Hacker News Reader

> She didn’t cut it and now is in a non-tenured position at a second rate institution.

What an awful reply to legitimate criticism. I infer that some of your friends at CERN talk about her (quite successful) book this way?

I don't see that line in the original comment, was it edited?

It was edited, but the comment is still ad hominem and avoids talking about the specific criticism.

> > She didn’t cut it and now is in a non-tenured position at a second rate institution.

Lee Smolin, who did "cut it" and who worked in several first rate institutions, wrote about the same thing in his 2006 book The Trouble With Physics. So Sabine Hossenfelder is not alone with these opinions.

If you read Jacques Distler’s blog you might be excused for thinking Smolin didn’t make the cut

That was uncalled for, so I took it out.

How dare you? None of us have ever said something in a comment that was over the line! Um... at least not so far this morning... I think... um... (furtively checks comment history) nope, not me. I don't do that kind of thing (mumbling) so far today...

We've all been there. Props for recognizing it, admitting it, and dealing with it.

It is pretty representative of the state of the academia. You should have left it in, if only as to serve as a reminder that science nowadays is all about status.

Would FIAS (https://fias.institute/en/, https://en.wikipedia.org/wiki/Frankfurt_Institute_for_Advanc...) even be considered a second rate institution?

Ignoring the fact that this is another ad hominem attack, judging a researcher by her/his home institution won't get you anywhere. I know dozens of outstanding mathematicians and physicists who are very well known in their respective fields and reside at less-than-well-known institutions. (Often, because their family is from there and they took the chance to go back.)

Generally speaking, in Central and Western Europe it doesn't matter which institution a researcher is from. They're all good and those at top-tier institutions are only marginally better at best.

She addresses many of your rebuttals quite well here http://backreaction.blogspot.com/2019/03/nonsense-arguments-...

Well, going on to see what should be done instead, we find she advocates for different experiments with the same type of argument she's so gleefully savaging here

> You already know we haven’t found anything yet—otherwise you’d have heard of it. But even null results are valuable guides for theory development. They teach us that some ideas—for example, that spacetime might be a regular lattice—are simply incompatible with observations.[1]

So this is really mostly about a "you had your chance with the toy, now it's time to let others play" argument, though she's very adamant that it's not. This is really the large problem with her criticisms: the case that exists is for particle physics getting less priority, but her argument is that it should stop completely, the collaborations disbanded, and everyone goes to find a new job because they can't be trusted to think clearly.

[1]: http://nautil.us/issue/45/power/what-quantum-gravity-needs-i...

> This is really the large problem with her criticisms: the case that exists is for particle physics getting less priority, but her argument is that it should stop completely, the collaborations disbanded, and everyone goes to find a new job because they can't be trusted to think clearly.

Huh? Where is she saying this? Everything I've read from her suggests that she's against building bigger accelerators, not stop investigating particle physics.

E.g. from http://backreaction.blogspot.com/2019/03/nonsense-arguments-...:

> Science is exploratory and to make progress we should study what has not been studied before, true. But any new experiment in the foundations of physics does that. You can probe new regimes not only be reaching higher energies, but also by reaching higher resolution, better precision, bigger systems, lower temperatures, less noise, more data, and so on.

> No one is saying we should stop explorative research in the foundations of physics. But since resources are limited, we should invest in experiments that bring the biggest benefit for the projected cost. This means the higher the expenses for an experiment, the better the reasons for building it should be. And since a bigger particle collider is presently the most expensive proposal on the table, particle physicists should have the best reasons.

Show we were she actually advocates for building any accelerator and you'd have a case. She violently opposes the FCC despite one of it's explicit goals being precisions measurements of the Higgs.

In fact in the very post you link to she dismisses the idea of high precision measurements: "Yeah, except that I am the one saying we could do better things with $20 billion than measuring the next digits of some constants"

Accelerators aren't the be-all, end-all of particle physics. She explicitly advocates for investigating particle physics in new ways, such as astronomical interaction observations.

Those aren't new ways, it's already being done by astroparticle physicists. You can look at Ice Cube, HESS and Pierre Auger Observatory for examples. Dropping accelerators simply leaves a hole, and those experiments all rely on cross sections from CERN for their predictions by the way.

>Show we were she actually advocates for building any accelerator and you'd have a case.

