Nima Arkani-Hamed was in Oxford two weeks ago giving two lectures for the philosophy of cosmology conference “Anthropics, Selection Effects and Fine-Tuning in Cosmology.” He also gave two talks in the maths department on scattering amplitudes and a talk in the physics department about building a 100 km circular collider. I think that similar versions of all these talks can be found online, so I’ll just give a broad outline and some nice quotes.
In his first conference talk, “Naturalness and It’s Discontents: Why Is there a Macroscopic Universe?,” he talked mostly about naturalness in relation to the cosmological constant and hierarchy problems, stating that “naturalness problems are not an inconsistency of physics—rather, they’re a guide for what to expect,” and “that something big and structural is needed to remove the cosmological constant problem”. He said, “the broad idea of naturalness is being put under more pressure by the LHC, but I’m not more worried than pre-LHC since people already had to make excuses.” However, “it’s a little disquieting.” He mentioned that nothing else new at the LHC would represent a 1% fine-tuning for the weak scale, which has happened before in physics e.g. the moon eclipsing the sun and the low quadruple of the CMB. However, if it goes to a 1/10,000 tuning, “it would make the Higg’s much more special that these crappy examples.” On the multiverse: “Asking if we’re part of a multiverse isn’t a theory but a caricature of what a future theory might look like.” In response to a question at the end he said: “the manifold structure [of spacetime] is all in our heads–it’s better to phrase things in terms of high-energy scattering amplitudes.”
His second conference talk, “Space-Time, Quantum Mechanics and the Multiverse,” was about the physics motivation for considering a multiverse. Basically, “it’s the only scientific approach to the cosmological constant problem and it happens in some theories e.g. chaotic inflation, string theory, simple toy models.” He said that he suspects that making sense of the multiverse will require equally as radical a step as going from classical to quantum physics. He thinks this will involve getting rid of space-time and thinking carefully about what are the precise observable: “the really big mysteries are cosmological observables.”
He gave a talk in the maths department on the Amplituhedron. He started with a long discussion about how gravity implies the lack of local observables (except at infinity). He then talked about how the least action formulation of classical mechanics helped connect to the next level of description of (quantum) physics and how something similar will probably be needed to get rid of space-time. This is linked to the idea that scattering amplitudes written in terms of Feynman diagrams are made to have manifest locality and unitarity, which requires incorporating gauge redundancy, but in the alternative approach scattering amplitudes for a particularly symmetric (N=4 SYM) theory come out more simply as the “volume” of some region in some space, which means that unitarity and locality are derived. He then spent quite some time defining the amplituhedron idea as a generalisation of the inside of a triangle. (There was a follow-up technical talk the next day that I didn’t attend.)
His collider talk, “Motivations for 100 km collider,” began with: “every physics point in this talk is obvious,” with his view that the main motivation for building a collider being that it’s the obvious future of the field. He said, “we’ve never had a consistent theory valid up to exponentially higher energies–this is a qualitatively different scenario from what we’ve seen in the past,” but that in every scenario he can imagine we will need a 100 TeV pp machine–there are deep structural issues in QFT at stake in finding out if the weak scale is natural. If we don’t find anything more at the LHC, there’ll be a 10% tuning and we’ll want to know if it’s more (the tuning goes as the square of the machine energy) since it’s significant to say that the weak scale is 100 times more tuned than other examples. He also said that he thinks that the Higgs find is undersold: a light Higgs boson means that our vacuum is qualitatively different to a random condensed matter system (“it’s not some crappy metal”). When talking about the 1% tuning in the fact that neutrons don’t bind he said that when he tried to learn nuclear physics as graduate student he found it really confusing. He also talked about his visit to China to discuss building a 100 km collider there and about the new centre for future HEP being set-up in Beijing, which he’s going to spend 2-3 months at every year for the next two years. He thinks there’s a greater than 1% chance that they’ll actually build it, which is why he’s spending that much time there. He noted that a good thing about building the next large collider under an authoritative regime is that you only have to convince a few people (unlike in the US). He also mentioned that he had a one hour conversation with Al Gore in which he apologised for the SSC cancellation (he was largely responsible for that). In response to a question about split-SUSY (I think), he said: “the psychological thing with model building is that you have to believe it’s right at the time so you’re motivated to work out its consequences, and when you’re done you forget about it. I wish I didn’t need these psychological crutches, but life is hard.”