The hype campaign marches on, just three very recent examples:
Quantum Theory, Groups and Representations
Not Even Wrong: The Book
Subscribe to Blog via EmailJoin 439 other subscribers
- Various and Sundry 1
- Quantum Supremacy 28
Peter Woit, Giovanni Ronchi, Peter Woit, Mark Hillery, Mike, Felix C. [...]
- string theory lied to us and now science communication is hard 59
SRP, Marion Delgado, Frank Wilhoit, Peter Woit, JE, Peter Woit [...]
- Kontsevich on the Hodge and Tate conjectures 8
Topological twist, Peter Woit, Simon Pepin Lehalleur, jsm, Peter Woit, Mike [...]
- This Week's Hype 6
zzz, Peter Woit, Peter Woit, Ling-Nan Zou, Chris H, David Roberts
- Various and Sundry 1
- abc Conjecture (16)
- Book Reviews (121)
- BRST (13)
- Euclidean Twistor Unification (7)
- Experimental HEP News (153)
- Fake Physics (7)
- Favorite Old Posts (50)
- Film Reviews (15)
- Langlands (42)
- Multiverse Mania (163)
- Not Even Wrong: The Book (27)
- Obituaries (34)
- Quantum Mechanics (23)
- Quantum Theory: The Book (7)
- Strings 2XXX (22)
- Swampland (19)
- This Week's Hype (132)
- Uncategorized (1,261)
- Wormhole Publicity Stunts (12)
- Alex Youcis
- Alexandre Borovik
- Anton Hilado
- Cathy O'Neil
- Daniel Litt
- David Hansen
- David Mumford
- David Roberts
- Emmanuel Kowalski
- Harald Helfgott
- Jesse Johnson
- Johan deJong
- Lieven Le Bruyn
- Mathematics Without Apologies
- Noncommutative Geometry
- Pieter Belmans
- Qiaochu Yuan
- Secret Blogging Seminar
- Silicon Reckoner
- Terence Tao
- The n-Category Cafe
- Timothy Gowers
- Xena Project
- Alexey Petrov
- Angry Physicist
- Capitalist Imperialist Pig
- Chad Orzel
- Clifford Johnson
- Cormac O’Raifeartaigh
- Doug Natelson
- EPMG Blog
- Geoffrey Dixon
- Georg von Hippel
- Jacques Distler
- Jess Riedel
- Jim Baggott
- Joe Conlon
- John Horgan
- Lubos Motl
- Mark Goodsell
- Mark Hanman
- Mateus Araujo
- Matt Strassler
- Matt von Hippel
- Matthew Buckley
- Peter Orland
- Physics World
- Robert Helling
- Ross McKenzie
- Sabine Hossenfelder
- Scott Aaronson
- Sean Carroll
- Shaun Hotchkiss
- Stacy McGaugh
- Tommaso Dorigo
Some Web Pages
- Alain Connes
- Arthur Jaffe
- Barry Mazur
- Brian Conrad
- Brian Hall
- Cumrun Vafa
- Dan Freed
- Daniel Bump
- David Ben-Zvi
- David Nadler
- David Vogan
- Dennis Gaitsgory
- Eckhard Meinrenken
- Edward Frenkel
- Frank Wilczek
- Gerard ’t Hooft
- Greg Moore
- Hirosi Ooguri
- Ivan Fesenko
- Jacob Lurie
- John Baez
- José Figueroa-O'Farrill
- Klaas Landsman
- Laurent Fargues
- Laurent Lafforgue
- Nolan Wallach
- Peter Teichner
- Robert Langlands
- Vincent Lafforgue
The TF1 Snowmass Report is intriguing: thirty-eight authors (some of them of a very high profile) for what amounts to (mostly) fluff.
Resistance is futile, Peter. Drink the Kool-Aid.
Speaking of which Ethan Siegel, evidently incensed at Sabine’s dreary (but correct, IMO) assessment of theoretical particle physics, wrote a rebuttal in which he says, if I interpret him correctly, that in the absence (in his view) of any clear way forward, a series of guesses (sparticles, axions, etc.) is an entirely valid modus operandi. It rules out all sorts of nonsense. So stay the course, full speed ahead, be one with the Borg.
As usual, when I consider all these guesses, and all future guesses, I have to ask: Were your guess proven correct, how would you explain that its veracity was inevitable? If you cannot do that prior to submitting it to experimental verification, then you haven’t a chance – just a job, and the moral support of like-minded ruminants.
Although there is definitely hype around cosmic strings, extending the symmetry structure of the SM is a tried-and-true strategy to address extant (not-just-theoretical) problems. After all, the SM is a (successful) theory based on SU(3)xSU(2)xU(1) gauge symmetries. Even assuming just the SM gauge group and the Higgs mechanism, there are measurable/phenomenological consequences of this phase transition and the associated topological defects (that said, the SM EWSB version is highly constrained). If there are additional gauge (or global) symmetries which are spontaneously broken at energy scales reached in the early universe, shouldn’t the hype associated with topological defects formed in these phase transitions (like cosmic strings, which have measurable effects e.g. effects on structure, lensing, gravitational waves, etc.) be differentiated from the hype around string theory which still has different “difficulties” in terms of phenomenological predictions?