In point 18 of this article: http://backreaction.blogspot.com/2019/03/nonsense-arguments-..., her first suggestion for an alternative avenue of research is

> (a) Astrophysical and cosmological observations attributed to dark matter. These are discrepancies between theory and data which should be studied closer, until we have pinned down the theory. Some people have mistakenly claimed I am advocating more direct detection experiments for certain types of dark matter particles. This is not so. I am saying we need better observations of the already known discrepancies. Better sky coverage, better resolution, better stats. If we have a good idea what dark matter is, we can think of building a collider to test it, if that turns out to be useful

> Isn’t it nice (...) advance our understanding of the fundamental forces of nature?

The criticism is that the ROI = (the potential advance) / (invested money) is too small. As you pointed out it can be argued that the real goal is elsewhere: in the engineering / technology or science of complex systems. Creating a big machine that verifiably works brings many unexpected advances in many fields. And what is more important it is a symbol of human advancement which unites people and inspires them to pursue science. Since humans need a clear goal to invest much resources in a constructive endeavor. It may not be that important though that the ROI in fundamental physics is small, as long as people work hard together. But it can be argued that the whole process begins to resemble religion more than a scientific method; which can be problematic as religions have a bad reputation among scientific community.

An organisation like CERN is extremely hard to replicate and easily destroyed and dragged down by a thousand cuts. The EU is funding similar moonshots in Neuroscience , Graphene, Quantum Computing and most recently machine learning. They all provide case studies what can go wrong in such a big science project (in fighting, diverging goals, huge administrative overhead, bureaucracy, political grand-standing).

The people working at CERN directly and their collaborators truly belief in the goal of understanding the fundamental forces of nature. And they had tremendous success doing so. They are the gold standard for a well managed long running big science project. Just recently they found the Higgs boson. Hossfelder worked for years on extremely speculative ideas in quantum gravity, they weren’t super well received, most people liked string theory better. Bottom up and top down reasoning in particle physics phenomenology suggests there are more discoveries to be made at the energy frontier. I like to see a future where eventually we build a solar system sized particle accelerator not stop at a radius of ~100 km

> ROI = (the potential advance) / (invested money) is too small

That's unknown in advance. To take example from maths, the conic sections, braid theory, number theory—they all used to be very abstract and un-applicable parts maths, until one day.

If it's unknown in advance surely it's more rational to spread the research out instead of putting most of the available eggs in a single basket.

CERN found Higgs - which was proposed decades before CERN was built. There is nothing like the Higgs proposal around today.

That's the real problem. Spend $20bn on some new models, and then you can spend $20bn chasing the most interesting prospects.

That's far more likely to get you somewhere exciting than the experimental equivalent of throwing more spaghetti at the wall and hoping some of it does something unexpected.

> Spend $20bn on some new models, and then you can spend $20bn chasing the most interesting prospects.

This is a very incorrect description of the problem. The situation in basic particle physics is that models are super plentiful; Six months in 2016 saw 500 different papers written to explain a novel signal that ended up not being real, so it's not models that are missing.

"CERN found Higgs - which was proposed decades before CERN was built." ---> This is factually incorrect.

   * CERN: founded 1954, as a lab for nuclear physics

   * PRL electroweak symmetry breaking papers: 1964

Professor Hossenfelder and others have written long books about why the probability of finding new physics based on not liking the aesthetics of the current maths (naturalness, supersimetry, grand unification) is extremely likely to fail, and has been constantly failing for ~50 years.

If our way of generating new theories has given bad predictions for 50 years, at what time should we stop and try out other ways of extending the confirmed theories?

It is unknown in advance if putting your hand through a fire will hurt you this time as well, but after 50 years of doing so, perhaps it's time to stop.

They make qualitative arguments, nobody has any realistic prediction of what the probabilities are.

And the problem has never been making new theories, the problem is always that we have a very good model, and no principled way for choosing between all the possible theories that gives rise to almost precisely that model.

So people tried to invent reasons and so far nobody has had any luck. And Hossenfelder is correct that in hindsight naturalness wasn't a very good rule.

But by the same hindsight, no theory that could be detected in previous experiments would have worked either. In this view, making the experiments at all was a complete waste, we should just have drawn the winning theory from a hat and stopped all further work.

