### The Cobbler’s Children

One of the disadvantages of chairing a session is that you end up concentrating on the clock, and trying to keep the speakers on schedule, rather than concentrating on the talks themselves.

Shamit Kachru talked about his beautiful paper with DeWolfe, Giryavets and Taylor on flux compactifications of Type IIA. A lot of effort has gone into the study of IIB flux compactifications, and the feeling was that these were somehow “typical” of (but more tractable than) generic string compactifications in which the moduli are stabilized. But these Type IIA flux compactifications behave qualitatively differently. Tadpole cancelation doesn’t produce a bound on the RR four-form flux. So the flux is a free parameter, $N$, and one finds an *infinite* family of vacua parameterized by $N$, accumulating at large radius ($R\sim N^{1/4}$) and weak coupling ($g_s\sim N^{-3/4}$)^{1}.

Given this (and other) qualitative differences between the IIB and IIA cases, we ought to be circumspect about naïvely hoping that the sorts of behaviour one finds in these backgrounds generalize to other (heterotic, …) classes of compactifications.

Vijay Balasumramanian gave one of his patented enthusiastic talks about some forthcoming work, in which he and collaborators have a rather detailed proposal for the AdS/CFT duals of the microstates of AdS blackholes. These are (multi-)traces of strings of $\mathcal{O}(N^2)$ fields in the gauge theory. The claim is that, to “most” probes, these states look precisely thermal. But, to a very few, there is a large, highly non-thermal, response. I hope I have the time to give a detailed account when the paper(s) come out.

While most of my colleagues were visiting Niagra Falls, I spent a few hours this afternoon in the company of Joe Clark, which is a whole 'nother story …

^{1}And, unlike in Freund-Rubin or related “compactifications”, even these “large-radius” vacua have a clean parametric separation of the Kaluza-Klein modes,
$\frac{1/H}{R} \sim N^{1/2}$
from the low-energy 4D degrees of freedom. On the other hand, despite the fact that the string coupling is going to zero and the radius is getting large, Shamit seemed to argue that the tunnelling probability between these vacua *is not* going to zero as $N\to\infty$. This is a bit subtle. The domain wall tension, measured in 4D Planck units, goes to zero quite fast at large-$N$. However, the bulk energy gain goes to zero, as $\Lambda$, itself, is going to zero even faster.