Yes hype about cosmic strings is hype about GUTs, not hype about string theory. But the hype in this article is the same kind of outrageous hype as the usual string theory hype, completely misrepresenting the situation. The author writes
“According to our best understanding of the early Universe, our cosmos should be riddled with cosmic strings. And yet not a single search has found any evidence for them.”
This implicitly claims that a GUT model, not the SM, is “our best understanding of the early Universe”. The situation is just like for monopoles: there’s supposed to be a huge problem of where the cosmic strings and monopoles are, when the SM predicts no cosmic strings and no monopoles. The theories that predict such objects are GUTs, for which we have lots of other falsifying evidence (eg no proton decay).
The math here is pretty simple: if the Higgs potential has a minimum on a manifold X, then you get cosmic strings when $\pi_1(X)$ is non-trivial, and monopoles when $\pi_2(X)$ is non-trivial. In the SM, $X=S^3$, so the two homotopy groups are trivial and you expect no cosmic strings, no monopoles, just as observed.
I think it’s kind of a remarkable document. The field is facing some difficult problems, and the point of the Snowmass effort is supposed to be to come up with a plan for the future. The authors act as if there is no problem to be faced, nothing wrong with what people have been doing in the past, the way forward is to just keep doing the same thing.
I’m onboard with Sabine that the best way forward is to identify problems with the structure/consistency of our best theory and find ways to address those problems. Completely mystified though why others don’t see that the analytic continuation of chiral spinor fields is such a problem, and addressing that should be the highest priority of all theorists.
More seriously, I’m also onboard with Ethan Siegel: the field is in a bad state, with nothing so far working (including Sabine’s favored approach), so there’s some justification for pursuing lots of different approaches, even ones that usually aren’t very promising ways to get successful new ideas. What’s important though is that such unpromising ideas at least are new, not what we have now which is a field that has institutionalized a research effort organized around a set of old ideas that failed long ago (see the last item).
I disagree with the idea of just trying anything (new) in the absence of guiding empirical evidence. I’m afraid that only a couple of things/strategies are worth pursuing, but will not be popular, since it amounts to basically doing nothing or very little (in terms of new, wild ideas… you know, the stuff that people like to hear about) and just wait until new experimental evidence finally contradicts our current best theories. But how do we reach to that desired point? The history of physics is illuminating.
Newton proposed his laws of mechanics in the late 1600s. As we know, they were a monumental success. And when it seemed they were failing, it was actually because of a planet that was not yet discovered. Only the small deviation in the prediction of the perihelion of Mercury resisted a newtonian explanation. The reign of newtonian mechanics lasted for a bit more than two centuries!
Now, did physicists of the time jumped all to speculate on wild theories that would replace newton during those centuries, like today’s physicists are doing with the SM? Maybe some did, but, regarding the rest, it seems to me their attitude was smarter. They either worked on applications of the theory (mainly in astronomy, to great success) or to rewrite/study its mathematical foundations (Lagrange, Hamilton, etc.) And I think that, in the part of the foundations of physics where no evident contradictions exist, that should be the adopted strategy today.
On the other hand, most other physicists were working on new and exciting phenomena, like heat and electricity and magnetism. This culminated in the theories of thermodynamics, statistical mechanics, and Maxwell’s EM. And the actual hard challenges to newton came from the last two (first with Planck, then Einstein). That’s because they exported the newtonian theory to realms for which it wasn’t designed to deal with. So, that’s where the new evidence will come from. So, the second strategy is to concentrate in realms not covered yet by our current theories. That seems a bit difficult, considering how amazingly encompassing they are, but there are still some holes here and there (e.g., quantum superpositions of ever bigger massive bodies, gravity mediated entanglement of similar bodies, and so on). The mathematics of both GR and the SM applied to other systems. More cross fertilization with other fields of science, etc.
So, I think there’s a clear path forward from the current situation, supported by history, but it’s not an easy one, it will not give you the possibility of activating the sausage machine of paper production and collect grants money with a bulldozer. Very little space for media super stars from the field. Instead, it requires a very careful and selective process regarding which problems to pick, the advance will be slow, and the general attitude should be more on the conservative side until news of something genuinely new finally appear.
You said: “Completely mystified though why others don’t see that the analytic continuation of chiral spinor fields is such a problem, and addressing that should be the highest priority of all theorists.”
Can you perhaps elaborate? Has it been discussed in a previous blog post? Or if not, in a paper somewhere? It sounds like the kind of thing I’d like to do.
@Alex: Great comment, I agree almost completely. Unfortunately, after almost ten (post-PhD) years in academia I have come to the conclusion that the only way (for me, someone who is just a decent physicist but not some genius) to proceed with this sort of serious, hard-work research on the foundations of the established theories is not possible within academia. So I am currently working on a strategy to leave the academic world and hope to find a way that allows me to do this sort of theoretical physics as a hobby.
