Sabine Hossenfelder watched a 43-minute video on loop quantum gravity that was posted two days ago. She wasn't too impressed.
Well, I am always impressed by low-budget or no-budget teams that manage to shoot a semiprofessional video of this size but it's hard to avoid some criticisms. Most of them are really criticisms of the topic they chose to cover, loop quantum gravity, so they shouldn't be used against the creators of the video. And I will avoid detailed criticisms of the imperfect sound quality (noise filters were used too much at some places), the subtitles (and whether it makes sense to have a video if there exists a written form of it), and the speakers' limited rhetorical abilities.
00:00 It's strange that loop quantum gravity is being connected with the birth of stars etc. because it isn't really capable of explaining the particle spectrum and other parts of physics that are crucial for our understanding of the early Universe.
00:40 Loop quantum gravity is said to be "treating gravity quantum mechanically". While this phrase is omnipresent and seems like an innocent one, and it can even be interpreted innocently, in the strict form – which is probably one chosen by most of the speakers – it is a seed of lots of confusion. A theory of quantum gravity isn't just the same classical physics "treated" differently. It is new physics, a brand new theory that leads to different conclusions and predictions. The procedures needed to find a quantum theory with certain properties are, in general, more complicated and different than just "placing hats on the heads of all objects in classical physics". Loop quantum gravitists sort of understand that a quantum theory differs from a classical theory. But they live in the misconception that the transition from a classical theory to a quantum theory is straightforward and one-to-one. It's surely not. The space of quantum theories with certain properties is completely different from its classical counterpart. Their consistency conditions are different, and so on.
00:50 LQG is said to tell us that space has discrete units. Well, it pretty much assumes it so it's misleading to say that it's a prediction. And this assumption has indeed far-reaching consequences – far-reaching destructive consequences, to be precise. It makes the theory incompatible with the Lorentz invariance, other continuous symmetries, and it is a property that guarantees that the infinite indeterminacy seen in the non-renormalizable Einstein's theory can't be cured by this attempted theory. It's unfortunate to hype this property of LQG without asking whether it's actually a good thing or a bad thing and without making a credible attempt to answer this question. But this hyping is being done not only by filmmakers but also by most LQG proponents.
01:00 Ivan Agullo is said to be a "rising star" of the field. I've never heard of the name and INSPIRE seems to confirm it's not my fault. I don't quite get the concept of a "rising star" whose most famous paper has 52 citations shared by 5 authors. It's surely an inflation of words. What about an average young researcher in this subpar discipline?
01:10 Abhay Ashtekar is a pleasant guy. In some of his features of the smile, he resembles Cumrun Vafa (Vafa is Iranian while Ashtekar is Indian by his roots). He read a book by George Gamow as a kid. So did I. Agullo loved algebra and calculus. I don't quite understand why we're told these stories about the childhood of someone whose contribution to physics still remains completely unknown to me – and all other viewers.
02:20 Are you afraid that Einstein failed and you're trying to solve it? Agullo says Yes but it's OK because many people are working on it. Well, many people with much lower capabilities couldn't replace the genius of one individual. What allows modern physicists to go beyond Einstein is their much better mastery or quantum mechanics and related portions of physics. Unfortunately, this can't be said about the LQG researchers which is another reason why they should be afraid of their inevitable failure.
02:50 Ashtekar: Einstein's GR is geometric and predictive, QM is different, probabilistic and algebraic. There should be a unifying theory. These sound like innocent words for every "unifier" of gravity and QM. However, the vague formulation about the two limits – which are completely different and which impose very different restrictions on the unified theory – is a seed of confusion within LQG. In particular, the postulates of QM have to be exact and they affect every dynamical degree of freedom and interaction in physics. On the contrary, the foundations of GR are about one particular field only, the metric tensor or the field communicating gravity, and its interactions with other fields while most of the other degrees of freedom remain unconstrained. Moreover, the original equations of GR don't have a reason to be exact and aren't exact. LQG folks don't understand these differences.
