. Is that engineering or physics? I'd call the scientific problem of figuring out the laws that emerge from quantum mechanics for these complicated systems physics.

In a similar vein, Maxwell "discovered" the free electron laser in the 1800s. He wrote down the laws of E&M and then everything in the world was solved. That's clearly a dumb way to think of it though, and the new physics that was discovered were the emergent laws that fall out of it when you apply those laws to a system of relativistic particles.

In particular, the emergent phenomena that had to be discovered for the free electron laser (FEL) was the micro-bunching instability. It's a subtle interaction between relativistic particles and a field of photons that will feedback and cause them to bunch together while dumping their energy into the field of light producing the coherent laser radiation. It was that phenomena and the realization that it could be used to create a powerful laser that are the new physics here.

On the other side, the advancement of the FEL to the point that it could make x-rays requires a lot of other new physics to create the machine. This is where there is a bit of a blurred line between physics and engineering. However, I'd still call what we're doing (developing new laws of physics that govern complex systems) physics. Just because we use them to build a machine doesn't make them engineering any more than a solid state physicist discovering a new phenomena in semiconductors and then Intel taking advantage of it in their next generation of processors.

In order to REALLY understand our physical world, we can't write down reductionist laws and say that we're done. It's also notable that the two developers of the free electron laser are rumored to be in the running for the Nobel prize in physics, so it's not just me who thinks that they are physicists, not engineers.

Edit: BTW, here is a great paper that reviews the physics that makes a free electron laser work if anyone is interested.

Huang, Z., & Kim, K.-J. (2007). Review of x-ray free-electron laser theory. Physical Review Special Topics - Accelerators and Beams, 10(3), 034801. https://doi.org/10.1103/PhysRevSTAB.10.034801


> I forget who said it

Paul Dirac. The quote is:

"The underlying physical laws necessary for the mathematical theory of a large part of physics and the whole of chemistry are thus completely known, and the difficulty is only that the exact application of these laws leads to equations much too complicated to be soluble. It therefore becomes desirable that approximate practical methods of applying quantum mechanics should be developed, which can lead to an explanation of the main features of complex atomic systems without too much computation."


Engineering is more about process and process improvements. If an engineer is breaking new ground they are really acting in a capacity closer to being a physicist.

Consider the famous case of Student's T Distribution - it was discovered via engineering, but it is clearly work that was done in the domain of mathematics rather than engineering.

That which is new is science. That which is re-learning an old lesson is engineering.


Science is the process of producing new knowledge and new questions from old questions, under consumption of coffee.


That seems rather dismissive of engineering practices and ignorant of scientific development.

If we want to get into the philosophical difference between engineering and physics, engineering at its core is solving problems and science is answering questions. There is obvious overlap and no clear distinction.


What is dismissive about it? There are a great many brilliant engineers who never do anything particularly novel.

In contrast, it is quite hard to be a brilliant scientist without doing something novel.


> I'm not sure how it's gotten this way, but I've actually been told as much by another grad student that my field should be moved to the engineering department because we aren't doing real physics.

That's a touch unkind of your friend ... but I think it hits the heart of the problem.

"High-energy physics" is pretty much "uniquely" physics and so is easy to describe to laymen as "physics".

Conversely, the line between something like "solid state physics" and "solid state engineering" can be really blurry (the solid state EE's and solid state physics folks at my alma mater took almost exactly the same classes at the graduate level). So it's a lot harder to describe the differences to laymen even if you point out that solid-state physics is what gave us semiconductors and computer chips.


Heh. Tell them that their field should be moved to the theology department.


I feel like modern high-energy physics as exemplified by CERN is a bit like NASA's Space Shuttle program: its main job is to hoover up all available money to fund the bureaucracy. Science outcomes aren't even a secondary goal, or a tertiary goal. They're waaaaaay down the list.

Similarly, ITER is going to suck up a hundred billion in science funding that could have gone into a hundred $1B programs intsead of one giant one.


Honestly, ITER is so bad that you're insulting the Shuttle with this comparison. Shuttle was super inefficient, but it actually launched stuff. Military, commercial, and scientific satellites. And only like a decade after the program was started, too. ITER started in 1985, won't achieve first plasma until 2025 at earliest, and they won't even start D-T fusion experiments until 2035. That's a full half a century until the START of real fusion experiments, and it could still slip. It won't even produce any electrical power. Any thermal energy produced will be vented.

Shuttle is a shining example of efficiency, competence, pragmatism, and execution in comparison.


The shuttle never had funding halted multiple times.

Also, ITER is a plasma physics experiment. It is being made to learn about plasma behavior. It isn’t going to fail or somehow not produce useful results before DT or high Q campaigns. A fusion reactor does not fall out of the sky. Was the V2 a failure because all it did was eat up a bunch of money and blow up a bunch of expensive equipment? No. The knowledge costs money. Tangible things are laughably low value by comparison.


Shuttle was an experiment in reusable launch technology. It was also a reentry science platform. It was a materials science tour de force, pioneering carbon fiber composites, carbon-carbon composites, metal-matrix composites, and high performance aluminum-lithium alloys among other things.

