Not Even Wrong

• I was sorry to hear of the death a few months ago of Tony Smith, who had been a frequent commenter on this blog and others. Unfortunately my interactions with him mainly involved trying to discourage him from hijacking the discussion in some other (often unusual) direction. Geoffrey Dixon did get to know him well, and has written a wonderful long blog entry about Tony, which I highly recommend (Dixon’s newish blog also has other things you might find interesting).
• On the Jim Simons front, the Simons Foundation has put together something to celebrate their 25th anniversary. It explains a lot about their history and what they are doing now, as well as giving some indication of plans for the future. On these topics, read pieces written by Jim Simons and Marilyn Simons. The Foundation has been in a high growth mode, having an increasingly large impact on math and physics research. Their main statement about the future is that the plan is for this to go on for a very long time:
  

  According to its bylaws, the Simons Foundation is intended to focus almost entirely on research in mathematics and science and to exist in perpetuity. If future leadership abides by these guiding principles, Marilyn and I believe the foundation will forever be a force for good in our society.

  My impression is that the Simons children have their own interests, and foundations with other goals to run.

  News from the \$75 billion source of the money (RenTech) today is that Simons is increasingly turning over control of that to his son Nathaniel, who has been named co-chairman. He has also added four new directors to the board, four of them senior Renaissance executives, and one his son-in-law Mark Heising.

• There are various IAS-related videos you might want to take a look at:

  Pierre Deligne explaining motives last night.

  Michael Douglas on the use of computers in mathematics.

  A Dutch documentary (not all of it is in Dutch…).

• If you aren’t regularly reading Scott Aaronson’s blog, you really should be. Latest entries are a detailed report from Davos and a guest post with a compelling argument about a major factor behind the problem of why women leave STEM careers more than men.
• For the latest on the “It from Qubit” business, see talks at a KITP conference. John Preskill notes “lingering confusion over what it all means”, which makes me glad to hear that I’m not the only one…

Sabine Hossenfelder has a new piece out, making many of the same arguments she has been making for a while about the state of fundamental theory in physics. These have a lot in common with arguments that Lee Smolin and I were making in our books published back in 2006. The underlying problem is that the way theorists successfully worked up until the seventies is no longer viable, with the Standard Model working too well, up to the highest energies probed:

The major cause of this stagnation is that physics has changed, but physicists have not changed their methods. As physics has progressed, the foundations have become increasingly harder to probe by experiment. Technological advances have not kept size and expenses manageable. This is why, in physics today, we have collaborations of thousands of people operating machines that cost billions of dollars.

With fewer experiments, serendipitous discoveries become increasingly unlikely. And lacking those discoveries, the technological progress that would be needed to keep experiments economically viable never materializes. It’s a vicious cycle: Costly experiments result in lack of progress. Lack of progress increases the costs of further experiment. This cycle must eventually lead into a dead end when experiments become simply too expensive to remain affordable. A $40 billion particle collider is such a dead end.

I have a somewhat different view about a potential next collider (see here), but agree that the basic question is whether it will be “too expensive to remain affordable.”

What has happened over the last forty years is that the way HEP theory is done has become dysfunctional, in a way that Hossenfelder characterizes as follows:

Instead of examining the way that they propose hypotheses and revising their methods, theoretical physicists have developed a habit of putting forward entirely baseless speculations. Over and over again I have heard them justifying their mindless production of mathematical fiction as “healthy speculation” – entirely ignoring that this type of speculation has demonstrably not workhat woed for decades and continues to not work. There is nothing healthy about this. It’s sick science. And, embarrassingly enough, that’s plain to see for everyone who does not work in the field.

This behavior is based on the hopelessly naïve, not to mention ill-informed, belief that science always progresses somehow, and that sooner or later certainly someone will stumble over something interesting. But even if that happened – even if someone found a piece of the puzzle – at this point we wouldn’t notice, because today any drop of genuine theoretical progress would drown in an ocean of “healthy speculation”…

Why don’t physicists have a hard look at their history and learn from their failure? Because the existing scientific system does not encourage learning. Physicists today can happily make career by writing papers about things no one has ever observed, and never will observe. This continues to go on because there is nothing and no one that can stop it.