The point is that there are known problems in the foundations of physics (the measurement problem, the reconciliation of general relativity and quantum mechanics to name the 2 best known ones) and that they should be by far the biggest focus for new research - not GUTs and solutions for un-naturalness.

If we have a theory that could solve the measurement problem (or detect quantum gravity etc.), and a plausible experiment for that theory, we should perform this experiment - it may well yield valuable data whatever result we get.

Conversely, if we have a theory that predicts that more particles may exist at higher levels of energy and nothing else, then there is no good reason to perform this experiment. Especially if the precise level of energy is a free parameter of the theory, so the theory won't change if the experiment fails.

> And Hossenfelder is correct that in hindsight naturalness wasn't a very good rule.

I believe that she is correct that naturalness wasn't a very good rule at all, not just in hindsight. It's as good a rule as it would have been to expect all numbers to be multiples of Pi.

To claim that the standard model has nothing to do with the quantization of gravity is very strange, part of the motivation for a GUT is that it would be a huge help towards making a theory of everything. And a TOE is precisely a unified treatment of the standard model forces and gravity.

The measurement problem isn't a physics thing, just like naturalness that's more of a life-style choice. In fact, the only reason to have problems with wave-function collapse is precisely a naturalness reasoning, technically it makes very accurate predictions just fine.

> part of the motivation for a GUT is that it would be a huge help towards making a theory of everything. And a TOE is precisely a unified treatment of the standard model forces and gravity.

A GUT is specifically the part of a TOE that has nothing to do with gravity... A TOE is a GUT + a theory of quantum gravity.

If we had the Standard Model + a theory of quantum gravity, this could be a complete theory of the universe (pending explanations for dark matter, dark energy etc.). These could very well be completely separate phenomena, there is no reason to believe they reduce to a single phenomenon - that there exists a TOE.

> The measurement problem isn't a physics thing, just like naturalness that's more of a life-style choice. In fact, the only reason to have problems with wave-function collapse is precisely a naturalness reasoning, technically it makes very accurate predictions just fine.

This isn't entirely accurate. There is a very clear quantitative question behind the measurement problem: what kind of system constitutes a measurement device, in the sense of invoking the need for the Born rule? There must be some precise size/kind of system that, when a particle interacts with it, you can no longer use Schrodinger's equation to predict its movement after that interaction - you must apply the Born rule. We know that this doesn't happen after a particle collision, but that it does happen after a collision with a "detector".

This is a clear empirical question (you don't need to call it a problem), well within the realm of physics. It may be extremely difficult to answer, but I see no reason to imagine it is a priori impossible to resolve.

>This is personal to me, some of my friends work at CERN

That is very clear. You didn't engage her argument at all.

Is is really too much to ask for scientific mega project to generate all that value, while at the same time same pursuing a central object that is scientifically viable?

Building a high energy collider is like building a larger telescope. The only things you can be sure of is that you will be able to understand whatever you have seen better than before. There is scientific consensus that we should be able to understand dark matter better by building one aswell. This is supported by technical arguments which are hard to communicate to the general public. Attacking this communication to the general scientific community is different from attacking the substance of these technical arguments. One can be done by a blog post, the other would need to be accepted by a scientific journal. Physical review letters would be more than happy to publish substantive attacks on prevailing beliefs about dark matter.

>>There is scientific consensus

Really - I thought that we were debating an article by a scientist, who along with many other scientists disagrees with that position.

>>This is supported by technical arguments which are hard to communicate to the general public.

And yet the general public is being asked to pay for it. Perhaps sorting out that gap and explaining how understanding dark matter will be as important as (for example) curing cancer or alzheimers disease, or more prosaically cystic fibrosis or malaria would be a good thing? Or even understanding ecologies and the global climate? Will the new accelerator be carbon neutral? Will understanding dark matter prevent the collapse of our planet and the deaths of our children?

If I were a politician funding this project I would want some pretty cold, hard information on these things.

Well, it's a scientist in a different field. More importantly, it's by a scientists that doesn't engage with the technical arguments made by those actually in the field, so it's unclear what exactly the label "scientist" brings.

If a physicist criticises recommendations by epidemiologist without reference to the theory of how diseases spread, should you count that as disagreement from the consensus of how to fight infectious diseases?

> Will understanding dark matter prevent the collapse of our planet and the deaths of our children?

