Hey, has anyone heard about this string theory landscape business, and the anthropic principle, and some sort of controversy? Hmm, I guess they have. Perhaps enough that whatever needs to be said has already been thoroughly hashed out. But, hey! It's a blog, right? Hashing stuff out is what we like to do. So I'll modestly point to my own recent contribution to the cacophony: Is Our Universe Natural?, a short review for Nature. To give you an idea of the gist:
If any system should be natural, it's the universe. Nevertheless, according to the criteria just described, the universe we observe seems dramatically unnatural. The entropy of the universe isn't nearly as large as it could be, although it is at least increasing; for some reason, the early universe was in a state of incredibly low entropy. And our fundamental theories of physics involve huge hierarchies between the energy scales characteristic of gravitation (the reduced Planck scale, 10^27 electron volts), particle physics (the Fermi scale of the weak interactions, 10^11 eV, and the scale of quantum chromodynamics, 10^8 eV), and the recently-discovered vacuum energy (10^-3 eV). Of course, it may simply be that the universe is what it is, and these are brute facts we have to live with. More optimistically, however, these apparently delicately-tuned features of our universe may be clues that can help guide us to a deeper understanding of the laws of nature.
The article is not strictly about the anthropic principle, but about the broader question of what kinds of explanations might account for seemingly "unnatural" features of the universe. The one thing I do that isn't common in these discussions is to simultaneously contemplate both the dynamical laws that govern the physics we observe, and the specific state in which we find the universe. This lets me tie together the landscape picture with my favorite ideas about spontaneous inflation and the arrow of time. In each case, selection effects within a multiverse dramatically change our naive expectation about what might constitute a natural situation. About the anthropic principle itself (or, as I much prefer, "environmental selection"), I don't say much that I haven't said before. I'm not terribly fond of the idea, but it might be right, and if so we have to deal with it. Or it might not be right. The one thing that I hammer on a little is that we do not already have any sort of "prediction" from the multiverse, even Weinberg's celebrated calculation of the cosmological constant. These purported successes rely on certain crucial simplifying assumptions that we have every reason to believe are wildly untrue. In particular, if you believe in eternal inflation (which you have to, to get the whole program off the ground), the spacetime volume in any given vacuum state is likely to be either zero or infinite, and typical anthropic predictions implicitly assume that all such volumes are equal. Even if string theorists could straightforwardly catalogue the properties of every possible compactification down to four dimensions, an awful lot of cosmological input would be necessary before we could properly account for the prior distribution contributed by inflation. (If indeed the notion makes any sense at all.) I was asked to make the paper speculative and provocative, so hopefully I succeeded. The real problem is that draconian length constraints prevented me from making arguments in any depth -- there are a lot of contentious statements that are simply thrown out there without proper amplification. But hopefully the main points come through clearly: calculating probabilities within an ensemble of vacua may some day be an important part of how we explain the state of our observed universe, but we certainly aren't there yet. Here's the conclusion:
The scenarios discussed in this paper involve the invocation of multiple inaccessible domains within an ultra-large-scale multiverse. For good reason, the reliance on the properties of unobservable regions and the difficulty in falsifying such ideas make scientists reluctant to grant them an explanatory role. Of course, the idea that the properties of our observable domain can be uniquely extended beyond the cosmological horizon is an equally untestable assumption. The multiverse is not a theory; it is a consequence of certain theories (of quantum gravity and cosmology), and the hope is that these theories eventually prove to be testable in other ways. Every theory makes untestable predictions, but theories should be judged on the basis of the testable ones. The ultimate goal is undoubtedly ambitious: to construct a theory that has definite consequences for the structure of the multiverse, such that this structure provides an explanation for how the observed features of our local domain can arise naturally, and that the same theory makes predictions that can be directly tested through laboratory experiments and astrophysical observations. Only further investigation will allow us to tell whether such a program represents laudable aspiration or misguided hubris.