04:05 Ashtekar: Gravity is encoded in the very geometry of spacetime. Well, the same is true for every theory that respects the equivalence principle, e.g. string theory. However, in string theory, one may derive such conclusions from a more fundamental starting point. Ashtekar says things like "there is no background and everything is interacting with everything else". These two statements aren't equivalent. The background is just a condensate that must be allowed to develop and exist. So if you can't get background from your theory, it's doomed. You can't mask this doom by demagogic identifications of the lethal flaw "there is no background" (bad) with the principle that "everything is interacting with everything else" (true, it does in pretty much every theory).
05:44 Comparison with string theory. LQG folks from the GR background, ST came from particle physics. So we're told by Ashtekar that string theorists are focused on the unification of forces and not the peculiar properties of the spacetime itself. That's just untrue. String theory has discovered and still studies lots of new remarkable properties of the quantum spacetime – T-duality, mirror symmetry, other dualities, holography, entanglement is a glue or wormhole, emergence of spacetime from other degrees of freedom, topology change – flops, conifolds, others. It's just a lie that string theory isn't fascinated and fascinating us with new insights about the peculiar properties of the spacetime in the quantum theory.
06:20 In string theory, SUSY and extra dimensions are essential, in LQG, they're not. Friendly speculations about the cooperation of LQG and ST. Not really realistic but OK.
07:30 Agullo about the Big Bang singularity. GR fails over there. He uses "loop quantum gravity" and "loop quantum cosmology" in the same sentence. It's pretty bad that they don't even explain enough to explain the viewer that these are two things and what the relationship between them is. Sabine has already complained about it. LQC isn't really a limit or special case of LQG; it is a simplified toy model of it. It suffers from all the problems of LQG and more. The "cure" of the singularity is presented as an extra quantum-caused repulsive force that creates a bounce. I find this 1) a wishful thinking, 2) a naive strategy to deal with the problem. Such a bounce only moves all the problems to (even more serious problems with) the pre-big-bang cosmology and it doesn't seem one has made any progress at all.
08:50 Sponge has a maximum capacity to hold water, and space has similarly a maximum density. That's just stupid. Even semiclassical GR is enough to see that the density is maximized for the minimum-size i.e. Planckian black hole when it's Planckian, too. But this is really just a transitory regime, not an extreme one. Lighter black holes are like elementary particles and their density isn't too well-defined or isn't finite because they're pointlike in some very specific sense. Larger black holes have a much lower than the Planckian density, and so do all other celestial bodies. There's really no way to make a whole "filled sponge", a macroscopic piece of the maximum-density "material" of space, and if a theory predicts otherwise, it's immediately dead. It's dead because the final stage of the collapse of localized matter is a large black hole whose density (mass/volume) is much lower.
09:50 Big bounce instead of Big Bang. Agullo says the naive things about the bounce, without really solving anything. One could really embed this "solution" even to classical GR as long as one adds some (non-local) interaction. But we really do know there was a Big Bang and by the second law, the entropy "before it" if there was any "before" couldn't have been higher than during the Big Bang. Agullo explains terms in a high-school-level equation but it's both incomprehensible and inadequate for analyses of extreme physical conditions such as the Big Bang.
12:00 Ashtekar says that this bounce force comes out automatically blah blah blah. He's happy about it. It's hard to see where this satisfaction comes from. Every new theory predicts some new effects. A generic effect is bound to get corrections from the new physics. This is not yet a reason to be happy. It is not a consistency check analogous to those in string theory. There are no successful consistency checks in LQG.
13:00 The female narrator starts to talk about cosmic inflation. What does LQG say? Ashtekar says confusing things about the probability of inflation that is small at some moments and so on. Bizarre. Inflation is a generic process that arises in theories with scalar fields and viable enough potentials that are surely not "infinitely unlikely". It arises whenever a region of space sees a scalar field sitting near the maximum of the slow-roll potential. It even arises without the slow-roll potential. Once the small region obeys this condition, which must be clearly allowed because it's an allowed configuration, exponentially large volumes of inflating space will be created. All the "inflation is unlikely" criticism trying to argue otherwise is complete unscientific gibberish, and so is Ashtekar's story about this topic.