It's difficult to know. General relativity is useful for GPS. Superconductivity is useful for NMR that actually sometimes saves children from cancer. Neither of the applications were clear at the discovery time.

(My favorite was Magnetoresistance. You start to explain it as a weird quantum boring effect of the spin of electrons in magnetic conductors, and when the other person is convinced that it totally unuseful you explain that it is used in hard disks. SSD are killing the fun.)

GR-> GPS ~50 years?

Superconductivity -> NMR ~60 years?

Do we have that long?

She is not a scientist active in this area. At this point she is for the most part paid to be a science communicator. A lot of money is spend on curing all these other things. Particle accelerators have been instrumental in developing the technology for hadron radiation therapy (there are now special purpose one just for cancer treatment) and superconducting magnets (useful for MRT medical devices). If you examine it closely a lot of advances in biology and medicine were thanks to technical advances made by physics. They have been a really solid bet so far and have created trillions of dollars in value. Why not let them have some fun and let them chase their McGuffin, which btw. might be totally real and interesting fundamental physics. The high energy particle physics community is doing ground breaking research on radiation hard imaging sensors, large scale super conducting power distribution (which might be useful for connecting solar power in a lossless manner to consumers).

These projects are a way to train high skilled STEM people (competition to stay at CERN is brutal, but most go on to industry) in a structured way that then work on a lot of other societally relevant problems.

That second paragraph is a really terrible justification.

> The only things you can be sure of is that you will be able to understand whatever you have seen better than before

I think it's more correct that you will be generating observations in largely the same way as in the past, but with higher resolution and fidelity.

> Web browsers and HTML were invented at CERN.

More structural funding for open source software would also be welcome. Software that deals with communication (browsers, protocols, email, filesystems, codecs, etc.) should be fully open and developed by universities and government institutions, imho.

I think the argument being made there is that when you gather smart people together, with resources to work on a hard problem, they will tend to invent things on the side which we didn't know we needed.

An academic committee to decide what software was worth funding, in the 80s, would probably have come up with a very different list to yours. And would have hired people from safer, more academic, backgrounds than Sir Tim.

>developed by universities and government institutions

The NSA called to say that they like your thinking there!

Legitimate security and cryptography work from the NSA is always welcome.

One could argue that spending a lot of money on a hyper advanced anything would lead to a bunch of cool technologies, even if the primary objective yielded no results. Why spend many billions of euros on something speculative when one could spend the money on something much more likely to produce both primary and secondary benefits?

Point me to something that has these characteristics and hasn’t received funding. Anderson made the case to fund many small scale experiments when it came to funding the SSC, but that doesn’t have the same secondary benefits.

Off the top of my head...

- Large-scale climate engineering

- Digitizing human consciousness

- Strong AI

- True life extension

- A replacement for the ISS

If you limit your characteristics to "involves high energy particle physics" it's pretty easy to put blinders on and pretend there aren't large challenges facing humanity that could potentially be solved by hard science progress and a lot of funding.

Well what I meant show me plans with 1000s of pages of itemized highly realistic goals that have a chance of succeeding and yield real technological advances over a period of 30 years with 100s of professors putting their professional careers on the line for a common goal. The overlooked value of the HEP community is that they have moved to the industrial age of scientific endeavour, while almost everyone else is stuck in the Middle Ages.

It would be great if people managed to organized around any of the goals you mentioned, so far it hasn’t happened at a comparative scale for various reasons.

> Web browsers and HTML were invented at CERN.

Theory of Relativity was invented at the Swiss Patent Office.

Which is a curious argument since it compares the free time activities of a patent officer with the activities of a scientist trying to improve his workplace communication. I.e., apples with oranges.

Well, apples are obviously superior, because you can't make an apple pie with oranges.

I think it is wrong to cherry pick the HTML example. It would have been invented outside (wasn't SGML already a thing?).

Compared to open source in general the CERN stacks are horrible, typical sloppy academic software of no use to the general public. Tons of stuff like the cling C++ interpreter that look interesting but are never polished to be generally useful.

cling ? clang perhaps?

Truth is, I did get your point. Perfect refinement is not necessary in every context. Your comment, like CERN's interpreter, though sloppy was fit for purpose.

https://github.com/root-project/cling

Thanks. I retract the specific objection.