ITER is supposed to get break-even. Plasma experiments can better be done at smaller, much cheaper facilities (many of which already exist).

ITER, as of 2006, was supposed to be followed almost immediately by DEMO, a power plant capable of producing electricity in 2033. But now ITER is so delayed that it's almost irrelevant. In order to address climate change, we'll have to have transitioned to, say, solar and storage before the 2050s when DEMO might be built, or we'll be stuck with terrible climate consequences.

It's just insane how bad it has been. ITER is so badly delayed that it's almost certainly not going to be a key player in fighting climate change.


Fusion power is not a replacement for solar. It is the technology that needs to be working in the next two hundred years if we want to see society continue on its current trend. It may be the difference between human society lasting billions of years or being a small footnote on a single planet in a dark universe. We’re already stuck with bad climate consequences. We might not destroy the entire habitat or run out of usable fuel if we stop penny pinching and actually pay what’s needed to solve a problem far harder than something like LEO launch vehicles.

As for breakeven: no plasma physicist I’ve talked to see it as some amazing barrier. It is what it’s always been: a publicity thing. It gets people excited so laypeople can point and say “See! It works!” without having to dive deep into physics and economics. The real physics goals of ITER are in testing models when pushed outside of previously measured parameters. You can’t just spend $50 billion on a reactor and hope it works. That’s why increasingly large and expensive science machines are necessary. Once a design is complete you don’t need diagnostics and scientists. You just need a tritium processing facility, some HTS coils, some vessels, and some plasma heaters.


> Fusion power is not a replacement for solar.

You're conflating ITER with fusion power in general. There're at least half a dozen competing approaches to fusion power with several orders of magnitude lower funding. Several of them are probably more promising than ITER, but ITER is primarily a political project, and only secondarily (at best) a physics project.


I am not conflating them. “Several of them are probably more promising than ITER”. Keyword: probably. There are no magic bullets. Would I be happier to see an optimized stellarator with a 10 meter major radius? Yes. A lot of the physics goals of ITER will apply to all MCF reactor designs. If you’re touting the benefits of non-MCF fusion: vaporware. I think many designs should be funded and explored, but it’s not like people haven’t tried clever ways to put particles together before.


None of them are particularly promising. They may be better than ITER, but that's damning with faint praise.

Fusion is a technology that's being pursued because it was being pursued, not because it's particularly attractive from a practical point of view.


> ITER, as of 2006, was supposed to be followed almost immediately by DEMO, a power plant capable of producing electricity in 2033.

This is not true. DEMO will not be able to breed Tritium.

Without tritium breeding, there will be no self-suffiency in term of fuel. You can't filter Tritium from sea water because it is not there - it is unstable. There is no fusion experiment which could achieve a Deuterium-Deuterium fusion within the next decades, the required densities are many orders of magnitudes away.

You can generate Deuterium by electrolysis and isotope separation of water, but to day, Tritium can only be produced in heavy-water generators. Which exist only a few in the world, and which are powered by uranium.

In theory, yes, it could be possible to breed tritium within a fusion reactor. But this is as theoretical as somebody in the 17th century saying that you could adapt a steam engine to flap wings and make it just a bit lighter to get a flying machine which could carry a few hundred people with near speed of sound across the Atlantic, at once.


> This is not true. DEMO will not be able to breed Tritium.

DEMO not only will breed tritium, it will not be feasible without tritium breeding. Otherwise, it would not be able to operate, as it will run out of DT fuel. Even getting enough tritium to start DEMO operations will be difficult, as there will be little left after ITER is done. The primary source of tritium, heavy water power reactors, will not long survive, and even all of those would not provide enough tritium for DEMO operation without DEMO being able to make its own tritium.

The need for breeding to work at DEMO means there's going to have to be an intermediate reactor, a Fusion Nuclear Science Facility, between ITER and DEMO to firm up blanket engineering.

I don't think this gets talked about much because, frankly, those in power in the fusion community will be retired by then.


Of course a DEMO will have tritium breeding. One of ITER’s tasks is testing lithium breeder blankets. Reactor studies have included the calculations of necessary thickness of lithium blankets to make sufficient tritium and not increase the minor radius by an unacceptably high amount.

>ITER is so badly delayed that it's almost certainly not going to be a key player in fighting climate change.

But is that the fault of the scientists and engineers working on it? Or of the politicians writing the checks (and their apathy and/or conflicts of interest)?


I’m not close enough to assign blame. It’s probably the management’s fault if I had to guess. And probably an overall lack of leadership and drive at the highest levels.


> The knowledge costs money. Tangible things are laughably low value by comparison.

I imagine Peter Venkman in Ghostbusters saying this to the Columbia dean as their lab is being disassembled and they are being fired.


Consider the difference in value between a fish and the knowledge of fishing. Now consider the value of a set of fishing gear to the invention of fishing.