This story brings up a lot of complex issues in the philosophy and sociology of science, but to me there’s one aspect of the problem that is relatively simple and deserves a lot more attention than it gets: how do you get theorists to abandon failed ideas and move on to try something else?

The negative LHC results about SUSY have had some effect, but even in this case it’s remarkable how many theorists won’t abandon the failed idea of a SUSY extension of the Standard Model. This was always a highly dubious idea, explaining nothing about the Standard Model and adding a huge number of new degrees of freedom and more than a hundred new undetermined parameters. Not seeing anything at the LHC should have put the final nail in the coffin of that idea. Instead, I see that this past fall MIT was still training its graduate students with a course on Supersymmetric Quantum Field Theories. You can try and argue that SUSY and supergravity theories are worth studying even if they have nothing to do with physics at observable energies, but it is a fact that these are extremely complicated QFTs to work with and have explained nothing. Why encourage grad students to devote the many, many hours it takes to understand the details of this subject, instead of encouraging them to learn about something that hasn’t been a huge failure?

The techniques one gets trained in as a graduate student tend to form the basis of one’s understanding of a subject and have a huge influence on one’s future career and the questions one has the expertise needed to work on. Besides SUSY, string theory has been the other major course topic at many institutions, with the best US grad students often spending large amounts of time trying to absorb the material in Polchinski’s two-volume textbook, even though the motivations for this have turned out to also be a huge failure, arguably the largest one in the history of theoretical physics.

To get some idea of what is going on, I took a look at the current and recent course offerings (on BSM theory, not including cosmology) at the five leading (if you believe US News) US HEP theory departments. I may very well be missing some offered courses, but the following gives some insight into what leading US departments are teaching their theory students. Comparing to past years might be interesting, possibly there’s a trend towards abandoning the whole area in favor of other topics (e.g. cosmology, quantum information, condensed matter).

• Harvard:
  Fall 2019
  PHYSICS 283B: Spacetime and Quantum Mechanics, Total Positivity and Motives
  PHYSICS 287A: Introduction to String Theory
  Spring 2020
  PHYSICS 211BR – Black Holes from A to Z
  PHYSICS 287BR – The String Landscape and the String Swampland
• Stanford:
  No courses beyond QFT in 2019/20
• Caltech:
  No courses beyond QFT in 2019/20
  [Actually, see comments, there’s Physics 230a last fall, which may or may not have been a course on SUSY models]
• Princeton:
  Fall 2019
  Phy 540 Strings, Black Holes and Gauge Theories (Klebanov)
  Spring 2020
  Phy 540 Strings, Black Holes and Gauge Theories (Polyakov)
• MIT
  Fall 2019
  8.831 Supersymmetric Quantum Field Theory
  Spring 2020
  8.851 Effective Field Theory

The places not offering string theory courses this year seem to have had them last year.

Update: Something relevant and worth reading that I think I missed when it came out: Jeremy Butterfield’s detailed review of Lost in Math, which has a lot about the question of why theorists are “stuck”.

Update: There’s some serious discussion of this on Twitter. For those who can stand that format, try looking here and here.

Update: Mark Goodsell has a blog posting about all this here, including a defense of teaching the usual SUSY story to graduate students.

Update: A correspondent pointed me to this recent CERN Courier interview with John Ellis. Ellis maintains his increasingly implausible defense of SUSY, but he’s well aware that times have now changed:

People are certainly exploring new theoretical avenues, which is very healthy and, in a way, there is much more freedom for young theorists today than there might have been in the past. Personally, I would be rather reluctant at this time to propose to a PhD student a thesis that was based solely on SUSY – the people who are hiring are quite likely to want them to be not just working on SUSY and maybe even not working on SUSY at all. I would regard that as a bit unfair, but there are always fashions in theoretical physics.

