GReat talk by Ashtekar at the kitp singularities workshop

[marcus]

http://online.kitp.ucsb.edu/online/singular_m07/

go here for Ashtekar talk
http://online.kitp.ucsb.edu/online/singular_m07/ashtekar/

I found it works best when I click on "download the whole movie", which takes 5 minutes or so
and then I have it on desktop so i can listen to it again.

the first 17 minutes are nontechnical enough that just about anyone here at PF would get something out of watching

after that it gets technical

then about 53 minutes into the hour talk he shows quantum states of the universe doing a bounce thru where the classical singularity used to be-----they did a bunch of computer modeling of different-size universes----so there are graphs showing how quantum states evolve thru where the bang was

so that's a nice part of watching near the end.

another nice thing is the skillful way that Ashtekar interacts with a room full of stunned and incredulous string theorists

 

[Chris Walters]

Hello Marcus: thanks for your excellent assistance and information on daughter universes and the conference. I am appealing getting the boot for
allegedly spamming but thought you would like to know one for the leading
Physicist said he would permit me to quote him in my publication

With Kindest Regards Chris Walters


Dear Moonbear: Many thanks for bringing your question to my attention on a violation for spam which I would like to reply too. I found several questions asked by other members on your website which were addressed by citing my website which quotes the leading experts in the different fields. I believe Universe Man asked asked for instance the smallest size.

I might ask if the questions asked by and addressed fellow subscribers by leading experts and Professors really need to be categorized as spam? You may be unaware my websites have been visited 1.8 million times and hardly need to spam anyone at Physics Forumn.

With Kindest Regards
Chris Walters

PS One of the leading Professors gave his permission for me to quote his book in the conversation

 

[kneemo]

http://online.kitp.ucsb.edu/online/singular_m07/
another nice thing is the skillful way that Ashtekar interacts with a room full of stunned and incredulous string theorists

Hi Marcus

Ashtekar also holds his own during the talk by Steve Shenker. Despite the title of the talk, the conversation came down to the viability of the AdS/CFT model. Many of Smolin's TWP concerns were addressed. It's a must see!

Struggling with Spacelike Singularities in String Theory
http://online.kitp.ucsb.edu/online/singular_m07/shenker/"

 

[josh1]

...another nice thing is the skillful way that Ashtekar interacts with a room full of stunned and incredulous string theorists

Can you give just one example of a claim made by ashtekar that left string theorists ''stunned''?

 

[Kea]

Hmm. Re Ashtekar at Shenker's talk - I was not so impressed. Anyway, I found it hard to hear what Ashtekar was saying. Shenker got nowhere at all with talking about singularities. There was his five minute example at the end only. I thought he made a great effort to really talk with the audience about the basic usefulness of AdS/CFT, which he claims is 'universal' in that any theory of QG should be able to describe it. If I was a String theorist, I would certainly not have been convinced that this idea is false.

 

[marcus]

I just watched the discussion about Do Baby Universes bounce from hole collapse...
it was split between Gary Horowitz who was at blackboard leading for the first half hour and Ted Jacobson in the second half.

In these cases everybody has to judge subjectively for themselves about tone of discussion, the sound in the room, and state of mind of participants. I can tell you how I took Jacobson's talk you might interpret the tone quite differently.

I liked Jacobson's attitude. It seems to me that he is skeptical about whether AdS/CFT applies to black holes! He kept asking them to convince him and never got anything very reassuring. I was left with the impression that AdS/CFT may actually FAIL in some interesting ways and circumstances. (edit: to clarify, it might fail to apply for example because there is a positive cosmo constant and nature is simply not AdS and the correspondence may not be adaptible in a useful way to the deSitter case)

The most interesting timeslot for me in the BabyU discussion was minutes 40-45.

If you watched it tell me your reaction (no reason they be the same)

 

[josh1]

I'm guessing this is your response to my post.

...everybody has to judge subjectively for themselves about tone of discussion, the sound in the room, and state of mind of participants.