14:40 Ashtekar also praises LQC for giving us a preferred instant, the bounce, while the Planck time after the Big Bang isn't special. Those comments are extremely special, too. The Planck scale surely is special in a theory of quantum gravity. Special things tend to occur there or its vicinity by dimensional analysis. On the other hand, the bounce in LQC isn't special, at least not the exact moment, because the evolution isn't time-reversal-symmetric (e.g. because the second law of thermodynamics) so the volume-minimizing instant is just a convention to choose a moment in the middle, and not the only convention. So the situations are equally special or non-special.
16:00 She asks a surprisingly sensible question – doesn't the observed positive cosmological constant rule out the recollapse, namely the bounce? Agullo says that LQC doesn't predict that there will be a recollapse. That's nice but it does say t – or at least they do say - that there was a bounce so it was a collapse of an earlier Universe. One has problems with this in a positive C.C. world, too, especially because the pre-big-bang universe had to be more or less maximally symmetric due to its low entropy.
17:40 Why was the entropy low at the beginning – which is not the beginning in the bounce models? A good question. Agullo says that the entropy was "reset". Wow. I suppose he means that an agent reduced a high positive value to zero. Jesus Christ? This is a flagrant violation of the second law and it is enough to kill any model with this feature. He talks this rubbish to a laywoman for many minutes and she has no way to learn that it is rubbish contradicting basic physics, even physics taught to undergraduates.
19:30 Ashtekar correctly says that we usually count matter entropy and not horizon entropy. But he doesn't elaborate upon this good point. One has to be a bit careful about double-counting here. Well, the matter entropy is always parameterically lower than the horizon entropy so if we count the horizon entropy, it beats the matter entropy, anyway. Ashtekar is trying to solve the same problem as Agullo by equally if not more childish comments. The entropy is "infinite" at the bounce and it drops to zero because the horizon disappears. But this is just an oxymoron. The event horizon is always defined relatively to the causal structure of the whole future spacetime. So if it's somewhere, it just can't disappear. Ashtekar's comments about the disappearance make no sense and they're added on top of Agullo's nonsense about the "reset entropy". Moreover, the idea that one may just ignore the decreasing matter entropy is also wrong. One may only ignore the disappearing matter entropy in GR if the matter ends in a singularity. But if it doesn't, the matter density separately can't decrease, either. Those folks are a community of sloppy folks who just support each other when they talk vague physical nonsense.
21:50 Resolution of black hole singularities. There are no precise calculations, we hear. They hype a Pullin-Gambini paper that happens to be about a similar question. The singularity is replaced by a new large spacetime. You know, that's a big problem, something that would revive all the information loss puzzles after they were solved.
22:50 She asks about Alan Guth who dared to publish a paper that bounces are prohibited. Agullo says that Guth assumed classical GR near the bounce while LQG violates it. Ashtekar agrees. She also asks about other problems of contracting universes. Agullo just says that there are "papers" and they imply that singularities don't appear. But this really means that the theory disagrees with GR in rather mundane contexts, a point that isn't noticed at all.
25:00 Ashtekar is worshiping the repulsive force. I can't get rid of the feeling that these people still live somewhere in the 17th century and they're adding terms to Newton's equations. Physics no longer operates like that and similar extra terms aren't considered fundamental or interesting. Ashtekar says that unlike previous attempts, LQC has a way to make a "cube of size 10 Planck lengths" uniform. I suppose that the mechanism he refers to is the Jesus Christ who resets the entropy and flattens the region. That's nice but the Jesuits already had the same mechanism (Jesus Christ) many centuries ago. All this stuff is ludicrous. Inflation and things equivalent to it is the only known actually operational mechanism to produce flattened, nearly empty large regions of space out of a different environment.