To start the new decade there’s an article very much worth reading by Misha Shifman, entitled Musings on the Current State of HEP. It’s somewhat of an update of something he wrote back in 2012, which I wrote about here. He starts off with:

Now, seven years later, I will risk to offer my musings on the same subject.The seven years that have elapsed since [1] brought new perspectives: the tendencies which were rather foggy at that time became pronounced. My humble musings do not pretend to be more than they are: just a personal opinion of a theoretical physicist… For obvious reasons I will focus mostly on HEP, making a few marginal remarks on related areas. I would say that the most important message we have received is the absence of dramatic or surprising new results. In HEP no significant experimental findings were reported, old ideas concerning Beyond the Standard Model (BSM) physics hit dead-ends one after another and were not replaced by novel ideas. Hopes for key discoveries at the LHC (such as superpartners) which I mentioned in 2012 are fading away. Some may even say that these hopes are already dead. Low energy supersymmetry is ruled out, and gone with it is the concept of naturalness, a basic principle which theorists cherished and followed for decades. Nothing has replaced it so far…

HEP, “my” branch of theoretical physics since the beginning of my career, seems to be shrinking. A change of priorities in HEP in the near future is likely as business as usual is not sustainable. The current time is formative.

I encourage you to take a look at the rest, there’s a lot more detailed discussion of the state of HEP and allied fields, especially about the central role of quantum field theory.

Shifman also includes a section very critical of Richard Dawid, the “non-empirical confirmation” business and talks given at the “Why Trust a Theory?” conference (discussed here):

With all due respect I strongly disagree with Richard Dawid and all supporting speakers at the conference and beyond… I object against applying the term “non-empirically confirmed” to science (the more so, the term “postempiric science”). Of course, we live in liberal times and everybody is entitled to study and discuss whatever he or she wants. But the word science is already taken. Sorry, colleagues. For “postempiric science,” please, use another word, for instance, iScience, xScience, or something else.

As for David Gross’s attempt to claim that string theory is, like quantum mechanics and quantum field theory, not testable just because it is a framework, not a theory, Shifman is having none of it:

David Gross is a great theoretical physicist, whose discovery of asymptotic freedom made him immortal, but I respectfully disagree with him. Framework or not, both QM and QFT have absolutely solid confirmations in all their aspects in thousands of experiments.

As for the once popular idea that string theory could provide a “theory of everything”, he writes:

Well… it never happened and – I will risk to say – never will.

At some point within the past couple years I noticed that one blog that had Not Even Wrong on its blogroll was the blog of Dominic Cummings, who was often getting credited with masterminding the political campaign that got the British to vote (narrowly) for Brexit in 2016. Cummings has had further success recently as Chief Special Adviser to British Prime Minister Boris Johnson, with a blow-out election victory three weeks ago putting him securely in control of the British state.

Today on his blog Cummings has, invoking Grothendieck, posted a job advertisement: ‘Two hands are a lot’ — we’re hiring data scientists, project managers, policy experts, assorted weirdos…. He’s looking for mathematicians, physicists and others to join him to change British society, working

in the intersection of:

• the selection, education and training of people for high performance
• the frontiers of the science of prediction
• data science, AI and cognitive technologies (e.g Seeing Rooms, ‘authoring tools designed for arguing from evidence’, Tetlock/IARPA prediction tournaments that could easily be extended to consider ‘clusters’ of issues around themes like Brexit to improve policy and project management)
• communication (e.g Cialdini)
• decision-making institutions at the apex of government.

For some other descriptions of who Cummings would like to hire, on the economics side there’s:

The ideal candidate might, for example, have a degree in maths and economics, worked at the LHC in one summer, worked with a quant fund another summer, and written software for a YC startup in a third summer!

We’ve found one of these but want at least one more.

He also wants “Super-talented weirdos”, with examples given from William Gibson novels, such as “that Chinese-Cuban free runner from a crime family hired by the KGB.”