I can tell you how I took Jacobson's talk you might interpret the tone quite differently.

I liked Jacobson's attitude. It seems to me that he is skeptical about whether AdS/CFT applies to black holes! He kept asking them to convince him and never got anything very reassuring. I was left with the impression that AdS/CFT may actually FAIL in some interesting ways and circumstances.

"tone of discussion"; "sound in the room"; "state of mind of participants";"Jacobson's attitude". It sounds like you're basing your judgements on things other than physics. But maybe I'm wrong and you can be more specific about in what sense they said that AdS/CFT can fail.

 

[Thomas Larsson]

But maybe I'm wrong and you can be more specific about in what sense they said that AdS/CFT can fail.

There is little reason to doubt that local theories may be reformulated as diluted theories in higher dimensions. Whether such a reformulation is useful is another matter. It seems like holographic duals of real, non-supersymmetric theories are horribly complicated and much less tractable than the local theory one started from.

However, there are good reasons to doubt that physical gravity is a diluted theory. If gravity has local degrees of freedom, it can obviously not be described in holographic terms. It seems plausible that gravity is in fact a local theory, since 25 out of 26 consistent quantum gravities in 2D are local (has correlators which depend on separation). I am of course thinking about the free D-dimensional bosonic string, which may be viewed as a toy model of quantum gravity in 2D coupled to D scalars. According to the no-ghost theorem, the theory is consistent (unitary) for all D <= 26, but it has local dofs except when D = 26. So the D=26 model is a holographic QG, and those with D<26 are consistent, local QG's.

The free string was also considered as 2D gravity coupled to scalars in http://www.arxiv.org/abs/hep-th/0501114 . For some reason, string theorists seem to like this identification when it can be used to discredit LQG, but not when it challenges the holographic dogma.

 

[josh1]

By my remark that "maybe I'm wrong", I didn't mean to say that there's something about ads/cft that I don't understand, but rather that it may be that marcus was in fact referring to something in particular said in the lecture and I was wondering what that was since he doesn't say in his post.

There is little reason to doubt that local theories may be reformulated as diluted theories in higher dimensions. Whether such a reformulation is useful is another matter. It seems like holographic duals of real, non-supersymmetric theories are horribly complicated and much less tractable than the local theory one started from.

Yes, I don't think anyone is claiming that the known theories exhibiting gauge/gravity duality are realistic theories of quantum gravity. These theories seem to be more useful as gravitational descriptions of otherwise intractable gauge theories.

For some reason, string theorists seem to like this identification when it can be used to discredit LQG, but not when it challenges the holographic dogma.

Baloney. On the other hand...

It seems plausible that gravity is in fact a local theory...I am of course thinking about the free D-dimensional bosonic string...

Would you think me unfair if I we're to suggest based on this that people working outside the string paradigm tend to invoke string theory only in support of their own opposing beliefs? In any event, the model used in the paper you quoted has nothing to do with any kind of realistic quantum theory of gravity. It is as you said just a toy model. It was meant to illustrate the ideas in that paper, but in rather unphysical terms having little to do with strings as a physically relevant theory.

 

[Thomas Larsson]

Would you think me unfair if I we're to suggest based on this that people working outside the string paradigm tend to invoke string theory only in support of their own opposing beliefs?

Yes, I think that would be unfair. I have never questioned technical results in string theory, only their relevance, and I do appreciate NPZ's argument against LQG. In fact, I made similar comments at the end of http://www.arxiv.org/abs/hep-th/0404033 . Besides, I am deeply impressed by CFT, which I guess counts as string theory. It was the desire to generalize CFT to higher dimensions that led to the discovery of the higher-dimensional generalization of the Virasoro algebra, eq. (13) in that paper.

In any event, the model used in the paper you quoted has nothing to do with any kind of realistic quantum theory of gravity. It is as you said just a toy model. It was meant to illustrate the ideas in that paper, but in rather unphysical terms having little to do with strings as a physically relevant theory.

