Abstract
Inflation, or exponential expansion of ~1030 between 10-36sand 10-32s, was invented to solve the problems of the horizon, the flatness of the Universe, and the non-existence of monopoles. As a bonus, the model gives a natural origin to the initial density fluctuations, which are at the root of all the structures in the Universe today: indeed, expansion is so strong that when a virtual particle-antiparticle pair is formed in the vacuum, causally connected regions are suddenly disconnected: particles can no longer annihilate each other, and density fluctuations are created. Planck's latest observations of the cosmic microwave background have confirmed the role of inflation in structure generation (as topological defects are eliminated). Early theories of inflation were based on phase change and symmetry breaking. Today, chaotic inflation no longer requires a phase change. What's needed is a scalar field that evolves on aslow roll. This field resembles that of a quintessence, except that the amplitudes and time scales are completely different. Current inflation models suffer from the problem of eternal inflation. It's hard to stop inflation everywhere. It can be stopped just for our Universe, with heating and particle creation, but it continues in multiple Universes. It's a fractal inflation, with a multitude of landscapes, all the more so in string theory as space has a large number of extra dimensions. What are the alternatives? Some are developing cyclic, or ekpyrotic, models, based on colliding branes, or cyclic, with rebound, like Turok and Steinhardt. A critical test would be the observation of primordial gravitational waves, which would prove inflation. This test has been attempted with B-symmetry polarization modes in the cosmic microwave background, but has not yet yielded any results. Future experiments may be sensitive enough to make a decision.