The remarkable things to me about this long document are what it doesn’t contain. In particular I see nothing at all about any specific policy goals. Usually a new government would recruit people by appealing to their desire to make the world a better place in some specific way, but there’s nothing about that here. The goal is to control the government and what the British population believes, but to what end?

In addition, a more conventional hiring process would be asking for candidates of high ethical values, with some devotion to telling the truth. Cummings seems to be asking for exactly the opposite: best if your background is “from a crime family hired by the KGB.”

Best of wishes to my British readers, now joining the US and other nations in a new dystopic post-truth era. It’s massively depressing to me to see how this has worked out here, I hope you do better. Maybe you should be sending in your applications to Cummings and hoping to sign up for a role in the new power structure. If so, tell him “Not Even Wrong” sent you…

Update: For more on Cummings, there’s a good Financial Times article.

For presentations a couple days ago at the latest HEPAP meeting, see here. One piece of news, from this presentation, is that there likely will be a delay in the scheduled startup of the HL-LHC, with the next LHC run (Run 3) extended for an additional year (through 2024), and the next shutdown (LS3) extended by a half year. The HL-LHC would then start physics in 2028.

Most of the HEPAP discussions have to do with funding. The pattern of recent years has been one of huge decreases in funding proposed by the Trump administration. These are completely ignored by both the Democrats and Republicans in Congress, which passes large increases in funding (then signed into law by Trump). For FY2020 this continues: at DOE the HEP budget for FY2019 was \$980 million, for FY2020 the White House budget request was a massive cut to \$768 million. This was taken no more seriously by anyone than the last few of these, with the FY2020 House Mark \$1,045 million, the Senate Mark \$1,065 million. The FY2020 budget remains to be finally finished and passed, in the meantime the federal government has been operating under a sequence of continuing resolutions.

Specifically on theory funding, JoAnne Hewett has a presentation on The State of Theory. It has no numbers in it, but the DOE numbers given here show an increase from \$60 million in FY2017 to \$90 million in FY2019 for Theoretical, Computational and Interdisciplinary Physics. But within this category, pure theoretical HEP is pretty flat, with big increases for Computational HEP and a huge new investment in Quantum Information Science (\$27.5 million in FY2019). There does seem to have been some sort of decision to de-prioritize conventional theoretical HEP in favor of newer trendy areas.

Hewett describes the general consensus on current problems with theory funding as

• Universal concern on ever decreasing levels of funding for university groups: concern that university programs are dying.

  -Private institutions attempt to offset cuts with non-federal funding sources.
  -Cuts to program further accumulated in 2019. Many postdocs learned in May 2019 that their contracts would not be renewed for the fall. It was then too late to apply for new positions.

• Lab theory programs are also losing researchers.
• Even distribution of cuts across U.S. theory program has indirect proportional effect to small programs.
• Large fluctuations cycle-to-cycle is making groups less cohesive and more inclined to opt for “safer” research projects.
• There is the perception that the recent emphasis on QIS comes at a cost to more traditional HEP theory research.
• Summer salary has been capped or reduced to 1 month in many cases. Removal of summer salary across the board is demoralizing.

and ends with

The situation is becoming increasingly unstable.
University-based theory is suffering its most serious crisis in decades.
Its future is in jeopardy.

It would be interesting to see some numbers on the size of new private research funding going to HEP theory (for instance funding from the Simons Foundation or the private funding of the CMSA at Harvard). I don’t know of such numbers but I’m curious whether what is happening is that the total funding level has seen reasonable growth, but increases in funding are going to a small number of elite institutions, with the rest of the field in decline.

On the question of caps or reductions in summer salary, I doubt that any significant number of researchers is reacting to only getting 1 month of summer salary by signing up for another job (e.g. teaching summer school) and not doing research during the other two months of the summer. There has been another huge influx of money to the field that in some sense replaces grant-funded salary supplements: the multi-million dollar Breakthrough Prizes. A sizable number of HEP theorists have now partaken in all or part of one of the \$3 million prizes. If you add in this money, on average HEP theorists may have been seeing significant increases in income, however with almost all of it going to a small number of people (at the same elite institutions that are doing well). What we’re seeing may just be the same trend as in the rest of the US economy: a move to a star system with ever larger increases in inequality.

