# Does a complete theory of quantum gravity require anthropic post-selection?

Does a complete theory of quantum gravity require anthropic post-selection? Certainly the black hole complimentarity and causal patch conjectures highlights the essential role of observers, at least in the asymptotic future of their future timelike trajectories. Does the measure problem in quantum gravity cosmology suggest that a "global god's eye view" of the universe might be an incoherent fiction? However, if observers are essential, don't we have to post-select to those states containing the observer of interest?

It might be suggested the S-matrix of string theory in a superselection sector with only a finite energy difference from the vacuum over a BPS background provides a counterexample, but can an asymptotic future state of noninteracting Fock space particles really support the future trajectory of an observer?

-
Anthropic? I don't think so. Probably experimental, that's it. –  Dimensio1n0 Aug 23 '13 at 19:51
It's fun to speculate about stuff like this, but I don't think it's realistic to imagine that anyone can say anything definite about characteristics of a theory of quantum gravity today. String theory and LQG both smell like failed research programs at this point. We don't even know what fundamental principles go into a theory of quantum gravity, e.g., whether it should be unitary or nonunitary. –  Ben Crowell Aug 23 '13 at 23:38
can an asymptotic future state of noninteracting Fock space particles really support the future trajectory of an observer? You need interactions to get an observer. For example, this is what stops us from having an observer made of massless, noninteracting particles all traveling along parallel geodesics -- which would be paradoxical, because you can't define a Lorentz boost for an observer moving at $c$. –  Ben Crowell Aug 23 '13 at 23:41

Quantum gravity is a gauge theory. The gauge in question is diffeomorphisms. This includes timelike diffeomorphisms. Quantum gravity states have to be invariant under timelike diffeomorphisms, no? Any postselection criteria also needs to be invariant under timelike diffeomorphisms, no? So, postselection criteria only at the conformal boundaries, no? Basically, only consider those diffeomorphisms which don't act upon boundary points. A boundary condition on diffeomorphisms.

If there is anthropic postselection, diffeomorphism invariance requires that to happen someplace on the conformal boundaries of spacetime. Could it be at the past boundary at the big bang? no. So only the future conformal boundary of spacetime then? This supports the final anthropic principle of Tipler and Barrow. Intelligent life will last forever and will never die out!

-
" Could it be at the past boundary at the big bang? no." Care to explain? –  Doug Packard Nov 26 '12 at 18:00
Quantum gravity states have to be invariant under timelike diffeomorphisms, no? You lost me here. Just because the theory is invariant under a certain group, that doesn't mean its solutions have to be. For example, Maxwell's equations are invariant under rotations, but that doesn't mean we can't have a plane wave. –  Ben Crowell Aug 23 '13 at 23:47
Any postselection criteria also needs to be invariant under timelike diffeomorphisms, no? So, postselection criteria only at the conformal boundaries, no? How come? E.g., if string theory is right, then we could apply the anthropic principle to the way the extra dimensions are curled up topologically in the vacuum. That's a purely topological criterion, so it's automatically diffeomorphism-invariant. You can apply the criterion at any point in space. –  Ben Crowell Aug 23 '13 at 23:49
This supports the final anthropic principle of Tipler and Barrow. Intelligent life will last forever and will never die out! Tipler proposed this in 1986. Isn't this hypothesis overtaken by events? See arxiv.org/abs/astro-ph/9902189 . In general, Tipler seems to be a complete kook. –  Ben Crowell Aug 23 '13 at 23:51