Predictions From Eternal Inflation

We investigate the physics of eternal inflation, particularly the use of multiverse ideas to explain the observed values of the cosmological constant and the coincidences of cosmological timescales. We begin by reviewing eternal inflation, the multiverse, and the resulting measure problem. Then follows a detailed study of proposals to solve the measure problem, both analytical and numerical, including an analysis of their predictions for cosmological observables. A key outcome of this investigation is that the traditional anthropic calculations, which take into account the necessity of galaxies and heavy elements to produce observers, are redundant in our framework. The cosmological coincidence problem, the seemingly coincidental equality of the timescales of observation and of vacuum domination, is solved for the first time without appeal to detailed anthropic assumptions: very general geometric considerations do the job automatically. We also estimate a 10% likelihood that evidence for eternal inflation will be found in upcoming measurements of the energy density of the universe.Encouraged by this success, we go on to construct a modified version of the light-cone time measure which has conceptual advantages but also reproduces the phenomenology of its predecessor. We complete our study of the measure problem by noting that for a wide class of proposed solutions, including the one developed here, there is an implicit assumption being made about a catastrophic end to the universe. Finally, as a by-product of this research program we find geometries which violate some of the accepted common knowledge on holographic entropy bounds. We point this out and conjecture a general result