There are a couple of peculiarities of our universe which it would be nice if the All-Universe Hypothesis (AUH) could explain, or at least shed light on them.
One is the apparent paucity of life and intelligence in our universe. This was first expressed as the Fermi Paradox, i.e., where are the aliens? As our understanding of technological possibility has grown the problem has become even more acute. It seems likely that our descendants will engage in tremendous cosmic engineering projects in order to take control of the very wasteful natural processes occuring throughout space. We don't see any evidence of that. Similarly, proposals for von Neumann self reproducing machines that could spread throughout the cosmos at a large fraction of the speed of light appear to be almost within reach via nanotechnology. Again, we don't see anything like that. So why is it that we live in a universe that has almost no observers? Wouldn't it be more likely on anthropic grounds to live in a universe that had a vast number of observers? The second peculiarity is the seemingly narrow range of physical laws which could allow for our form of life to exist. Tegmark writes about this at http://www.hep.upenn.edu/~max/toe.html. He shows a chart of two physical constants and how if they had departed from their observed values by even a tiny percentage, life would be impossible. In the full paper linked from there he offers many more examples of physical paramters which are fine-tuned for life. So why is this? Why does it turn out that our form of life (or perhaps, any form of life) can exist for only a tiny range of variation? Why didn't it turn out that you could change many parameters a great deal and still have life form? I don't see anything a priori in the AUH that would have led to this prediction. Now, it may just be one of those things that "happens to happen", a fundamental mathematical property like the distribution of primes or the absence of odd perfect numbers. Self-aware subsystems just mathematically turn out to only be possible in a very tiny region of parameter space. Now, you might be able to make the argument that "tiny" is not well defined, that there is no natural length scale for judging parameter ranges. Tegmark could as easily have zoomed in on the appropriate region of his graph and shown a huge, enormous area where parameters could be moved around and life would still work. However I think there is a more natural way to put the question, which is, what fraction of computer programs would lead to simulated universes that include observers? And here, if we follow Tegmark's ideas, the answer appears to be that it is a very small fraction. (Of course, you still need to use your own judgement to decide whether that is "tiny" or not.) In a way, then, these two questions are both related, and perhaps the same. They both ask, why so few observers? One question looks around the interior of our universe, and the other looks at the set of all universes. In each case, it seems that intelligent life is terribly uncommon. Hal Finney