> a series of guesses (sparticles, axions, etc.) is an entirely valid modus operandi. It rules out all sorts of nonsense.
The problem with this, of course, is that there is an infinite amount of possible nonsense. The more you rule out, the more nonsense you subsequently generate.
I’m currently trying to finish writing something much more explicit about the analytic continuation issue.
And another question very few people seem to be thinking about:
“Why is the Standard Model one of the very few Lagrangians that is compatible with Alain Connes’ non-commutative geometry constructions?”
The answer is probably that most physicists don’t know anything about either analytic continuations of chiral spinor fields or non-commutative geometry, and their colleagues tell them that these aren’t the answer, so why should they bother learning about them (something that is quite hard; it’s much easier to continue thinking about stuff they know about).
But if one of them is the answer …
I can only speak for myself, but as one of these ignorant physicists refusing to learn more about potentially promising – but hard – mathematical concepts such as non-commutative geometry or twistors, my reason is not that I don’t believe that one of them could be the answer.
Rather I believe that this is not the most promising way to do theoretical physics.
It’s worth noting that both for the discovery of GR and QM the first 20+ years were about fiddling around with the “old” mathematical tools, e.g. statistical mechanics or Maxwell’s theory. Only after there was a certain amount of understanding of the new physics (described in “old” terms) did people begin to discover its relationship to new mathematics.
Riemann was already dead for a quarter century before Einstein started thinking about space and time, nonetheless, the thought of differential geometry as a tool came much later – after already having discovered the principles of special relativity, and arguably Einstein knowing more differential geometry would not have changed much about the history of this discovery.
For the formulation of QM, the discovery of many mathematical tools was historically happening around the same time (e.g. Hilbert, Weyl), and yet, the first half century of the new theory was dominated by phaenomenological models that were strongly based on the established mathematics of classical physics.
Hence, I am very happy if mathematical physicists are working on ideas like Connes’ or Penrose’s, but I would be quite surprised if the next big discovery comes from one of these approaches. Most likely, I expect it to come from some experiment – maybe one of the macroscopic quantum systems in the lab type things – but if it comes from theory, then my bet would be on a deeper understanding of issues and inconsistencies with QFT in curved spacetime vs. GR or in the heuristic ideas around studying quantum mechanical systems with gravity. In my opinion, there are plenty of situations that are only very poorly understood in our existing theories (talking about fundamental issues, not complex systems etc.).
I’ve just deleted a large number of incoming comments that had nothing to do with the posting. Please keep in mind that you create a problem by posting extensive comments that may be interesting, but have nothing to do with the posting. When I allow such comments, this quickly leads to a lot more of them as everyone decides this is a good place to try and get a discussion going of whatever interests them. This then puts me in the position of having to spend a lot of time I don’t have trying to sensibly moderate a discussion that has gone off into areas I’m not interested in.
The Snowmass report contains this paragraph, with intriguing last sentence:
“ Grand unification seems very natural in the heterotic picture (where the MSSM gauge group naturally arises as a subgroup of the larger ten-dimensional gauge group). It is less natural in intersecting brane models, and may be achievable in F-theory models with 7- branes. In all cases, details of making the unification work are subtle. (Current data is also not conclusive on the question of whether achieving unification should be a central goal or not.)”
I think this can be translated as “nearly forty years of intensive research has shown string theory is compatible with just about anything, specifically either GUT theories or non-GUT theories”. Since this is a document aimed at guiding funding decisions, one would think the implications of this would be obvious…
Do people know if any sociologist or historian has done a study of string theory as a research program? What I have in mind is a study that tries to provide a social-historical explanation of it/framework for understanding how it persists, starting from the assumption that string theory’s prominence has not been justifiable on rational grounds for a very long time? This is something sociologists do well–explain why a social institution that doesn’t fulfill the ostensive functions legitimating its existence (in this case, producing physics knowledge), nevertheless persists as if it actually is fulfilling those functions, despite convincing demonstrations that it isn’t. Given the prestige and cultural influence of string theory, and the importance of the questions it purports to help answer, it seems like this is something worth understanding from a sociological and historical point of view.
I don’t know about sociologists, but various historians (Sophie Ritson is one example) and philosophers of science have written about the controversy over string theory. But, as long as most of the leading figures in the field continue to argue that string theory is not a failure, I don’t think academics from other fields are going to do work starting from the assumption that string theory is a failure.
At some time in the future, such academics may very well be looking back at what happened in fundamental theory starting in 1984 and analyzing it more from the point of view of what this can tell us about how science can go seriously wrong.
An article in the New York Times is now hyping ER=EPR, among other things.
I did see that, and as usual with this kind of hype, can’t even figure out what the hypsters are referring to. Supposedly it’s now all about wormholes, but what’s the theory governing these wormholes?