Another problem for the field of HEP theory may be that funding is stagnating because the DOE and NSF are skeptical about its intellectual health. Hewett notes that “Formal theory resides solely in university environment and has undergone significant funding cuts.” Trying to make the positive case for this part of the field, she lists three areas of advances, but oddly, the first two are identical. The two areas of advances in formal theory she describes are:

Advances in strongly coupled quantum field theory (gravity/field theory duality, bootstrap program, amplitudes) has implications for particle physics, cosmology and beyond.

Geometric advances in particle physics constructions from String/F-theory has implications for the “swampland program”.

For the second of these, it’s quite possible that most physicists don’t see this as an advance at all.

Update: Physics World has more about the delay here. It is supposed to be announced on Tuesday. The cause evidently is a budget gap caused by some planned contributions from non-member countries now not happening. The story doesn’t explain which non-member countries are involved or why their planned contributions are now not expected.

The great German artist Anselm Kiefer now has a show up in London at the White Cube Bermondsey gallery, with a review in the Guardian entitled Terrifying Odyssey Through a Cursed World. The review describes some of the works as follows:

Another room is given over to panoramic blasts of brown and black that map sweeping vistas of desolate fields. A road twines through a morass of mud and collaged sticks. Lines of fence poles vanish in the distance. These scenes are drawn in black on a vertiginous scale. Kiefer uses perspective, the Renaissance technique of showing the real world shrinking towards a single vanishing point, to define his landscapes – but the perspective view is a transparency on top of a muddy tumult of colour and texture, with real, 3D stuff stuck over that in turn. From the right distance, the picture of a landscape can be read clearly, like a painting by Van Gogh. Go closer and the picture dissolves in a mess of bulges and muck.

What’s the inspiration for these works (besides the Holocaust)?

These landscapes are entitled Superstrings, a reference to string theory, an influential idea in contemporary physics that seeks to unify quantum mechanics with Einstein’s relativity.

and the show is entitled Superstrings, Runes, The Norns, Gordian Knot. The gallery website explains:

White Cube is pleased to present an exhibition of new work by Anselm Kiefer. The exhibition brings together many of the interests that have characterised Kiefer’s work for decades, including mythology, astronomy and history. Located across the entire Bermondsey space, it features a large-scale installation and paintings that draw on the scientific concept known as string theory.

The Guardian review continues:

The main gallery at White Cube Bermondsey is already pretty bleak in its featureless emptiness. Kiefer makes it work for him by heightening the chill, turning the White Cube into a morgue for Europe. Snow-covered landscapes with none of the cheer of Bruegel stretch away to infinity. They are marked with sticks as black as gravestones and nets that catch at nothing. Kiefer’s science reading clearly hasn’t cheered him up. The curvy grids of space-time become horrible wire traps in a devastated nowhere. We might be on the no-man’s land of the Ukraine border. Anyway, this place has got death in its hard black furrows.

Another review, at City A.M. tells us more about Kiefer’s motivations:

This is where we come to string theory – the monolith of Kiefer’s new show. Though he admits that he doesn’t quite understand what string theory is, Kiefer professes complete fascination with the idea that there is a scientific equivalent to the allegorical Gordian Knot – an idea that he picked up after thirty years of subscribing to Spektrum, a German monthly science magazine…

String theory cannot be verified empirically. Rather, it is an attempt to provide an all-encompassing description of the universe. And that, says Kiefer, is just why it is beautiful. “I suppose it’s like painting,” he says. “You cannot prove if a painting is good or bad. That is the point of it – it is descriptive… and there’s something sublime in that.”

From the images available, the work does look quite amazing. Kiefer quite possibly has gotten to the very heart of superstring theory, seeing in it a dark, desolate and blasted mythology which “cannot be verified empirically.”

