Saturday, February 08, 2014

Juan Maldacena's NYU colloquium on QM, GR

Some hot promising yet divisive conjectures mentioned

Readers with 66 spare minutes are recommended to watch a colloquium that Juan Maldacena, an inaugural Milner Prize winner and the father of not only the AdS/CFT correspondence, gave at the New York University a week ago:

He's introduced and the sound quality improves immediately afterwards. In fact, it seems like the participants at NYU didn't enjoy the same quality as the YouTube viewers like you.

After some introduction to general relativity and quantum mechanics, Juan offered some jokes about the same criticism applied against the big bang theory as well as string theory.

He immediately jumps on the journey towards the quantum gravity topics – quantum field theory in curved spaces is an intermediate step. After five minutes or so, he already talks about the black hole temperature, including some fun numbers. A continental-mass sub-micron black hole will emit light at visible frequencies like a flashlight; so Juan's lightning technology is rather inefficient, not ready to replace LED lamps. ;-)

Someone interactively asks whether you may feel the temperature in the flat space, too. You can.

Inflation. Quantum fluctuations as WMAP non-uniformities. These inhomogeneities are scale-invariant. A clever way for Juan to discuss the importance of scale invariance at many places of condensed matter, statistical physics, and particle physics. The scale invariance is the important subgroup of the conformal invariance – one behind string theory's world sheets, AdS/CFT, and so on.

He asks whether the wave function of the Universe (as a function of the background metrics) may be equal to the partition sum of fields defined on the geometry. Strominger, Witten, Anninos, Hartman, Juan, and others are said to vote Yes while Dyson, Kleban, Susskind, Polyakov etc. vote No.

Juan refocuses on the hyperbolic or de Sitter space and explains the maximally symmetric spaces.

This picture of the hyperbolic space by Escher is mentioned often. If you don't know how it's called, what do you see in the picture? Answer before you continue reading.

The fun fact is that the picture contains two kinds of objects – black demons and white angels. The optimists only see the angels, the pessimists only see the demons. Where did you belong? ;-) Were you also stunned that you were overlooking 1/2 of this Escher world?

Around 18:00, Juan turns to AdS/CFT, offers various names of the duality, the brane argument that originally supported the duality, and other things. By proposing the holographic duality, Juan de facto reconciled the gauge picture of the 3-branes by Polchinski; and the curved-black-brane picture of the object by Polchinski and Horowitz.

He sketches the behavior of scale-invariant theories and praises Georgi's "unparticle" terminology (something like aparticle with its position and size depending on the world line parameter; it may be automatically translated to a genuine particle in a higher-dimensional AdS space).

Has quantum gravity become cheap because every scale-invariant theory produces one via AdS/CFT? Juan answers just like I do (e.g. when a frustrated Steve Shenker once raised this "cheapness" bitter remark): Yes and No; it is "some" quantum gravity theory but it may be very strongly coupled, far from having a usable flat-space limit. A large number of fields is pretty much enough to make the bulk theory weakly coupled. You need \(10^{12}\) or \(10^{120}\) fields for the curvatures of the same order as in the early Universe or today, respectively. No accident that the numbers are the inverse cosmological constants in the Planck units. However, the coupling has to be strong as well for the three-graviton vertex in the bulk to be Einstein-like.

Juan is asked what he means by the boundary; he has never said that the AdS had any (conformal) boundary LOL, so he tries to fix this pedagogic flaw.

\(\NNN=4\) super-chromodynamics, the most often studied example of the duality, deserves a special discussion. Integrability. Calculable dimensions. Practical applications: difficult strongly coupled CFT problems translated to simpler gravity problems. Thermal configurations become AdS black holes, entropy agrees on both sides. Dissipiation, viscosity calculable.

Entanglement on the boundary may be translated to geometrical shapes in the bulk. Ryu-Takayanagi and minimal surfaces. Israel's and Juan's entangled state for the eternal AdS black hole; thermal double. Eternal Schwarzschild i.e. ER bridge; wormhole interpretation. A very secret, fatal meeting inside the black hole. ER=EPR.

Someone asks a question how the bridge may be classical in GR as well as highly quantized. Well, it's the point of ER=EPR LOL. The connectedness of the two black holes is no feature of the dynamics, just a fact about their entanglement. Final theory not known; the connection between entanglement and geometrical connections is likely to matter.

55:00: conclusions and various interesting open questions. Questions from the audience. How can we build entangled black holes i.e. the ER bridge? Either by feeding two tiny black holes with nicely entangled pairs. Or by pair creation.


  1. Oh yes, this I will immediately watch Wednesday evening when back home :-)))

    Juan Maldacena is a very nice and fun speaker, I know it from his FFP talk.

    Thanks for posting this and explaining what he says!

    Darn, when reading too stringy TRF posts while being away, I always miss my Zweibach book too LOL :-D, which must feel quite lonely at present ...

  2. Thanks for this blog post entry.

  3. Hey Lubos, many problems in physics got solved when someone got rid of infinities right ? Why not try to get a model of a black hole with no singularity at the center? There is this model where the inside of a black hole is a star made of dark matter, it describes some of the observed phenomenons. Obviously I'm not an expert, but maybe you can reply and explain why there has to be this singularity. :) Greets

  4. Dear Jonny, one may prove that singularities form in classical GR when things like stars collapse, via "singularity theorems" (founded by Penrose and Hawking). To revert this conclusion, one would either has to reject local special relativity or introduce voodoo-like "action at a distance" designed to avoid singularities, or something else that looks similarly implausible.

    Quantum gravity only modifies the physics "Planck length" near the singularity, sort of, but it is enough to avoid infinities in principle. No one knows how do describe the singularity region precisely, anyway, but it's clear that the "strict infinity" doesn't exist in the quantum theory in any well-defined sense.

    Purely physically, the singularities are really harmless. Observers die near them, anyway. And most singularities are hidden behind event horizons (cosmic censorship conjecture which is known to be false to a large extent, but it's "often" a good intuition, anyway).

    A full theory of quantum gravity like the CFT description of AdS gravity using AdS/CFT is free of infinities that would render any predictions meaningless. But that doesn't mean that the classical prediction that singularities are formed is "complete denied". It probably can't be denied.

  5. Wow thank you for reply :)

    Have a nice day !

  6. Juan Maldacena careful extrapolations toward understanding the ads/cft on the boundary as a conformal field theory approach gives a interesting perspective about what is happening inside the blackhole.

    Honestly, I did not understand how people did not see this graduation?:) This geometrical propensity, as a mathematical basis?

    At the same time, I see Juan Maldaena as being very careful to explain this foundational approach as if to support the ongoing talk regarding the firewall?



  7. Thanks I will definitely watch this...

    FYI a new book about String theory history is coming:

    I'm planning to buy it but why it's so expensive?

    I don't get it...

  8. quizz answer: bat/angel duality in AdS space

  9. and a new conjecture: the existence of a bat/demon correspondence in the bulk

  10. CentralCharge15

    This is really ridiculous: Lumo and Qmechanics have answered this group theory question on Physics SE

    which does not prevent other high rep users from migrating it away...

    They blatantly give a damn about the judgement of experts. I have never seen a site disrespection people and their contributions like this before. On both, Math SE and MathOverflow they respect people who are willing to offer their time and efforts by contributing and explicitely say they dont migrate questions away against the will of the OP.

    Maybe it would be a good idea for people seriously interested in physics at a technical level, to shift their activity and business to the mathematical sites indeed, as Physics SE has become rather hostile to them. Basic up to good decent level questions could go to Math SE and high enough level stuff to MathOverflow ...

    On MathOverflow, there are good theoretical physicists too indeed, whereas on Math SE too physicsy questions are rather ignored though at present ...

  11. Thanks. It is expensive. But I preordered it. I could instead waste that quickly with books by .... Never mind. I will hope it was not as not even wrong a purchase as some.

  12. I saw both bats and angels at the same time. Think that just makes me confused...

  13. Thanks for the link Lubos, I really enjoyed this video. For the first part of the talk I thought Juan did an excellent job giving a colloquium. I believe a good colloquium should allow the brightest undergraduates to understand something and the first year graduate students to understand a lot. This implies many things - not speaking too fast, not putting too much on each slide, defining all jargon, all of which Juan did well - although the question asked at 31:12 seemed to amuse Juan, as he realized he had not defined the boundary, which is kind of important for this talk.

    I really liked the discussion at around 23 minutes connecting conformal invariance, unparticles, and the corresponding action with its connection to AdS. When I read Georgi's paper on unparticles in 2007, I wrote to my group "I have a vague idea that Georgi's unparticles could have analogs in the CFT that is something like the QGP"; it's nice to see I was not completely wrong.

    I of course also enjoyed the discussion on ER=EPR, but here I think Juan may have lost part of the audience. Note that the Penrose diagrams that start appearing around 41 minutes aren't defined or explained. Readers of TRF may be familiar with them, but I would not assume a canonical colloquium attendee would be.

  14. I jsutd noticed, they;ev block'de Idear!

    "Voting Irreguraklities"?! Seriously? It is obvious to anyone with an IQ of at least 90, that Idear is not such an idiot to cerate sockuppet accounts to uppvote his own posts!!!

    They just nweed excuses to block poepple doens't they? !

    I can't believe it, the only real reasons I can see are:
    - His question about the unexpected breakthroughs was closed, which infuriated a whole lot of people, against the moderators.
    - His meta questions were closed, and once in a while he questioned moderator actions, which was not liked.
    It would be obviously a lie to block him for cooling down, or soemething like that, so they chose a reason that which no - body could ever verify, but everyone who has read and seen Idear's postzs would know is false.

  15. I agree with Gordon regarding mixed primary schools. In Ireland most boys are trouble for the female teachers (great majority), the reason being there isn't enough outdoor physical activities during the day for them and the pace of teaching is too slow...
    (I found a solution to this problem by putting my son into a boys school, best move ever. He loves it. Even female teachers have to adapt their teaching to the boys...).

    My daughter is in a mixed secondary school school and she was telling me recently how surprised she always is by all the boys' ability to understand complex technical things even among the most silly boys ;-). I totally agree with her.

  16. Those SE trolls with too much power are obviously going completely nuts now. This is so abominable and almost makes me vomit :-(0).S

    They want to continue banning and harrasing people who are much smarter than them and capable of posting good research-level questions and answers, to dumb the site down to just a popular level playground for SE politicians and disciples who like to play physicists ..? ! Maybe this should not surprise me, as the APODs give a damn about experts being there and their expertise since more than a year ... So be it, happily we are making good progress with PhysicsOverflow ...

    Something to cheer us up: Urs Schreiber will be there on PhysicsOverflow too (even though he has not the time for being a moderator) :-). He told me in a nice email.

  17. Nice talk, thanks.
    How does time in the bulk is related to time in the boundary?. I mean, movement in the extra dimension in the bulk seems to be related to change of size in the boundary, so the time coordinate of both are somewhat correlated. But then, the boundary is in the far future of the bulk.

  18. Dear Juanjo,

    no! The AdS boundary is infinitely far in space, not time. (De Sitter space without "anti" has boundaries in inifinite past or future, but dS/CFT isn't quite as well-established as AdS/CFT.) Once may always choose many different time coordinates but the most natural "global time" coordinate is the same thing in the bulk and in the boundary.

    One must realize that this assertion is lacking some beef because we can't identify all the spatial coordinates as well. That's why we can't identify "points" or events - at most, various regions in the boundary may be linked to "causal diamonds" in the bulk, and so on.


  19. I see now.
    Thanks a lot!

  20. Eelco HoogendoornFeb 9, 2014, 1:03:00 PM

    You clearly havnt attended any university courses in recent decades... There may be exceptions to the rule, but most degrees are nothing but degrees in regurgitation.

  21. Why he says that the correspondence
    is applicable for AdS geometries? We only care if it is asymptotically AdS. AdS
    is just the vacuum state of the geometry we can fluctuate it as much as we

  22. Great! I really hope he and many other such experts will take part in PO!

  23. In the slide appearing around 0:30:45 it is written:

    “Then in AdS we have a theory of particles of spins less than two in almost flat space”

    I guess this is a typo, he means less than or equal to two…

  24. Hmm, that's irony, right? (courtesy of Sheldon :))

  25. kashyap vasavadaFeb 9, 2014, 5:27:00 PM

    Hi Lubos: I am trying to understand these matters as much as I can without a good background of ST. Can you say something more about why " dS/CFT isn't quite as well-established as AdS/CFT" If I understand, DS (positive Lambda) is our known universe . But it might be that for some reason DS/CFT does not work mathematically well enough. Also what do they mean ADS asymptotically only. What about nearby, non -asymptotic region (nearby may not be proper word opposite to asymptotic!)?

  26. LOL, thanks for your lessons, and I haven't attended a regular college course for nearly a decade although I would teach one as recently as 7 years ago.

    But I refuse to believe that colleges are already at the level of the high schools in this respect, at least colleges similar to those I have been connected to in my life. And at least when compared to the - euphemistically speaking - shitty high school I had attended before.

  27. Dear Giotis, AdS/CFT is relevant for all the bulk geometries and their excited states that are asymptotically AdS.

    Empty AdS is dual to the vacuum state in the CFT; non-empty but still asymptotically AdS configurations are dual to non-vacuum, excited states in the CFT.

    One may refine this discussion and distinguish the different rates at which we demand the AdS geometry to be approached at infinity, and this refinement forces us to distinguish the two dual operators in CFT, normalizable and non-renormalizable modes, and so on. This stuff has been known rather well from the first months of AdS/CFT - Witten; Gubser and Klebanov and Polyakov were already pretty sure about those matters, I guess.

  28. LOL, if you could prove that God is malicious, it would probably follow that dS/CFT has to be the one case inaccessible to the (A)dS/CFT methods, to screw us. But I think there's no way to prove that God is evil in this sense...

    More seriously, dS/CFT implies that the new holographic dimension has to be time-like; and the boundary of dS really has two components - in the past and in the future. Both of these features (not shared by AdS) make the dS case more problematic.

    Yes, it means that we're not using the holographic duality directly to our visible near-dS universe. The dual CFT, even if it were AdS, would have 10^{120} strongly coupled fields so the dual CFT description wouldn't be terribly useful as a method to calculate. ;-)

    Non-empty geometries that are asymptotically AdS correspond to excited (not vacuum) states in the CFT. Obviously, the holographic duality would be useless if it allowed us to map a single state (empty AdS to the CFT ground state). It's only interesting because it allows us to translate any state - any CFT state is dual to a corresponding state with bulk matter and bulk curvature (which is inevitably, e.g. because of backreaction) which still preserves the asymptotic AdS structure of the bulk geometry. See also another comment in this thread I just added.

  29. If you mean that gravitons are the last ones to be included, I agree. ;-)

  30. Lubos,

    Aren't you being a little too hard on Aristotle?

    Charles Murray wrote a book called Human Accomplishment in which he quantitatively measured human accomplishment through history. Murray said that he believes Aristotle was the most accomplished person who ever lived:

  31. Yes :-)

    I have favorited it such that I can put it back into the reopen queue, as soon as 3 APODs or a mod have said leave closed (which they certainly will ...) ... ;-). I have explicitely seen only 2 leave closed votes till now, so I still wait a bit ...

    Yeah, thangs God and Rod Vance (missed him in the review queues) :-)

    BTW this should get reopend too (it has only two votes at present and the APODs already kicked it out of the queue)

    It is really too bad that only the APODs are vigorously hyperactive in the review queue, whereas reasonable people such as Rod Vance are seen only sporatically.

    Maybe we should start a "Request for Reopen Votes", such as Math SE and MathOverflow have it, too somewhere ...

  32. @Dilaton or anyone else,

    Could you help kick this back into the reopen vote review queues?

  33. I thought the black things were bats, so I'm more inherently scientific than you, who said they were demons. :-)

  34. I voted to reopen shortly after the outkicking event, so it should be back in ;-)

    We need more reopen votes however ...

  35. LOL, and I was the more culturally accurate because "demons" appears in the name of this work.

  36. I would reopen, but I do not have enough rep.

  37. The PO site looks interesting, am not knowledgeable enough in that field I fear though

  38. Damn! He's as smart as Lubos! Looks like the better you understand a problem the clearer it gets.

  39. LOL, and I was the more culturally accurate one ...

    Nice use of the definite article! This is evidence that you're internalizing it.

  40. Who are you on Physics SE ?

    If we successfully get many of Lumo's friends who have been active on TP.SE back, I will most of the time have to shut up and learn too ... :-P
    But I want a nice troll free higher level site for the (theoretical) physics community to be there by all means :-)


  41. I am always learning, and love all aspects of Physics (and Mathematics as well). The steeper the learning curve, the more I enjoy the challenge.

  42. Well, I don't hold an academic position either (and have a PhD in Physics) - I do so by choice. I also appreciate critiques on my ideas, especially when they are followed with a learning opportunity. Definitely sounds like an interesting gentleman.

  43. Rereading that question, I still can not figure out how and why it is closed - mind you, I have a personal bias, a lot of my friends and I discuss and debate (and research) that quite often. I feel that it is a legitimate Physics question.

  44. I think we should all stop adding the "(theoretical)", since the scope of PO is supposed to be not just theoretical, but, experimental and phenomenological, too. Otherwise, +1.

  45. Have you seen this

    David z and the usual APOD are now seriously ganging up (as Idear calls it ;-) ...) to ban many high-level technical questions officially in chat ... :-)

    This is just horrible that not a single reasonable person is there to defend legitimate questions of people who want to learn physics at a technical level, against David Z s attacks ... :-/. But we two should better keep out ..

  46. Lubos around 0:17:50 Juan says that the duality holds for any quantum theory of GR.

    Do you agree with that statement?

  47. Yes. Well, for any consistent QG theory. It largely means string theory. I am almost sure that Juan agrees in these matters in detail.

  48. Now, I am even more enticed to participate - experimental physics is my playground so to speak. Though, I dabble in TP and phenomenological aspects too.

  49. This is a very nice beautiful and quite accessible talk (applying the method of pushing the pause button to reread slides / ponder about stuff a bit).

    Juan Maldacena is such a nice speaker, clearly and patiently explaing things, and he has a good portion of humor too :-)

  50. CentralCharge15

    First I got really upset abou this vzn guy picking at MathOverflow in their chat and meta, but for some reason he now makes me just chuckle

    Even quid has stopped talking to him, and it is some kind of funny to observe how he manages to keep continuously self excited without external reason, LOL :-D

  51. CentralCharge15

    Now Maishearth explicitely attacks high-level technical questions

    I will have to immediately import all of user43126s question ...

    Manishearth's point of view, that all questions have to be useful to the large laymen crowd of random googlers nicely fits in with SE trying to dumb the level of each site down to popular by means of this off putting hot list. Now I understand why he claimed that your Area51 Popular Science proposal is a duplicate of Physics SE: he is on a crusade to make Physics SE a popular physics site, where technical questions, that are of interest to people who want to learn physics at a deeper than equation free level, are forbidden.

    Maybe we should start a bet about how long it will take him (and other SE representants) to start attacking all research-level questions on Physics SE, as those are only useful to researchers (to be) and not to the whole crowd of random googlers ... ;-/

    BTW is the place for making the suggested edits of PhysicsOverflow on your Wikia ready, and could you give me the link? The Technical Private Beta will most probably come this week :-)

  52. Yes I have already seen that. "Way to misintrepret things" uis always his standard, meanginless respoonse, which doesn't need to be relevant to the situation at all...

    As you have said below, I indeed sometimes suspect these people to be bots, or at least very similar to such, bot-minded, botl-ike people.

    Meanwhile, see what some idiots call themselves here: Just a funny sidentoe...

    The suggesting edits page is ready (it functions):, if you have any ideas, e.g. to make it more user-friendly, etc., you can just edit the page, or add a comment below the page,...