In the NYT, Overbye does refer to two identifiable papers as the crucial advances. The first is
where the authors have an actual theory (JT gravity coupled to conformal matter), but it’s a 1+1 dimensional QFT. 1+1 d quantum gravity is very much a toy model, with no physical degrees of freedom. Maybe some phenomenon in this toy model corresponds to wormholes and resolves the bh information paradox, but if so, the bh information paradox is getting resolved in a way that doesn’t have anything to do with the problems of 3+1 d physical quantum gravity theories.
The second is
There the author just invokes “AdS/CFT” as the theory he is using, but doesn’t use anything about a specific theory of the quantization of the 4d gravitational degrees of freedom. His conclusion is that understanding the bh information paradox can be done semi-classically:
“In this paper, we have argued that the key expected features of unitary black hole evaporation in AdS/CFT can be derived from the bulk semiclassical description of an evaporating black hole, so long as we assume entanglement wedge reconstruction.”
In both papers, the conclusion seems to be that you don’t need a well-defined 4d quantum gravity theory, that you can just work semi-classically or in 1+1d. Hard for me to follow the details here, but the original goal of having a well-defined theory of quantum gravity seems to have been replaced by something else.
I’m getting the impression that in the same way Susskind successfully argued for the abandonment of the idea of having a well-defined theory of particle physics (what we see is a random artifact of the unknowable “landscape”), what’s going on here is the abandonment of having a well-defined theory of gravity (with gravity only understandable semi-classically as an emergent approximation to something unknowable).
Peter, there surely are good arguments that a unified theory of relativistic quantum gravity has no equation of motion. The minimum length yields a minimum measurement error, and thus no way to define a precise equation. Also, for the same reason, single degrees of freedom are unobservable (in a way that recalls the unobservability of single gravitons). As a result, there is no way to define an equation of motion (or a Lagrangian) for the underlying degrees of freedom. Thus, the unified theory of relativistic quantum gravity can only be statistical, emerging from some underlying degrees of freedom.
But there still is the standard model. It arises, like general relativity, from these underlying degrees of freedom, with the precision that we know. In other terms, the underlying degrees of freedom determine the fundamental constants uniquely. Thus, there is no landscape, but a uniquely defined world with quantum theory and general relativity.
I’m not asking for observable degrees of freedom, an equation of motion or a Lagrangian. If you want to tell me that
“the unified theory of relativistic quantum gravity can only be statistical, emerging from some underlying degrees of freedom” and that “the underlying degrees of freedom determine the fundamental constants uniquely” you have to tell me what those underlying degrees of freedom are, what laws govern them, show that gravity emerges and explain how to compute the fundamental constants.
Maybe you think you have a way to do this (please don’t use my blog to promote it), but I don’t see any evidence that Susskind et al have a program to find such a thing. Rather their program seems to be to argue that string theory is such a thing, but you can never know what it really is or how it works.
Coincidentally I am starting to read about the works of some schools of Sociology of Science/Scientific knowledge: the Edinburgh school and its Strong Program, and the Relativism of the Bath School being the more controversial ones. In a bibliographical search on theses matters I found an essay by Weinberg (1995) Night Thoughts of a Quantum Physicist. In it it mentions his view on the Strong Program and related schools and also this passage: “String theory is mathematically very difficult. I think it’s a sign of the intrinsic health of physics that even though most physicists of the older generation don’t learn string theory and can’t read papers in it, young string theorists continue to get tenure at leading American universities. Our field is not (as some imagine it to be) hegemonic, dominated by an old guard of fuddy-duddies. It is very much alive to new possibilities” …the irony
Peter, what I wanted to point out is that the statement “the unified theory of relativistic quantum gravity can only be statistical, emerging from some underlying degrees of freedom” can be made *independently* of whether these degrees of freedom are known or not. Nature *must be* like that – because of the measurement uncertainties in nature at the Planck scale and because of the impossibility to reach the Planck scale. There is no alternative.
The same holds for “the underlying degrees of freedom determine the fundamental constants uniquely”. Nature must be like that – there is no alternative.
Finding theose degree of freedom is a different step. It is of course necessary as well.
I think there’s something wrong with that reasoning. The angular momentum of a single particle is unobservable (except statistically), but we have perfectly good equations of motion for angular momentum.
Quantities that are only statistically observable but have exact equations of motion are everywhere in quantum mechanics. If Schrödinger and Heisenberg had followed reasoning like yours, we would not have the theory of quantum mechanics that we do today.
quantum theory and relativitsic quantum gravity differ.
Quantum theory and experiments show that we are able to say whether a system contains just one spin or just one particle. We can isolate a single particle and follow its motion.
In contrast, we are not able to say (or even imagine) that a systen contains just one Planck-scale degree of freedom (those which make up black hole entropy) or just one graviton. We cannot isolate one of those degrees of freedom, because the minimum length cannot be realized in any experimental set-up. (We are far away by many orders of magnitude.) We cannot follow the motion of such as single degree of freedom, because the minimum length cannot be realized, not even approximately.