> Honestly, ITER is so bad that you're insulting the Shuttle with this comparison. > [ ... ] > It won't even produce any electrical power.

Worse. It won't even address the question what could be the so far unknown near magic materials which a breeding blanket would need to be made of. to support sustainable Tritium re-generation, which would be needed a long time before any demonstration could be hoped to produce any electrical power.

In short, D-T fusion - the only process which can be imagined to be accessible by magnetic inclusion - needs a supply of Tritium and this needs a working Tritium breeding system, with a neutron loss factor below 1. It is completely unknown how to achieve that.


I know nothing about physics but I was under the impression ITER was our best shot for clean sustainable energy.

I'm not saying you're wrong btw, just that I had high hopes.


Neither ITER, nor fusion in general, are our best shots for clean sustainable energy. They are at best very long shots, and at worst complete wastes of effort.


What's your option?

If you say solar or wind, I'll tell you not without solid state physics experiments to multiply battery capacity by at least 10x. (Not impossible, but as much pie in the sky as fusion.) Or a global electric network. Or space solar with concomitant energy sending infrastructure.

Nuclear is good for maybe 300 years. (And not exclusive with other sources.) Hydro, geothermal are nice where we can have them.

Anything else, we don't even know where to start.


Solar and wind. Your comment about batteries is bizarre. Batteries don't need higher capacity for stationary storage, they need to be cheap. That is not "pie in the sky" and is certainly a much easier engineering problem than fusion. And batteries are not what you'd use for long term storage; that would be hydrogen (which can be stored underground for less than the equivalent of $1/kWh of storage capacity.)

Given reasonable projections of where renewables, batteries, and electrolysers will be in 10 years, one will be able to synthesize artificial baseload sources at a lower cost than nuclear fission, anywhere in the world (in some places much less). And fusion will have a very hard time being competitive even with fission.


Yes. Also for grid- level storage space and weight are not at a premium - so cheap, comparatively in efficient solutions like sodium-ion batteries or redox flow systems is where the field is moving to.


These are really expensive on a scale needed for powering current world with solar and there's nothing to suggest we'll stop there.

I'm not sure you understand the scale involved. We're talking billions tons sodium. Millions tons aluminum and steel.

Hence the 10x.


Let's see... producing 1 kg of sodium metal uses about 10 kWh of electrical energy. 1 TWh would produce 100,000 tonnes of sodium. A billion tonnes of sodium would involve 10,000 TWh. World primary energy consumption is below 20 TW, so you're proposing batteries capable of storing hundreds of hours of world energy demand. This is a gross overestimate -- batteries would be used diurnal leveling, with perhaps 12 hours of storage needed, and that just for the grid. Industrial power demand in a renewable world would likely be more dispatchable, and any demand for thermal energy would use bulk thermal storage, not chemical batteries.

If you meant sodium compounds, not sodium metal, note that annual world production of sodium chloride is about 200 million tonnes.

Oh, and about "millions of tons of aluminum and steel": 2019 world steel production was 1.89 BILLION tonnes. Aluminum production in 2018, 60 million tonnes. The world economy is very large. An energy infrastructure capable of powering that economy will also be large, be it renewable, nuclear, or unicorn power.


Yes. It's not impossible. Just really damn hard to pull off.

Sodium chloride is useful for many things beside batteries, if we were to use that as the source. Say, sulphur as well, other traces too.

10x batteries make this a no brainer and super easy to implement, and economical. Heck, 3x batteries even.

You need to make and place enough panels too, doable but needs a lot of political will. And will of course take some time as well.

I think we shouldn't put the cart before the horse though. Fix the immediate problem with a few smaller nuclear plants and work real hard on our own pace to replace them with renewables. If fusion comes in the meantime, great. In the meantime, we can use the spare resources to remove other sources of GHG vent. (Say optimize transportation.)

See the other branch on how I view fusion. I see DD or enhanced DT fusion as endgame and mostly a space technology. (Fusion engines would be rather nice. Think 50x better ion engines, the technology is related.) Few hundred years of work.


> Yes. It's not impossible. Just really damn hard to pull off.

Yes, any global energy system will be extremely expensive. Trillions of dollars worth of expensive. That it is a lot of work is not a showstopper.

I don't see nuclear being part of that solution, though. Why waste money on something that is demonstrably inferior? It had its chance. We don't need to invest in losers.

Fusion is not an endgame. Fusion is bullshit. It's Rube Goldberg on stilts. From an engineering point of view, it's pretty much the opposite of attractive. Keep It Simple, Stupid is not just a slogan.

DT fusion makes no sense in space. The energy comes out as neutrons, which get turned to heat. You can make heat with fission in space much more easily, and with much less reactor mass, than you can with fusion. The notion that fusion is somehow good for space comes from science fiction more than anything. It's a trope, not a good idea.


> Fusion is bullshit.

Especially because the Tritium fuel problem is not solved, and nobody has the faintest idea how to solve it. The materials needed to breed Tritium in something different from the Sun (or thermonuclear weapons) just do not exist for now.