There have been a lot of bad ideas in markup languages. Check out some of the ideas tossed around in the creation of HTML. Plenty of disasters averted by pragmatic thinkers and good luck.

The web browser enabled my whole very comfortable career so I'm willing to overlook a sloppy C++ interpreter here and there.

Depends on who you ask I guess, present value of root has been estimated[1] at around 2.8 billion Euro.

[1]: https://www.youtube.com/watch?v=0Zpeh9wuzMs&feature=youtu.be...

I've never heard of anyone outside of physics using ROOT. Have you? The tech open source data stacks are their own quickly maturing but completely separate stacks.

> Generally speaking developing a new [HUGE PROJECT] leads to excellent technical advances at all levels of the stack.

It's true for accelerators and a huge number of other things too. If this isn't unique to accelerators, it's not really a case for accelerators specifically.

> Isn’t it nice that a fraction of the smartest people on the world are not working on getting you to click Ads and are using machine learning to advance our understanding of the fundamental forces of nature?

This isn't a good faith argument. What if we got our smartest people working in hospitals or schools or orphanages instead of getting you to click ads? Maybe that's where the money should go. Think of the children.

Agree.

> Generally speaking developing a new accelerator leads to excellent technical advances at all levels of the stack. Web browsers and HTML were invented at CERN.

The exact same argument can be made for military spending. The internet was invented at...

It is always such a dishonest argument to talk about spin-off benefits.

Its not zero sum, but the sum isn't infinite, especially for research without direct commercial applications. It would be nice to fund everything, but scientists have to get lot more politically engages before that sort of political support is even on the table

> Whatever theorists predict is not the only reason for them to do it, higher energies are a technical challenge, a data processing challenge, something to focus your attention and creativity on.

I don't understand this argument. We should prefer a particular experiment not because it is more likely to answer a research question but because it, itself, is more challenging?

There are few other ways of probing high energy physics systematically and a lot of them are funded and done aswell. This is not a zero sum game. We can bury detectors in the antarctic, in mines and have CERN. In fact technology developed at CERN helps those other experiments aswell.

Fine, but that's not the interesting part of Hossenfelder's claim. Suppose we couldn't do it all and could only do one thing first. What is wrong with her claim that a bigger collider shouldn't be that thing? Is she wrong that that is not the most promising investment for discovery or is she wrong that "most promising investment" should be the main criterion for choosing what to do first?

But we can do it all. Moreover a large number of talented people really want to do it. Whatever money they are asking for is a fraction of the cumulative benefits these people have provided to humanity. When arguing about money you never argue about "science discoveries", no-one in their right mind cares about them when it comes to fundamental science, when you talk about this level of funding. If you manage to build a 100 TeV detector you need to figure out a mind boggling number of practical technical challenges along the way. All of those sum up to a value that is far larger than 10 billion. For other projects like the Human Genome project or the Brain projects the ROI seems more immediate. But in fact physics projects have routinely yielded higher ROIs. There is an extremely long tail of physics / technology research that is funded by a new detector.

As another example take the EU quantum initiative, they managed to get ~1 Billion and have promised to get ~5000 people working on it for ~10 years. That works out to ~20000 per person / per year. This only works because a lot of these people are so passionate about this project, that they are willing to self exploit themselves for years.

Fine, but if we couldn't, does she have a point about the priorities, and if not, why not?

(And, BTW, the fact that some investment has a good ROI doesn't mean you should pick it over another with a higher ROI)

I honestly don’t think she is qualified to speak on the priorities of a discipline she has never been involved in. They will do and prioritise what they think is reasonable based on fierce internal discussions and consensus building. I personally was never taken by the prospect of working on a huge project like the various experiments at LHC or any of the successors. If you defund HEP these people won’t magically go work on something else together. A large pool of people will simply stop doing physics and dissipate. Let them do what they want, it has worked out pretty well so far. There is plenty of other research that is well funded and competing for talent and attention. Why give up on something where Europe has clear worldwide leadership.

I agree with you. I would still love to work in academia, but when the work required to get proper funding is tied to the extension of your contract it can lead to some sleepless nights.

In the book God Particle, the author describes how the funding for colliders in USA stopped after the cold war and it was extremely demoralizing even as a reader.

You are repeating the nonsense arguments Sabine has countered before http://backreaction.blogspot.com/2019/03/nonsense-arguments-...