In relativistic quantum gravity, only systems with huge numbers of degrees of freedom can be described: black holes, curved space, etc – in contrast to quantum theory, which allows describing single particles.
So, in relativistic quantum gravity, there can never be an equation of motion for a single degree of freedom, in contrast to quantum theory. The constituents of space are not observable one by one, but only in huge numbers.
“… , what’s going on here is the abandonment of having a well-defined theory of gravity (with gravity only understandable semi-classically as an emergent approximation to something unknowable).”
Peter, this is not at all what is happening in the recent computations of the Page curve using the semi-classical description plus HRT formula. It is like saying that because you compute the entropy of a many-body system using thermodynamics and equations of state, you have given up on there being an in principle microscopic description of the entropy. The AdS/CFT community does not think semi-classical GR is a low-energy EFT of something unknowable, because we know exactly what the microscopic degrees of freedom are: they are the degrees of freedom of a conformal field theory.
This has gone way off topic, enough.
You know exactly what the microscopic degrees of freedom are? Great, I couldn’t get that from reading any of these papers. What are they exactly, and what exactly is the conformal field theory?
It depends on what AdS/CFT example you are looking at, so for different bulk dimensions and matter fields you will have different CFTs. But compared to most things in QFT, CFTs are extremely well understood. The prototypical example is N=4 SYM. In this case semiclassical (super)gravity in an asymptotically AdS_5 x S^5 space is an EFT for N=4 SYM in certain limits. The microscopic degrees of freedom of quantum gravity are the degrees of freedom of N=4 SYM. Sure, these don’t live in spacetime, but that’s the whole point of emergent gravity. Of course, I cannot tell you in full generality what the emergent gravitational field in the bulk are constructed from in N=4 (but in some cases we can, ref HKLL and the extrapolate dictionary). But this is not a fundamental problem, any more than it is a problem that I don’t know how to write the pion field of the chiral lagrangian in terms of QCD operators. There is some (extremely hard to compute in practice) field redefintions of N=4 SYM in terms of new variables, whose dynamics become Einstein gravity at large N and large ‘t Hooft coupling.
Let me add that we know the exact dual for JT gravity. It is the matrix model described in 1903.11115. So the matrices described there are exactly the microscopic degrees of freedom of JT gravity. As for Penington’s case, his argument is sufficiently general that you don’t need to know the exact matter spectrum of the bulk theory (or the dual CFT).
In the toy JT case I see a well-defined theory (JT + conformal matter in 1+1), even independently of the dual.
When asked for a specific theory that’s going to be quantum gravity in 3+1d spacetime (dS), you’re telling me the standard story about a supergravity limit of superstring theory on AdS x S^5 being dual to N=4 SYM on the boundary 4-sphere. I know that story well, and I’ve seen 25 years of people trying to get a 4d physical quantum gravity theory out of it, which somehow never seems to lead to a specific theory you can write down which does this. What is this theory?
This subject has clearly evolved recently and there are new aspects I don’t understand (ER=EPR and the wormholes, for instance…). If there is a specific conformal theory you can write down which is supposed to have as dual 4d physical gravity, it would be helpful to know exactly what it is, and how this duality is supposed to work.
But now you are changing the complaint. You are of course 100% right that AdS/CFT does not give you a dual to 4-dimensional quantum gravity in deSitter space, but that was never the complaint I was responding to. You seemed to say that the AdS/CFT community had given up on the microscopic degrees of freedom, which I was disagreeing with. You were saying “but the original goal of having a well-defined theory of quantum gravity seems to have been replaced by something else.”. The point I was making is that it was recently realized that the Page curve can be computed without knowing the microscopic degrees of freedom. Just because you decide to carry out this computation doesn’t mean you have given up on microscopic degrees of freedom. Furthermore, since the Page curve computation is completely agnostic to the number of dimensions, so it is pretty reasonable to be hopeful that this solution works more generally.
For a research community to “give up” on a goal doesn’t require that members admit this publicly or even admit it to themselves. It’s a fact that the string theory community has given up on particle physics, in the sense that leading figures have voted with their feet and no longer work on this, whatever they may say to the public, to their colleagues, or to themselves about the question of whether there’s any hope get particle physics out of string theory.
The fundamental physics goal of “string theory”, whatever it is these days, has moved to “find quantum gravity”, but this goal is being pursued in a way that I have trouble understanding and that increasingly appears to me to be describable as “giving up”. For many years the main motivation for this program was given as the occurrence of a spin-two state in the spectrum of a weakly-coupled 10d superstring, nowadays instead it seems to be “emergence” of an effective gravity theory from something else. My problem has been that I can’t find a specific explanation of what the “something else” is, and how this emergence is supposed to work. Your pointing to 4d N=4 SYM doesn’t really help.
The NYT article points me to Susskind’s
“Dear Qubitzers, GR=QM” (https://arxiv.org/abs/1708.03040)
which, honestly, makes zero sense to me. The NYT headline is “Black Holes May Hide a Mind-Bending Secret About Our Universe”, but the story it tells about a semi-classical resolution of the BH information paradox seems to imply that studying this question about black holes tells us only about the emergent semi-classical limit, nothing about the fundamental theory. So, the big success here is showing that the “mind-bending secret” about the nature of the fundamental theory is hidden from us rather than revealed to us by black holes.