At this point fusion is an elaborate high energy physics stimulus package.

Max we are looking at 1000 TWh, that's two days of global energy demand. Rule of thumb we need 300gm of sodium per kWh. That's 3 billion metric tons. Which is a lot - but for the entire global energy demand it's peanuts

> Batteries don't need higher capacity for stationary storage, they need to be cheap.

Not only cheap, but also sustainable.

Coal is plenty cheap.


Why do you need increased storage density for grid level storage?


Because we don't have enough batteries. Lithium as a minable element is too limited to have enough without increased capacity. (If lithium fusion were ever available, we'd run out in 2000 years or so. Batteries take more.)

Even redox batteries are, as you need rather pure salts. Even pump storage, cheapest option, would require tons of concrete, which means we'd have serious lime shortage. (Once we run out of natural mountains to use.) Plastics cannot be used as they're either not strong enough or require too much fossil fuels.

Any of these problems can be attacked, and they're all about as easy and solvable. Prediction would give about 40 years until that battery capacity is reached. Fusion could be there just as soon.

This is not to say we should stop developing renewables, but we need 100% clean power yesterday. Or rather 20 years ago. Currently the only open path is nuclear with enough other renewables (expected 25%-33% is doable depending on location).

We're talking gigawatthours of battery capacity installed and running per city. Even if you have everyone an electric car with best available batteries, that would maybe come close. (And the network would have to be heavily as adapted. I've accounted for battery wear and manufacturing with recycling of said batteries.)

Top end sodium-sulphur batteries would work too, in equivalent amount. (They do like 250 kWh per bank. Car lithium-ion do ~120 kWh, but you probably want to drive them in the morning.) Mind you the renewables to power them won't materialize overnight either. You get to pay energy and material to make them and modify the grid to adapt to them.

Since we're late as all hell, the only way right now is to nuclear. We can decommission it in 50 years easily enough as needed.

We had the technology to go full renewable since 80s...


We don't have enough lithium for DT fusion then, either. If we assume fusion reactors have a wall loading of 1 MW/m^2, and assume 50 cm of LiPb in the blanket, then fusion reactors capable of supply (as heat) 10 TW would need the 6Li from 80 million tonnes of lithium. Current Li reserves are a fraction of this.

That ARC reactor design, if extended to world energy demand, would require 100x known beryllium resources.


Yes. Fusion is not short game and should not be treated as one. Not funded like it either.

It will be very useful when we get to space, for example, unless we figure out something better. Much fewer problems with it than with nuclear up there. (Especially with getting fuel.)


Why would it be useful when we get to space? When we get to space, we'll be 1 AU from the Sun. At that distance, solar is vastly superior to fusion. And DT fusion will always be inferior to fission in space.

It pisses me off that despite all the risks they are willing to cut corners in using underpaid foreign workers to pour the concrete...


Yes, I have a friend who left plasma physics and is still salty about the situation with ITER. She also said that by the time they're done, the technology is going to be outdated since it all began in the 80's.

I'd disagree that they are there to pump money into administrators. Some of these problems are just really hard and requires monstrously large organizations to make headway. At worst, I'd say that they are misguided/focusing on the wrong scientific questions or in the case of ITER stuck on a path that should have been abandoned a while ago.


If you have the chance, ask her what a better vision of the use of funding would be. Keep in mind that the physics goals are to test high neutron flux divertors, tokamak scaling laws, and burning plasma regime stability.


AFAIK you can’t operate a tokamak continuously, only pulsed.

A stellarator like Wendelstein 7-X looks like a much more promising design.


I know. I work at the first MHD optimized stellarator :)

That doesn’t mean backing out of ITER 15 years ago would have been a good idea. It would be an even worse idea to back out of it now.


Stellarators also have ruinously low volumetric power density, so they're not any less economically dead than tokamaks.


So ITER is an iterator design :)

It's almost as if they are saying that their 'real physics' isn't science, because in real science observation is first and foremost.

I'm really looking forward to what XFEL can do for material science. Perhaps we can finally replace plastics with something like eutectic systems, bulk metallic glass, or the classic transparent aluminum.


Actually in real science theory is first, observation comes later as a way to judge competing theories.

This is because all observations are theory-laden. They don’t tell you anything without a framework to interpret them in.


Isn't this an idealization myth of the scientific method? I'm pretty sure some kind of observation prior to forming the theory will account for many theories.


True that. It is like cat and mouse. You present a theory and validation and usefulness will come from experimental verification. And some theory comes from unexplainable observations. Michelson–Morley experiment and special theory of relativity will be taking the second path.


In a lot of situations, the observation is that the current favorite theory doesn't work for some observation, but the new theory explaining it only comes much later.


I disagree because we don't entertain theories which make no testable predictions. It seems logical that this rule makes observation the ultimate qualifier.


Yeah, as long as there's no experimental confirmation, it's metaphysics, which is a sub-branch of philosophy, not science. (And it doesn't matter how refined their mathematics are...)