Now back from a trip to the West Coast, here are some accumulated things that may be of interest:

• One thing I didn’t do while there was attend the 2020 Breakthrough Prize symposium. For videos of three talks about supergravity, see here. At the time of the award I wrote here about why a \$3 million prize for a failed idea about particle theory was a bad idea. Listening to the talks, I think an even worse idea is telling the public that this is a great example of why they should trust science.
• For another dubious idea from the West Coast, in January the KITP is bringing high school teachers to Santa Barbara to teach them about Spacetime, Holography, and Entanglement. Most of the programs the KITP has run for teachers (see here) have been devoted to explaining important, solid science. Back in 2001 when they promoted string theory I thought that was a bad idea, this latest one isn’t much better. Again, when the credibility of science is under attack, why go to the public (or, in this case high school teachers) to promote a highly speculative research program? Is it really a good idea for high school teachers to be exposed to this kind of hype, presumably with the hope that they’ll somehow transmit it to their students?
• On the evergreen topic of bad multiverse science, Scott Alexander here defends multiverse speculation, responding to Jim Baggott’s article. He and the authors of the more than four hundred comments debate at length a red-herring issue. Alexander writes:
  

  My understanding of the multiverse debate is that it works the same way [as respectable paleontology]. Scientists observe the behavior of particles, and find that a multiverse explains that behavior more simply and elegantly than not-a-multiverse.

  Yes, if theorists had a simple, elegant multiverse theory with lots of explanatory power, you could get into interesting arguments about its testability and whether the idea was solid science or not. The problem is that no such multiverse theory exists. If you want to talk about the MWI multiverse, your problem is that solving the measurement theory problem by just saying “the multiverse did it” may be “simple” and “elegant”, but it’s also completely empty. If instead you want to talk about the cosmological multiverse, the problem is that you don’t have a theory at all (and the actual fragments of a theory you do have are complicated and ugly). For more about this, see my posting and article on Theorists Without a Theory.

For something more positive, while traveling I noticed two quite interesting articles which explain in a detailed technical way approaches to two of the great unsolved problems of our time, while carefully discussing why the approaches have not (yet?) worked, leaving the great problems unsolved.

• For mathematics and the Riemann Hypothesis, see Alain Connes and Caterina Consani’s article The Scaling Hamiltonian, about the attempt to understand the zeros of the Riemann zeta function in terms of the properties of a specific Hamiltonian operator, which in some sense is a generator of a group of scaling transformations.
• For physics and quantum gravity, see Donoghue’s A Critique of the Asymptotic Safety Program, which has a detailed discussion of the problems with making sense of both quadratic gravity Lagrangians and the idea of a non-trivial fixed point gravity theory theory. I was interested to see that he has a lot to say about the Lorentzian vs. Euclidean signature issue, something often ignored.

Finally, I recommend reading Elizabeth Landau’s interview at Quanta with astronomer Virginia Trimble and Trimble’s excellent advice to us all:

Pay attention. Someday, you’ll be the last one who remembers.

Update: Two more math-related items well worth a look:

• Dan Rockmore at the New Yorker on Where do ideas come from?
• A profile of Terry Tao at the Princeton Alumni Weekly.

Update: Yet another blog entry from Scott Alexander about the multiverse, with more hundreds of comments. What is it with the fascination for this empty argument?

At BBC Science Focus, more multiverse promotion from Sean Carroll. He does end though by getting to the real point (note that when theoretical physicists say a question is “hard to answer”, it means they have no idea how to answer it):

Many-Worlds is a lean and mean theory, but it’s possibly too lean and mean; there is very little structure to rely on, so questions like “Why do probabilities behave the way they do?” and “Why is classical mechanics such a good approximation to the world we see?” are hard to answer.

This is exactly the problem that those arguing over this at Slate Star Codex and elsewhere don’t seem to understand: saying “all is the Schrodinger equation” doesn’t tell you how to connect the theory to the world we observe. Adding in an ontology of multiple universes does nothing at all to solve this problem.