Having seen what happened with particle physics and string theory, I don’t think it’s unreasonable to interpret what is going on here as people abandoning the goal of search for a fundamental theory by voting with their feet to move to working on problems that can’t address this goal (or, when they do, such as in Susskind’s preprint, do so in a way that can’t be taken seriously).
Peter, I agree that many people (although not all) have given up on getting particle physics from string theory, so let me respond only to the part about emergent gravity.
“My problem has been that I can’t find a specific explanation of what the “something else” is, and how this emergence is supposed to work. Your pointing to 4d N=4 SYM doesn’t really help.”
I see two complaints here: (1) what is the “something else”, and (2) how does emergence actually work in terms of the microscopic degrees of freedom.
(1) I am still confused by this complaint. For quantum gravity in AdS spaces, we are always claiming that the “something else” is some conformal field theory. Sure, we don’t always know what the relevant CFT is, but that says that we haven’t mapped out the space of all AdS/CFT dualities. But we do have precise examples, like N=4 SYM. I am not sure why this example doesn’t help. Sure, its dual is not four dimensional gravity, but it is an example of a quantum gravity where we have know exactly what the microscopic degrees of freedom are. If you want four dimensional gravity, you can do a KK reduction of AdS_4 x S^7, and the dual is the ABJM conformal field theory.
(2) As for the lack of explanation of “how this emergence is supposed to work”: this is the difficult question of bulk reconstruction and “entanglement wedge reconstruction”, which is a topic of active research. It’s hard, because it in some ways is morally similar to the problem of writing the pion field of the chiral lagrangian in terms of the QCD fields. But we do actually know quite a bit about this problem. We know how the gravitational field and matter fields near the asymptotic region of AdS can be written in terms of CFT variables. We know how to write bulk operators deeper the bulk in terms of CFT variables, provided we are not behind a horizon. We haven’t solved this problem, but progress has been continuously made over the last 15 years. Some milestones are: 0606141, 0907.0151, 1601.05416,1612.0039, 1704.05464, 1704.05839.
To detail one example, 1612.00391 gives an algorithm to reconstruct the bulk conformal structure (outside the black hole horizon) from Lorentzian CFT correlators.
Thanks, that’s very helpful. My previous attempts to understand what has been going on here were not helped by mystifying claims for radical reformulations of fundamental theory (e.g. Susskind’s GR=QM, Arkani-Hamed’s “space-time is doomed, to be replaced by amplituhedra”, etc.). If you’re telling me that the basic idea is the same as it always has been post-1997 (try to understand 4d gravity in terms of holography and a dual 3d CFT), I can understand that (as well as see the well-known problems with the idea). It appears the recent advance is to show a way to resolve the bh problem within this context. I’ll be curious to see what the future holds. Right now it looks to me like this just leads to a quite complicated way of quantizing gravity, one still far from working in the physical case, with no hope of testability (either via predictions about gravity or via connection to particle physics).
The trouble is that we don’t live in AdS but supposedly in dS so the program of emergence of gravity as in AdS/CFT is just left in AdS. And I don’t quite see how you can get a similar duality in dS because there are no asymptotic observables. One can wish for a similar emergence in 4d dS but we need something better and more precise than that.
Of course, happy to elaborate. And I agree that it is certainly complicated. But no one ever said quantum gravity would be easy. As for no hope of testability, this is true, since these quantum gravity theories never were claimed to describe our universe. On the other hand, people are building strongly coupled many-body systems in the lab these days, so if we obtain the ability to simulate CFTs in the lab, we might be able to probe simulated quantum gravity theories experimentally. Sure, it is not our universe, but there are many really interesting problems that we can hope is shared between different quantum gravities, like: is there such a thing as approximate geometry when quantum effects are important? Can topology change happen dynamically? What happens to a black hole at the very end of evaporation? Can a classical singularity be resolved by quantum effects? What are even natural observables in quantum gravity? How does approximate locality emerge? How is information about a black hole interior encoded in the Hawking radiation? My hope is that new qualitative lessons about these questions and quantum gravity can be learned from AdS/CFT once we can actually simulate CFTs. Of course, we need to understand deSitter at some point as well, but a better understanding of the former questions might get us closer to deSitter. AdS/CFT is quantum gravity in a box, and while the box is unphysical, it seems unlikely to me that the lessons learnt with the box are completely irrelevant to our universe.
Funny thing is that I was just this afternoon at a talk here at Columbia by John Preskill. Fine general talk, pretty much the same as
He ended with a slide claiming
“Deep insights into the quantum structure of spacetime will arise from laboratory experiments studying highly entangled quantum systems”
This kind of “we’re going to study quantum gravity in the lab” claim I think is dangerously misleading. There’s a huge difference between testing a proposed fundamental theory of physical space-time and testing your modeling of a strongly-coupled condensed matter system using geometry of some space that has nothing at all to do with physical space-time.