A buddy of mine in grad school (he was an HEP postdoc), once told me (computational condensed matter), that I was working in the dirt left over from the "real" physics. He was a nice guy, but enjoyed ribbing me a bit.

That said, I do recollect mostly condescending attitudes from the HEP folks.

After finishing my Ph.D., I moved on to a successful career with building, deploying, supporting supercomputers, and the codes that run on them. My friend meanwhile, wound up working in condensed matter physics, modeling semiconductors (which is what I wrote my thesis on) for a large chip company.

It seems reality sometimes has a sense of humor.


There's a lot of interesting research (and progress) even going on with Classical Mechanics. E.g. in the 90s it has been shown that even classical Newtonian Physics can yield divergent results which even has a Philosophical impact. Also normal (1-2 particle) Quantum Theory theory is progressing, much more is known about the border between QT and Classical Physics. But of course most resources at Physics departments are dedicated to Solid Matter Physics and High Energy Physics. At least the theoretical flavours of both have a lot of overlap actually and Solid Matter Physics also has applications in the industry. But teaching really focusses on going to one of the 2 topics, at least where I studied.

> physics has a communication problem

It puzzles me how in 2020 it's still common to believe that the world is a fully deterministic place. That kind of proves that the results stay inside Universities despite a lot of popular science publications.


I've heard about the Newtonian physics thing, but I forgot the specific reference. Could you provide a link to an article?


They may be referring to Norton's Dome, a thought experiment involving a stationary ball ontop of a cone in which the ball can spontaneously begin moving at any time and still allegedly be obeying Newtonian dynamics: https://en.m.wikipedia.org/wiki/Norton%27s_dome



The advances in physics are minute compared to fields such as biology. The most minute and useless part is particle physics where the findings are completely unuseable in normal life.

Solid state phishing sees some advances, but they are not earth shattering, compared to biology. The day where there will be a substance you put on a cavity of a tooth and it cleans the wound and settles will be something worth of news (that's just an example).

At some point, with limited funding, one must make a choice on "fundamental studies" and study what has a chance of having a practical use.

Just in case : I have a PhD in particle physics done at CERN and I regret not having taken a field which makes more sense than such impractical studies.


Physics doesn't need another Einstein. Einstein explained Brownian motion, the Photoelectric effect, created special relativity and general relativity, the cosmological constant, helped found quantum mechanics, served as an invaluable critic of quantum mechanics. Then he foundered. Einstein never accepted particle physics, refused to follow new developments and became a dinosaur. This sadly is what most "physicists" are doing today, the followed in Einstein's footsteps, mostly the dinosaur part.

What does physics need? The world doesn't know. Nobody knows but someone will do it, someday in a manner no one else thought possible or could really anticipate. In fact, that's not entirely true but new developments will happen and only a handful of people will be in the loop. Lorenz, Poincare etc. e.g. laid some vital groundwork for relativity.

My own two cents on the matter is that we really don't understand our theories well enough and are badly in need of a firmer foundation. The situation is analogous to calculus before Weierstrass, Cauchy, Dedekind and Cantor.

Of course, mathematics wasn't completely stuck just because calculus wasn't fully developed. Probability and non-Euclidean geometry were stunning developments which predated a truer understanding of real numbers.

So it is with physics right now. Unification, strings, etc. isn't working out so well right now. Quantum computing is now a thing and Quantum mechanics is enjoying a second revolution not unlike the General Relativity Renaissance led by Penrose, et. al.

We can't predict the future. We don't know the sequence things we need to take the next step in AI or even if there is one. Will some form of deep learning be all we need? Probably not but possibly yes.

Physics is right where it should be. Frustration is part of the process. We're feeling some pain because our approach isn't working. Instead of having answers to everything maybe we should focus on better questions.


I'm not a physicist or even a researcher of anything but I sometimes watch youtube videos of people experimenting with material sciences and chemistry and a trend I have seen is that even things that we consider to be very basic have a huge lack of good quality information available.

One youtuber I saw spends time looking at new research papers for ideas and has found that they are just impossible to reproduce. Following the steps exactly as well as every possible variation doesn't produce anything close to the results described in the paper. In another video [0] he attempts to make glass and finds that almost all of the information available on the internet about making glass is not enough to actually make glass since they contain only enough information to register a patent but not enough to make from scratch.

I wonder if we will see huge gains by just making complete and unobfiscated information available to the public since it looks like even the foundations of society are secret company internal information.

[0] https://www.youtube.com/watch?v=mUcUy7SqdS0


The world should have a "crafting" wiki, which explains how to build anything from scratch.

From how to make glass from sand to how to build a quantum computer.


Would the crafting wiki contain instructions for crafting the wiki itself?

(An attempt at humour referencing the "set of all sets" paradox. That is all, as you were.)


The Whole Earth Catalog ceased publication in 1971. The Whole Mars Catalog however............

> Physics doesn't need another Einstein. Einstein explained Brownian motion, the Photoelectric effect, created special relativity and general relativity, the cosmological constant, helped found quantum mechanics, served as an invaluable critic of quantum mechanics. Then he foundered. Einstein never accepted particle physics, refused to follow new developments and became a dinosaur.