I agree with her. These are nonsense arguments.

> Science shouldn’t be a zero sum game.

(2. The “No Zero Sum” argument)

> new accelerator leads to excellent technical advances at all levels of the stack

5. It is not a waste of money

11. It may lead to spin-offs.

12. A big particle collider would benefit many tech industries and scientific networks.

BTW her examples of experiments that received insufficient funding is really curious: FAIR she just happens to work at a university that is pretty directly involved with it, the cost overruns were so bad that the original people lost management of the project . The James Webb Telescope was in part delayed because it shares the platform with the next gen of American spy sattelites. I happen to know someone that worked on the sensors for both. Sensors that btw. share a lot of similarities with the sensors used in particle detectors.

Her main point is that from physics perspective there are better ways to use money. Whatever success and failures in projects, we should keep throwing money into places that are more likely solve important physical questions.

Any argument that sidesteps that is nonsense. If goal of investing in fundamental physics is not R&D it's just a side effect as I assume.

If I read your argument correctly it's "We can't give argument that this is the best way spend money if you want to solve these problems, but it will generate other side benefits that may help indirectly". It's honest argument if put that way.

The particle community has enumerated why new accelerators are the best way to answer particle physics questions, Hossenfelders position is however that those arguments aren't worth engaging with because particle physicists are slaves to "group think".

As for the argument that cosmology or astronomy are more promising ways of yielding new insights, that's not something she ever bothers to discuss in detail. There's no articles about how the extremely noisy and uncertain constrains from astronomy are more worth to gamble on that than the much cleaner direct detections from accelerators.

And there's certainly no arguments about how we should stop giving so much money to the exoplanet researchers and their instruments instead of making satellites to measure the cosmic radio background. (There's not even a little post about how SpaceX will hamper radio astronomy by making sure there's no radio quiet place left on earth).

>The particle community has enumerated why new accelerators are the best way to answer particle physics questions

They have not given a good answer why particle accelerators they are asking for would give any answers at 100 TeV.

It's just biggest they can hope to get.

There are worthwhile undertakings other than particle physics

But we wouldn’t know what the things explored by those undertakings are made of.

I usually have to bite my tongue when it comes to Sabine's ramblings. BTW, I would love if anyone could point me any phenomenology paper of hers, I can't find them. Maybe when she single-handedly finishes with the slaying of the particle physics dragon she might tackle idk, condensed matter physics, she has the same kind of background there so who knows.

Anyway, I'd like to link Dorigo's post where this anti-science stirrup might have shown up first:

https://www.science20.com/tommaso_dorigo/what_it_means_to_be...

I don't find it surprising tbh.

That is a nice article! Thanks for pointing it out. Most of the time I try to bite my tongue aswell. It is just sometimes really upsetting.

" Super conducting magnets at scale were pioneered for LHC" ... wait... no love for HERA (DESY) here?

People really need to stop quoting tangentially invented technologies as the justification for CERN to exist, if I was to gather the same group of scientists to ask them to build the best Lego set possible with 50bil they would have also produced the same tangential technologies.

> Science shouldn’t be a zero sum game.

Poor people "shouldn’t" die from simple diseases. Is this the game, we are wishing away things we don't like?

Science 100% is zero sum. Especially talking the figures we are here.

How about we start with creating free science text books (that are not garbage) for the world then we'll let you blow multi-billions on your religion?

Pretty simple neither Hossfelder, nor politicians, nor you get to tell people what they would like to do. At this point it becomes a power struggle and so far mainstream physics has won that power struggle, because we have good arguments on our side. Investment in particle physics has had large net positive effects on society at large. Why kill the golden goose just because you aren't interested in whatever these people are interested in. Science doesn't enter this argument at all, purely from a technology and human capital level this is close to optimal use of their talents. You have a bunch of high y motivated, talented people working towards a common goal. Even if you don't perfectly understand why they would do that, their results alone are convincing enough to let them continue.

Good arguments? It was won in the same manner that the first motorway built in GB happened to be between Oxford and London (in that order).

Right good arguments in the political sense. A highly organized, well connected, motivated group of people that write excellent funding proposals, with a track record of Nobel price worthy results and the ability to execute on >10 billion dollar projects on the time scale of decades. There aren’t that many people like that around and they happen to like to smash atoms.