Again, my problem is that we’ve already seen the destruction of fundamental particle theory by string theorists declaring that they’ve figured it out, the landscape’s the answer, too bad it’s untestable. We may be on our way to seeing the destruction of the subject of quantum gravity as “string theorists” declare that the answer has to be an untestable holographic dual (only game in town!), and that they’ve solved the testability problem by testing holographic duals in the lab, so that’s the end of it.
“that has nothing at all to do with physical space-time. ”
This is where we disagree, but I suspect we won’t be able to resolve our differences here. Ref. my comment about quantum gravity in a box.
I think the problem is that
“… AdS/CFT is quantum gravity in a box, and while the box is unphysical, it seems unlikely to me that the lessons learnt with the box are completely irrelevant to our universe.”
contains two hypotheses which holographers have been claiming since the inception of these conjectures decades ago but for which, despite all the effort and money invested, not a single piece of concrete evidence (of any type, not even theoretical, purely mathematical, let alone empirical) has been provided for us non-believers. These are “… AdS/CFT is quantum gravity in a box…”: you cannot know that since we don’t have the real QG deal. Maxwell’s eqs in a box are Maxwell’s eqs in a box because we start with them and only then put those boundary conditions and then it follows trivially that the results are indeed Maxwell’s eqs in a box. QG is such a non-trivial matter that I wouldn’t run to make such hasty general claims. Second, “… it seems unlikely to me that the lessons learnt with the box are completely irrelevant to our universe.” Again, just a gut feeling, basically a variation of the same previous melody. You guys speak as if you already had the true physical QG theory. You don’t. You have a specific theory which is highly unphysical for a number of well documented reasons. Besides, even its interpretation as a QG, even in that realm, can be questioned, as has been done.
We have been listening to this stuff for decades. Piles of conjectures over gut feelings over even more conjectures. And then you get Susskind or even Preskill with these overhyped-elevated-to-the landscape-power claims. Really? We have to buy this and not even compain? That’s too much.
In the absence of empirical data, you have two options: give up, or pursue a theory based on its theoretical merits. Assuming you are the same Alex as posting earlier in this thread, you seem to advocate the former. I am not in favor of this approach, as I do not see how we are ever going to come up with ideas for things to measure if we don’t even explore the various possible theories of quantum gravity. Sure, we can sit back and wait until inconsistencies with GR show up, but it seems much more efficient in the long run to actively look for them. I am not saying that AdS/CFT currently tells us what to measure, but understanding the quantum gravity theories that we know and that resolves existing puzzles (like the BH info paradox in AdS) seems like a good start. Let me add that I am all for people investigating alternative approaches, like what Peter is doing.
Next, I am going to disagree with your claim that we don’t know that holography describes quantum gravity in a box. The CFT is a UV complete quantum theory whose dynamics in certain limits is semi-classical Einstein gravity. If that doesn’t classify as a quantum theory of gravity, then I am not sure anything does. How else do you propose we define a quantum theory of gravity? Also: sure AdS/CFT itself is not proven, but it has so many chances of being disproved. Every time we compute something that can be computed on both sides of the duality, the two results are consistent. That sounds like theoretical/mathematical evidence to me.
Finally, AdS/CFT can be justified according to the principles you were championing earlier. “They either worked on applications of the theory (mainly in astronomy, to great success) or to rewrite/study its mathematical foundations”. It is clear to everyone that we need to understand quantum field theory better, and that this has obvious relevance to the parts of the world currently empirically accessible. AdS/CFT has given us a lot of new insight into conformal field theories at strong coupling (and their real time dynamics). Since QFTs typically are conformal in the UV and IR, and since phase transitions are described by CFTs, its clearly very interesting to understand them better. What other approaches do we have to describe real time dynamics of strongly coupled CFTs? As another example, AdS/CFT gives an explanation for the very low empirically observed value of the ration between the shear-viscosity and entropy density in quark gluon plasma. It also helps us understand the emergence of hydrodynamics in strongly coupled field theory, through the fluid-gravity correspondence.
Sorry but for those – like me – who do not know much about the field, the discussion is not so clear: is the “CFT whose dynamics in certain limits gives Einstein gravity” (advertised by Dumans) the same as the ” specific theory which is highly unphysical for a number of well documented reasons” (as referred by Alex).
If so, the main of these well documented reasons is that we live on dS rather than on AdS, right ? Are there other “big” reasons ?
A side question related to an earlier post of Peter: what is the S^3 on which the Higgs field lives in the standard model ? The spatial section of the FRWL model?
This is way off-topic, but the SM Higgs is valued in $\mathbf C^2=\mathbf R^4$. The potential depends on the norm-squared, minimum is at a fixed non-zero value of the norm-squared, so at a minimum the field takes values on an $S^3\subset \mathbf R^4$.