If you measure a scientist not by what they accomplished, but by what they got wrong, you'll quickly find that there are no good scientists.


> What does physics need? The world doesn't know. Nobody knows but someone will do it, someday in a manner no one else thought possible or could really anticipate.

I think that is precisely what the professor meant when he talked about Einstein, not the person himself with all his brilliance and mistakes, but someone capable of revolutionary way of thinking to jolt physics out of it's current stagnant state. Which is specially important since, physicists have turned into almost religious believer of their own, untestable theories, and incapable/unwilling to think outside of that narrow box.


Something important to know about Einstein: he was a patent clerk. He didn’t have much in the way of original revolutionary ideas. He read what other physicists had submitted patents for and combined the work. Without a doubt he did good work and all of mankind greatly benefited from it, but it goes a bit too far to claim all of the revolutionary ideas that came out of him were original.

The most important work, relativity, should be credited to Lorentz and Poincaré.


Einstein was only human... doing "better" and never get stuck on some pet theories might not be humanely possible.


To dimm willis936's post (cancel culture) is an example for the cause physics is stuck. Here is another try to give a reason to dimm:

"Density of mass is _not_ the source of gravity. A difference in density is. There is no difference in density of the Universe (as a whole) compared to 'outside the Universe' since there is no 'outside the Universe'. And therefore gravity does _not_ dominate the Universe. More precisely: there is no gravity of the Universe (as a whole) at all. - To apply the field equations to the Universe as a whole (Friedmann) is nonsense."

It is simple and well known in Germany for years (of course it is still under the regime of U.S.-cancel culture). It doesn't need another Einstein to know his cosmology was wrong. http://www.hashsign.eu


http://cns.physics.gatech.edu/ChaosBook/extras/CNYang.html

"In the next ten years, the most important discovery in high-energy physics is that 'the party's over'."


The party's over because we've done all the easy things and now we need mind boggling machines to test our theories.


Can you explain for a layman what that means?


Experiments needed to confirm "fundamental" physics predictions (eg. large accelerators a la CERN) too costly to be pragmatic, or at least to continue the tradition of timely theory-experiment interplay.

Bear in mind though that the comment was made in the 1980s; people on both sides will debate how prescient he was.

Some background and counterpoints in the context of the proposed Chinese super-collider:

http://ias.ust.hk/ias/files/pdf/1508403651_b2.pdf


Let's finish working on our current experiments first, and then declare that we're stuck. We have particle accelerators, gravitational-wave telescopes, solar probes, and planetary probes which we're either still building or still operating, and we haven't finished falsifying-or-verifying all of our predictions.

It's certainly wrong to imagine that we need a Great Man in order to come and fix everything. Perhaps we need a Great Insight, but also perhaps the insight has already been published and it simply hasn't been well-adopted yet. After all, people are still today fighting against basic century-old quantum mechanics, hoping that reality is more material and determined than it actually is.


It is hard to believe how entrenched locality and realism are in physics. A semester of philosophy should bury them for good. A more modern concern is Boltzmann brains, but those too should offer no barrier to a philosophically aware mind. Perhaps a mere PhD or decade of contemplation would enlighten them to mysteries of the anthropic principle.


> It is hard to believe how entrenched locality and realism are in physics

Not really, since every valid theory must be testable and falsifiable. The idea that all of reality has been a fleeting random agglomeration of particles that experienced this exact moment, of you reading my sentence here, before dispersing and taking all of suggested reality with it - in other words, a Boltzmann brain - well, it's a cool idea, but where does it get you? It's not falsifiable, ever, so it's not science. It's philosophy, and as such it's not something that a scientist, let alone a whole field of scientists, should be spending their time on. Instead, they should be doing actual science, grounded in realism.

If they want to put on a philosopher hat after work and talk about Boltzmann brains, on the other hand, that's great. I, for one, will lap up the crazy philosophies generated by the mind of a physicist, with gusto. But I'd rather they (or so least, most physicists) stick with actual science, aka, their job, most of the time.


Proclaiming "let's kill locality" doesn't give you anything. I mean sure, kill locality. But replace it with what? You need some precise definition that can make predictions. It's like saying "you have to let go of three-plus-one dimensions" or "you have to let go of space being smooth". Okay...and? You need a new, better, testable constraint to replace it with.

Boltzmann brains imply that I should wake up tomorrow with an equal probability of the sun being an orange circle and the sun being a blue unicorn. I've made that bet with myself every night I go to bed for the last twenty years, and it's landed on orange circle every time.

Cute concept, but not buying it.


It's entrenched because relativity specifies a maximum speed. Remove locality and relativity falls apart. Removing it for some cases like correlation but leaving it for causation and communication seems arbitrary. Why? Saying "just because" feels like a hack. What untertheory underlies what is bounded by locality and what isn't? What new predictions does that make and how do you test it? Solve that, and you've got a Nobel Prize waiting for you. Saying "we don't need locality" gets you a noble eye roll.