25:45 Ashtekar is asked about the proposal by Smolin – whom he doesn't like – and a physicist with a Polish name that black hole singularities are big bangs of other universes. Asthtekar thinks that this can't be the case because the singularities are different.
26:50 Ashtekar is asked whether inflation may be eternal according to LQC. He doesn't seem to have an answer. He has no intuition, we hear, and asks others to do a strange calculation.
28:35 Agullo is asked why the previous Universe contracted. He says we don't have observational access to this question. That's a fun excuse because if you claim to have a theory replacing the Big Bang singularity, it should give you a mechanism for that process, too. After all, we don't have the observational access to the Big Bang singularity which doesn't mean that we can sweep it under the rug. He's using different standards for his pet theories and the (more viable) conventional theories. Agullo says that quantum effects wash out the information about the previous phase. Great. But the theory should have the information about the overall evolution, otherwise it says nothing about the early Universe.
29:10 The female narrator complains that a paper by Guth et al. (Borde, Vilenkin, Guth) didn't refer to LQG. Well, Alan Guth rarely cites crackpots in general. Agullo meaningfully explains the paper, saying that it claims that inflation can't be eternal in the past. At the end, he says that the theorem may be ignored but there is no justification except for foggy claims that we don't have the final theory yet. A great power of the theorem is that they may prove many things even before we know everything else, a point that Agullo clearly fails to grasp. Ashtekar says that the theorem doesn't apply to LQC because LQC has a contracting phase, something assumed to be absent in the theorem.
32:25 It's hard to test quantum gravity experimentally; interesting that no one dares to criticize these folks for this fact of Nature. Random wrong papers about the Lorentz violation by gamma-ray bursts, wrong ways in which black holes should evaporate, and so on are hyped. The CMB is most promising for Agullo except that his LQG has nothing to say about it.
34:00 Ashtekar says that everything we observe is compatible with LQC. Wow. An unexpected conclusion after half-an-hour of hyping the difference between LQC and the theories that actually do agree with the data.
35:40 Violations of Lorentz invariance due to the discreteness space were experimentally falsified "while other experiments have confirmed it" (the latter is bullshit, all these positive signals have been debunked or explained as being due to a more mundane effect). Ashtekar says that this violation of relativity is a robust prediction of LQG, in analogy with crystals in condensed matter physics. Very true: LQG imagines that the world literally is a sort of a crystal. So why doesn't he see that LQG has been ruled out? He wants to believe that the correction only contribute at the second order. But even the (experimentally falsified) first-order correction is a wild underestimate of the huge violations of relativity that a fundamentally Lorentz-violating theory predicts!
37:10 Detection of gravitational waves, polarization of CMB, etc. What does LQG have to say about these matters? The only honest answer, nothing, isn't heard.
38:50 Small black holes at the LHC. They would be great, Agullo says. All this stuff is incompatible with the LHC because black holes may only appear at the LHC with the extra Randall-Sundrum warped dimensions which aren't a part of LQG. Agullo doesn't explain this fundamental thing – either he doesn't understand it or he finds it beneficial to hide the fact.
39:50 The narrator talks about a Planck press conference, anomalies, differences from WMAP. The video doesn't look honest concerning the question who authored these comments. They're not hers, are they? Agullo talks about some perturbations from the pre-big-bang cosmology and while the gap between the conventional physics and LQC is deep, he blurs the boundaries so that it is often completely unclear whether he is talking about the LQC research or proper physics. Ashtekar admits he doesn't understand these anomalies so he offers general wishful thinking about learning things about the initial state instead. We're told that Planck data are compatible with LQC except that LQC predicts nothing that looks like the real world, and talking about details of Planck data in this context is way over the edge.
The narrator wraps the video by comments about an entirely new vista that the bounce may open.
I don't plan to proofread this text.
Sabine Hossenfelder watched a 43-minute video on loop quantum gravity that was posted two days ago. She wasn't too impressed.