Ever since the dawn of AdS/CFT, it has been tempting for many to claim that this solves the problem of quantum gravity: QG is just whatever the dual is to your CFT. If you try and ask how this is supposed to work for physical 4d gravity, there are lots of problems. The most obvious one is you get AdS not dS. Another is that AdS/CFT is best understood when the CFT is N=4 SYM in 4d, so your space-time is the wrong dimension (5). Dumans Fladskote says “If you want four dimensional gravity, you can do a KK reduction of AdS_4 x S^7, and the dual is the ABJM conformal field theory. “, but this has its own problems. As this shows, in this game you’re really dealing with an 11d space-time, and have all the problem of not really knowing what the 11d theory actually is, having to pick a specific unmotivated background (for a supposedly background independent theory…), etc.
From the beginning I would often ask string theorists about this: what is your 4d theory, and why are you not doing physically relevant calculations in it? I’ve never gotten much of an answer, so never took time to look into the details. Perhaps someone more knowledgeable can point to a source where people use a 4d AdS/CFT to do QG and which discusses exactly how this works, what is known, what is unknown, what the problems are.
The problem with the statement
is that we know from the AMPS thought experiment that any satisfactory theory of quantum gravity has to violate one of unitarity, locality, or the equivalence principle. Any laboratory experiment studying highly entangled quantum systems has to obey all three of these principles. So how are we going to learn anything about quantum gravity from experiments that obey these conventional physical principles?
I suppose it’s possible, but certainly none of the experiments proposed to date are going to tell us anything interesting about quantum gravity.
Preskill refers to AdS/CFT, where there is manifest unitarity, but only emergent locality in the extra dimensions (while locality is exact in the boundary directions). The resolution of AMPS in AdS/CFT is through (spacetime) wormholes, which violate locality. The fuzzy dream is that somehow you’ll be able to interpret quantum simulations in terms of gravity in AdS: both through conventional gravitational phenomena, but also more speculative ones, like wormholes and other things that we don’t yet know about.
It is a challenge in AdS/CFT to make the size of the internal manifold M (in AdS_4 x M) small (compared to the AdS radius). The best known construction in string theory is by DGKT [arXiv:hep-th/0505160], but the dual CFT is not understood, and according to my estimates, roughly half of the community believes that something is wrong with DGKT and one can’t make M small. The good thing is that there is steady progress in building CFT tools (bootstrap and the like) to answer this question on a 5-year time scale.
Peter: sorry for having insisted on the off topic question, but that was to understand better your comment 🙂 Thanks for the answer !
Y’all are quibbling over old hat. String theory has moved on, and is currently viewed mostly as a tool for analysing strongly coupled field theories, via S-matrix bootstrap methods coming from strings (Simmons-Duffin, Zhiboedov, Sinha), or the large-N methods applied to Hilbert spaces of field theories (Sonner, Verlinde, Jafferis, Kolchmeyer) or the spectral functions that arise from analysing strongly interacting theories (Sachdev and collaborators and others), something that was the offshoot of the JT-gravity/SYK-model duality as pioneered by SSS and Witten and Maldacena and others. So please get yourself up to speed on these, and then we could have a conversation going.
You seem to be agreeing with me that string theory first gave up on explaining anything about particle physics, and moved on to a theory of only quantum gravity, now has given up on that and moved on.
If string theory is now only a “tool for analysing strongly coupled field theories”, irrelevant to particle physics or gravity, those of us who aren’t condensed matter physicists have good reason to no longer pay attention. After decades of hype about particle physics (ending in failure), more decades of hype about quantum gravity (ending in failure), hard to find a reason to pay attention to what looks like hype about a tool for dealing with certain strongly coupled QFTs.
String theory has moved on ? Here is the abstract of the seminar that Vafa will give this week at the physics department of Genoa’s university:
“String theory landscape of vacua point to new consistency conditions that a quantum gravitational system must satisfy. There are only a small number of quantum field theories that satisfy these conditions and all the rest belong to the `Swampland’ which cannot be consistently coupled to gravity. In this talk I review some of these conditions and their implications for cosmology and particle physics”.
And why is Vafa in Genoa ? Because he is one of the exceptional guest of the Science Festival, which is going on all around the city for 10 days (very nice festival by the way). He will give a “general audience talk” with abstract (google translated from italian):
“There are many puzzling things in the Universe. What if we could better understand the fundamental laws of the world through puzzles? The puzzles can reveal the greatest mysteries, revealing the simplicity of the physical laws and the elegance of their mathematics. Here is an original insight into the evolution of scientific thought, from the first great questions of antiquity to the brilliant theoretical synthesis of recent years on the laws of nature. Accompanied on this journey through time and space by one of the world’s leading experts in String Theory, you will thus appreciate the key ideas of modern understanding of matter – for example, symmetry and its break, but also, precisely the Theory of Strings and research on the nature of black holes – thanks to a series of exciting questions, that will capture you with all the charm of great scientific adventures.”
@Somdatta: do you seriously believe he will advertise String Theory to a general audience as a tool for analysing strong coupling field theories ?