It's not like math has no concepts without locality. Locality is absent from most of mathematics. The question is, what specifically is the math that models reality? The thing that has locality almost everywhere and for almost everything but a couple exceptions? And what other exceptions can we expect? So it's not that physicists don't accept non locality, it's that so far no mathematical theory that explains the relationship between locality and non locality of correlation precisely in a way that matches reality and makes other testable predictions.


The anthropic principle is one thing, but Boltzmann brains is so close to solipsism that it's worthless for scientific inquiry. While some philosophy background should be mandatory for physicists, I don't see why they should waste so much of their time on Boltzmann brains ?


It is a decent model for ontology, and ontology is good training for observing cause and effect. We grow up to make a lot of erroneous assumptions so we need to train to become good observers. Basically wrestling with the idea lets you observe some of your own biases.


Reminds me of Philip Anderson's testimony against the funding of superconducting super collider.

(need to register to read the article) https://www.the-scientist.com/opinion-old/the-case-against-t...

Incentives in academia seems to be directed towards bringing in grant money. I like Barto and Sutton's discussion of 'nomadic' researchers (8:54, but the entire video is great):

https://www.coursera.org/lecture/sample-based-learning-metho...

I don't know if this is a positive or negative for scientific progress, but, those are just two links that add perspective to academic research/funding.


I have read that in the 1800s the upper crust were being told to not allow their children to pursue a career in Physics. The commonly held view was that Newton had uncovered all the good bits and there was really nothing left to discover.

Then someone asked "What are those blue sparks I see when I take off my jumper (sweater) in a dark bedroom?"

We just need to find the next sparks. :)


This is an anecdote usually told about Max Planck (and may have originally been recounted by Planck!) and his adviser, Philipp von Jolly. Some decent sourcing here:

https://hsm.stackexchange.com/questions/2129/who-said-that-e...

Electricity was a subject of intense study throughout the 19th century and before (think of the people we've named units and effects after - Galvano, Volta, Ampere, Ohm, etc)


It reminds me of a story about Max Planck [1] being told by his advisor to not pursue theoretical physics because it was only the small details that were left to work out. Of course, he went on to help start the revolution in quantum physics.

> in this field, almost everything is already discovered, and all that remains is to fill a few unimportant holes

[1] https://hsm.stackexchange.com/questions/2129/who-said-that-e...


I have a PhD in physics but work on sustainability -- what I consider the most important role for someone who understands nature today. I constantly meet people who are surprised to find that turning on a light in their homes causes fossil fuels to be burned. Their ignorance is their business, but the greenhouse gases become everyone's.


Can I ask what you're working on specifically? As a lapsed graduate-level physicist turned entrepreneur, problems related to climate change have been making me frantically search for a place where I can make meaningful contributions.


Nearly everyone says we know what to do, we just lack the will to do it. We lack leadership -- not telling people facts, figures, doom, gloom, and instructions. That's management, which is important, but leadership -- helping people do what they want to but haven't figured out how. No leaders or people with authority understand nature and I think it's easier for scientists to learn leadership than leaders to learn science.

I support more science, but it's overwhelmingly clear. We don't need more science to create a vision and strategy. People need to feel meaning and purpose, to have an expectation of success. That comes from what leaders work on -- stories, images, role models, systems, emotions, and culture.

So I focus on leadership -- bringing what people like Mandela, Eisenhower, Ali, and Deming brought but I don't see anyone doing today. My TEDx talks https://joshuaspodek.com/tedx and podcast illustrate what I'm working on, though I've advanced a lot since then.

Here are podcast episodes describing my strategy:

- https://shows.pippa.io/leadership-and-the-environment/episod...

- https://shows.acast.com/leadership-and-the-environment/episo...

- https://shows.pippa.io/leadership-and-the-environment/episod...

- https://shows.acast.com/leadership-and-the-environment/episo...


Yeah common people think emissions come mostly form cars. Globally all transportation represents like 15% while electricity production is like 25% of our emissions.



It seems to me that the idea of supersymmetry will have the same fate as the idea of an aether. But physics wasn't stuck at the heydays of the aether theory and it is also not stuck now. There is much more to physics than just supersymmetry:

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

But maybe some fresh ideas from other fields of physics are necessary for particle physics. The problem it seems to me is that the people working in high energy physics and string theory are often very narrow minded and are ignoring important developments outside of their own field.


The time for another Einstein will only be after there has been another Michelson–Morley experiment.


To be fair, Einstein did not know about (or even had a need in) the Michelson-Morley experiment when he created SR.


