David Goss wrote to tell me that Physics Web has an article about physics weblogs. The theorists quoted are Sean Carroll, Paul Cook and Dave Bacon.
There’s an article by Michael Green in the latest Nature Physics reporting on the recent 23rd Solvay conference (mentioned here and here). Green notes that “Much of the discussion focused on string theory” but that “the structure of string theory is still so badly understood that it does not yet deserve to be described as a ‘theory’ at all — it is more a ‘work in progress’.” He describes the discussion at Solvay about anthropic explanations of the CC as “lively”, but that Polchinski et. al. have yet to carry the day: “there is a strong body of opinion that holds that it may be premature to decide which parameters are environmental within string theory, as the structure of the theory is still poorly understood and will surely hold further surprises.”
The same issue of Nature has an essay by Lawrence Krauss entitled Anthropic Fever where he starts by explaining the standard Landscape story that fundamental physics is really an “environmental science”, but then goes on to write:
But I have been wondering whether things might actually be much worse. It could be that many different combinations of laws could allow life to form, and that it is a pure accident, not favoured by any particular probabilistic explanation, that the constants of nature result in the combinations we experience in our Universe. Or, it could be that the mathematical formalism is so complex that the ground states of the theory might not be mathematically determinable, even in principle.
Whether or not nature is ultimately ‘undecidable’ in this strong sense, these ideas point to a possible future for particle physics that is very different from the past. Fundamental physics might not be restricted by any underlying grand mathematical structure that would ‘explain’ why the Universe is the way it is. It’s a possibility that I hope will be wrong, but it’s a possibility nonetheless.
This kind of worry seems to me completely misplaced. There’s not a shred of evidence for it, and the only reason people have been engaging in this kind of speculation is that string theory led them down this sorry path. Finding a trackless swampland at the end of the path doesn’t mean that physics is destined to spend the rest of eternity mucking about in a swamp, it is far more likely that a wrong turn was taken quite a ways back.
Last night a write-up by Nati Seiberg of his rapporteur talk at Solvay appeared at the arXiv, entitled Emergent Spacetime. It does a very good job of laying out the reasons that people often say that string theory suggests that our standard concept of “space” needs to be revised, that perhaps space is an “emergent” concept. Personally I don’t think the different reasons that Seiberg lays out add up to a very convincing case. For one thing, string theorists have been trying to come up with a new “stringy” version of space for twenty years now, without much success at all. They are still far from anything like a consistent proposal of what this new idea about space will be, and the various evidence given by Seiberg is rather incoherent, leading to different kinds of generalizations of space, not pointing to any one of them in particular.
The past two days there was a conference at Harvard on Black holes, topological strings, and invariants of holomorphic submanifolds, which also included some lectures in memory of Raoul Bott by Sir Michael Atiyah, S. T. Yau, and Dan Freed. Lubos has reports on two talks there, one by Robbert Dijkgraaf about a hoped for “Universal Wave Function”, a Hartle-Hawking kind of wave-function that would give the amplitude for the universe to be in various parts of the Landscape. The second was by Frederik Denef who spoke on “D-brane ground states, multicentered black holes, DT/GW correspondence, and the OSV conjecture [or: why OSV is probably right].” Lubos reports on a conversation with Frederik about his two forthcoming papers with Michael Douglas on computational complexity and the Landscape. It seems that what Denef-Douglas show is that, even if everything one would like to calculate is in principle calculable, the problem of identifying a specific string theory background realizing anthropically small values of the CC is NP-hard. This means that in practice you’ll never be able to do what landscapeologists would like to do: use the observed values of the CC and maybe some other standard model parameters to identify a tolerably small number of backgrounds, then use the properties of these backgrounds to make predictions. I believe it is this possibility that Krauss is alluding to in his quote above about how the “ground states of the theory might not be mathematically determinable, even in principle.”
Update: The last paragraph has been modified to better reflect reality. In its initial version I had assumed from Lubos’s blog entry that these computational complexity issues had been what Frederik’s talk was about.
Update: Polyakov also has a new preprint based on remarks at the Solvay conference, entitled Beyond Space-time. It’s a mixed bag, mostly about various ideas related to AdS/CFT, as well as the cosmological constant. He begins with critical but not completely dismissive comments about anthropic arguments:
Another danger is to get distracted by non-dynamical anthropic arguments, which recently acquired some popularity. I find the anthropic principle irrelevant. It is unlikely to uncover fundamental ideas and equations governing the universe. But, in spite of these misanthropic remarks, I believe that in special cases anthropic arguments may be appropriate.
At the end of the paper, he characterizes the various topics he has discussed as follows:
As it is clear from the list of the references below, these ideas (except for the gauge/strings correspondence) did not attract any attention. Perhaps they don’t deserve it. My best hope, however, is that some of them may serve as small building blocks of the future theory, the vague contours of which we can
discern at the horizon.