A toy model can still be useful as a counterexample. So the next time you hear David Gross and others say that there are no local observables in quantum gravity, you know that counterexamples exist.

 

[josh1]

Yes, I think that would be unfair.

Good, because you'd be right.

So the next time you hear David Gross and others say that there are no local observables in quantum gravity...

There's a very good reason that you're not going to hear many string theorists straining to convince people that there can be no local observables for quantum gravity. The reason is that the observation that our universe has a positive cosmological constant means that space is probably compact thus requiring a much more local definition of gravitational observables than strings seems to provide right now. This is because in it's present form, the only gauge-invariant observable for a realistic treatment of quantum gravity in string theory is the S-matrix, and it's definition requires the presence of a spatial infinity. The hope was that there might be some analog for de Sitter space of ads/cft, but such a dual description has yet to be found. The problem then becomes one of finding solutions of string/M-theory that have a positive cosmological constant. As you probably know, the construct known as the landscape is supposed to consist of such solutions. However it is not yet known whether the members of the landscape are indeed such solutions.

If string theory is right, it's hard to believe that the ugliest possible set of candidate solutions are in fact the physically relevant ones, at least for doing quantum gravity. This suggests that if string theory is correct, there is something extremely strange about the gravitational physics of the vacuum that we simply do not yet understand. In fact, all of the fundamental difficulties encountered in doing quantum gravity can be traced back to our failure to understand what this strange thing might be.

 

[marcus]

There's a very good reason that you're not going to hear many string theorists straining to convince people that there can be no local observables for quantum gravity. The reason is that the observation that our universe has a positive cosmological constant means that space is probably compact thus requiring a much more local definition of gravitational observables than strings seems to provide right now. This is because in it's present form, the only gauge-invariant observable for a realistic treatment of quantum gravity in string theory is the S-matrix, and it's definition requires the presence of a spatial infinity. The hope was that there might be some analog for de Sitter space of ads/cft, but such a dual description has yet to be found. The problem then becomes one of finding solutions of string/M-theory that have a positive cosmological constant. As you probably know, the construct known as the landscape is supposed to consist of such solutions. However it is not yet known whether the members of the landscape are indeed such solutions.

If string theory is right, it's hard to believe that the ugliest possible set of candidate solutions are in fact the physically relevant ones, at least for doing quantum gravity. This suggests that if string theory is correct, there is something extremely strange about the gravitational physics of the vacuum that we simply do not yet understand. In fact, all of the fundamental difficulties encountered in doing quantum gravity can be traced back to our failure to understand what this strange thing might be.

Let's assume that the ads/cft correspondence fails to apply to the real universe because reality is de Sitter, not anti-de-Sitter----i.e. positive cosmo constant. And also let's assume that the conjectured correspondence is not successfully adapted to form an analog version applicable to nature.

Then what remains of the arguments that baby universes can't branch off?

What Ted Jacobson was presenting (when he wasn't being distracted by talk about ads/cft) was a pretty commonsense position, I thought.

Basically his main message, that I heard anyway, was time may continue down a black hole.

I don't think the participants presented convincing reasons to imagine that it doesnt! And as to what the continuation of spacetime might look like, he seemed thoroughly agnostic.

 

[josh1]

AdS/CFT is a particular example of a more general conjecture that goes under the name of gauge/gravity duality or the gauge/gravity correspondence. The conjecture is that all theories of gravity have a dual description which is complete and in the form of a CFT.

If the gauge/gravity correspondence is true, then since QCFT's are unitary, physical processes on the gravity side that lose information are forbidden. We focus on the AdS/CFT case since there is an overwhelming amount of evidence that this particular instantiation of the conjectural gauge/gravity correspondence is true. It’s worth pointing out that it’s possible that there are examples of AdS/CFT that are different then the one we already know about and that do provide realistic descriptions of gravity. For example, there may be instances in which the compactified space is something other than a sphere or torus and gives rise to realistic particle phenomenolgy. We just don’t know.

Let's assume that the ads/cft correspondence fails to apply to the real universe because reality is de Sitter, not anti-de-Sitter----i.e. positive cosmo constant. And also let's assume that the conjectured correspondence is not successfully adapted to form an analog version applicable to nature. Then what remains of the arguments that baby universes can't branch off?

What Ted Jacobson was presenting (when he wasn't being distracted by talk about ads/cft) was a pretty commonsense position, I thought.

Basically his main message, that I heard anyway, was time may continue down a black hole.

I don't think the participants presented convincing reasons to imagine that it doesnt! And as to what the continuation of spacetime might look like, he seemed thoroughly agnostic.

Well, you can certainly choose for whatever reasons to believe that nature does allow nonunitary processes (which may or may not exclude baby universes, though the focus of the seminar was a bit more general than the title suggests) and simply forget about the gauge/gravity correspondence, or you can take the majority view that AdS/CFT is perhaps the best (and maybe only really good) clue we have about quantum gravity and build on it. Which road would you choose?

 

[marcus]

My window on the Ted Jacobson discussion, I said earlier was minute 40-45 but I can actually narrow it down some. I hope other people will focus in and have a look---anyone interested in these issues who happens to be reading this thread. On my quicktime movie player the key timeslot is
from 42:10 thru 43:40

this is one and a half very revealing minutes that I recommend everybody interested in the Kitp spacetime singularities meeting should listen to carefully.

I have played them over several times with increasing pleasure.

whether one chooses to share TJ's agnosticism depends somewhat on how one imagines the basic choices---not everyone would define the alternatives the same way.

As time permits I will quote a few seconds of the dialog right around minute 43.

If you didnt get the link yet, it is
http://online.kitp.ucsb.edu/online/singular_m07/babybh/
this is for the BabyU discussion, Jacobson comes on the second half of the hour.

 

[marcus]

Time 42:10 is where Ted Jacobson is doing a crit of Steve Shenker's attempt to study black hole singularities in an AdS/CFT framework, and to apply the conjectured correspondence. And TJ has just pointed out questionable and contradictory stuff---difficulties with applying the complementarity to that situation, so he says:

This is the sharpest puzzle and objection that I can see to the point of view you are trying to maintain. So far I see no way out of this...

That is, no way out for you Shenker...your framework just won't fit the BH situation and runs into too many contradictions. So then Steve Shenker comes back with what is essentially a challenge. You Jacobson have been showing contradictions in my setups, but what do you think really happens when a black hole forms?

He asks TJ what is his BASIC HUNCH that he starts out with. Nobody has any certainty or any idea that they actually believe in---but you can still ask a theorist about his initial take on something before getting into detailed analysis.

TJ says OK my basic hunch, that I approach the problem with, is that when a black hole forms "evolution has to continue at a singularity, because there's a timelike direction there"

I'll get some more quote as time permits.

 

[josh1]

Time 42:10 is where Ted Jacobson is doing a crit of Steve Shenker's attempt to study black hole singularities in an AdS/CFT framework, and to apply the conjectured correspondence. And TJ has just pointed out questionable and contradictory stuff---difficulties with applying the complementarity to that situation, so he says:

This is the sharpest puzzle and objection that I can see to the point of view you are trying to maintain. So far I see no way out of this...

That is, no way out for you Shenker...your framework just won't fit the BH situation and runs into too many contradictions. So then Steve Shenker comes back with what is essentially a challenge. You Jacobson have been showing contradictions in my setups, but what do you think really happens when a black hole forms?

He asks TJ what is his BASIC HUNCH that he starts out with. Nobody has any certainty or any idea that they actually believe in---but you can still ask a theorist about his initial take on something before getting into detailed analysis.

TJ says OK my basic hunch, that I approach the problem with, is that when a black hole forms "evolution has to continue at a singularity, because there's a timelike direction there"

I'll get some more quote as time permits.

Marcus, you've completely missed the boat here. The question isn't whether AdS/CFT can handle black holes. In fact the discussion centres around an anti-de Sitter black hole. (In fact these sorts of black hole solutions provide vis a vis AdS/CFT a gravitational description of changes of phase in quantum conformal field theories at finite temperatures). The question is whether AdS/CFT allows things that fall into this black hole to form a baby universe and escape. This is an open question and really does not bear in any way on the basic validity of the gauge/gravity correspondence including on whether it can describe black holes.

 

[Thomas Larsson]

There's a very good reason that you're not going to hear many string theorists straining to convince people that there can be no local observables for quantum gravity.

I think you are wrong here. That locality is incompatible with quantum gravity seems to be a deeply rooted belief. Not only among string theorists and AdS/CFT people, but also among LQGists. The standard point-moving argument shows that non-trivial correlators are incompatible with proper diffeomorphism symmetry. But locality is equally incompatible with proper conformal symmetry in 2D - the only Virasoro rep with c = 0 is the trivial one. This was my hint how to avoid the no-go theorem.

Anyway, if QG is a local theory, then something like AdS/CFT cannot apply, simply by counting degrees of freedom.

My understanding is that locality is the reason why Gerard 't Hooft considers hidden variables, see the third slide at http://online.kitp.ucsb.edu/online/kitp25/thooft/ . If you look at the video at the same place, you see why I mentioned David Gross in my previous post; 't Hooft quotes Gross saying "There are no local observables in quantum gravity, so forget it."

Of course, combining locality with gravity leads to several heresies: consistent gauge anomalies, diff anomalies in 4D, replacing QFT by QJT. But the fact that 't Hooft considers hidden variables makes me feel conservative in comparison.

 

[josh1]

I think you are wrong here. That locality is incompatible with quantum gravity seems to be a deeply rooted belief.

I wasn’t talking about what they believed. I was talking about how much effort they put into convincing people that this belief must be true given the current situation.

't Hooft quotes Gross saying "There are no local observables in quantum gravity, so forget it."

In fact, string theorists are indeed now wondering whether string/M-theory can or should provide an explicitly gravitational prescription that is in some sense more local (as opposed to completely local in the conventional sense) than the S-matrix. As another reason people are thinking this way, nobody yet knows how to compute the S-matrix without assuming that the spacetime geometry at spatial infinity remains fixed.

That locality is incompatible with quantum gravity seems to be a deeply rooted belief. Not only among string theorists and AdS/CFT people, but also among LQGists.

AdS\CFT people are string people. Also, are you sure about the lqg people? The reason I ask is that the fundamental degrees of freedom of lqg include the vertices of spin networks and these measure spatial volume. This means that lqg is not a holographic theory of gravity and holography and locality are closely related.

 

[Thomas Larsson]

Also, are you sure about the lqg people? The reason I ask is that the fundamental degrees of freedom of lqg include the vertices of spin networks and these measure spatial volume. This means that lqg is not a holographic theory of gravity and holography and locality are closely related.

IIRC, the observables in LQG are Wilson loops, hence non-local. Locality in the sense of correlation functions that depend on separation (which is what I mean by locality) are not compatible with background independence in the sense of LQG. I once had a discussion with f-h about this; he can probably provide references.

Anyway, you are smart enough to realize that infinite conformal symmetry in the strict sense is also incompatible with locality. In order to have a non-zero anomalous dimension, you need a conformal anomaly. The diffeomorphism algebra in higher dimensions works in the same way. This is why we need the Virasoro-like extension of the diff algebra to combine locality with background independence.

 

[josh1]

IIRC, the observables in LQG are Wilson loops, hence non-local. Locality in the sense of correlation functions that depend on separation (which is what I mean by locality) are not compatible with background independence in the sense of LQG. I once had a discussion with f-h about this; he can probably provide references.

...infinite conformal symmetry in the strict sense is also incompatible with locality. In order to have a non-zero anomalous dimension, you need a conformal anomaly. The diffeomorphism algebra in higher dimensions works in the same way. This is why we need the Virasoro-like extension of the diff algebra to combine locality with background independence.

Yeah, I've been thinking about this and realized you're right. Thanks.