That’s unclear—Einstein himself made contradictory remarks on this. See http://adsabs.harvard.edu/full/1974Obs....94...81J


I've always been amaZed by the collaboration and synthesis of ideas that happened during Einstein's time. Consider how important and contributive the following were to the famous relativity theories and the other names involved -

- Mach's principle

- Michelson - Morley experiment

- Maxwell's equations of electromagnetism

- minkowsky space-time

- Lorentz transformations

- Riemann's geometry

- Ricci curvature tensor

- Christoffel's connection symbols

- Newton's observation and attempt to account for the perihelion precession of mercury's orbit (not to mention his motion and gravity laws in the first place)

- Schwarzschild's solution to the field equations

- Kerr metric and solutions

- .. and amongst all that we have the Einstein field equations


Yes. It seems that folks forget that Einstein didn't invent all this things.

See also the (now decade old) book The trouble with Physics by Lee Smolin.


I prefer Woit's Not Even Wrong, which has a lot of interesting historical detail and personal anecdote, in addition to the headline polemic.

This one is a fun read


Here is a similar article on powerful members of a scientific community seizing control of the mainstream view and inhibiting scientific progress:

https://news.ycombinator.com/item?id=21911225


Around when Einstein started there were only a few unsolved problems in physics. One gave up quantum mechanics, another relativity. That was basically chance.

Maybe dark matter research will just find dark matter and yield neither more questions nor useful applications. Maybe it will yield a whole new area of enquiry with 1001 useful applications.

We just don't know. No one should assume that X Hours/Dollars/Papers in physics represents an actual amount of "discovery".


Physics needs more data, this is not a single person effort. Actually i’m those who believe this a golden edge for Physics and the constraint is not having another single brilliant guy but resources that could help us validate even further or break what its already known.


Shout out to the sophons from Trisolaris interfering with research.


I always thought it interesting that Liu was tacitly saying that new breakthroughs in physics can only be done in places like CERN with a LHC-like project. Humans couldn't make theoretical breakthroughs anymore.

In some sense, the modern institute of science is somewhat like an orthodox church: a few pre-approved lines of thinking with a swift punishment for heretics.

This is the biggest strength of science: this rigidness of thought is what protects it from clever charlatans. But this is also its biggest weakness because this risk-averse behavior, when scientists can't risk saying unapproved things without torpedoing their own reputation, is why science makes only tiny steps. Right now physics has reached a rather big obstacle on its way and the usual risk-averse-one-tiny-step-at-a-time tactics won't work.


Wasn’t it always like that? If someone actually believes in non-orthodox ideas will simply face the backlash, follow the path anyway and later be recognized for the extraordinary accomplishments if these ideas turn out to be worth the salt.

It’s much better to give a hard time to dedicated people than to risk giving an easy time to charlatans.

Since you can’t really judge novel ideas, you can make the process just hard enough that charlatans drop out to some area where they can get fame and money for cheaper.


One of the challenges with unorthodox ideas is that occasionally some of them lead to great new understandings, but it's a very low signal-to-niose ratio. The orthodoxy can perhaps be excused for their reluctance for entertainment.

Professional scientists can't bet their reputation on a novel idea: a 1% chance it'll work out and 99% chance the scientist will be sent to academia's exile. It's better than what would've happened to them in medieval ages, though.

In past, many of the breakthru ideas came from aristocrats: they were rich enough to not give a shit about their reputation in academia.


Fluff 'science' from Salon is not exactly a resounding endorsement for the article's content.


We need a research satellite around Mars to better understand long range forces. The immediate problem is being bound to Earth distance experiments, including astronomy from two long range reference points. Also need a Venus or Sol orbit solar satellite with a massive capacitor on it for doing obscenely high power experiments.


What long range forces? Almost all probes that go a long way away can (and are) used to test general relativity.

The energy levels needed to brute force more interesting physics is quite possibly beyond being achievable by humanity.

I don't know enough Quantum Field Theory to calculate it on the spot but I have read that to detect any gravitons at all we would need a planet sized detector orbiting closely around a black hole


> obscenely high power experiments

Like the energy released when 2 black holes collide? We're going to need a bigger capacitor.


What would you want studied with that satellite? Which forces?

afaik gravity is pretty well studied by predicting the planets


Gravity is poorly tested at the smallest scales: https://arxiv.org/pdf/1408.3588.pdf . Perhaps space based experiments could help. I also think astronomy would greatly improve.


> Physics is stuck

I agree.

> needs another Einstein to revolutionize it

Nope. The smartest minds today need to work on two topics: AI/AGI, and anti-aging/longevity

Once these two problems are cracked, we'd be on a smoother transition path towards Type-1 Kardashev civilization.

Then the smartest (long-living) minds can sit down with their favorite AGI collaborators in their ivory towers and have all the fun they want to have.


Although another 'Einstein' making breakthroughs is totally possible, AGI gives me a lot of solace as a sort of back-up plan.

If we have indeed hit some type of human wall, AGI at least gives us some long term option to potentially leap it.

I'm not sure how much longevity research would help on that front, given the age distribution of scientists making breakthroughs, but it's nice in its own right.


The longevity is really healthful and youthful longevity, not longevity for its own sake.

I recommend looking into longevity forums, SENS foundation, Aubrey de Grey talks, etc, etc.


I dunno, it might be interesting to have conferences where Einstein, Newton, Hawking, and others were in attendance at the same time.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: