As in the case of super cooled water, the metastable state cannot survive long, because things tend to come to thermal equilibrium. Thus, eventually water becomes ice and a transition from one phase (liquid) to another (solid) would have occurred. In this case energy is released. In the case of the universe, a phase transition occurred and was symmetry breaking, in the sense that forces and particles started to differentiate (break the symmetry that existed before). As water when it freezes doesn't always make a perfect ice crystal, but voids, cracks and other defects are created, so in the phase transition of the universe defects such as mono-poles (an isolated magnetic charge that has never been observed but has been postulated to have been produced in quantity in the Big Bang), cosmic strings etc. were produced. Our universe is just one part of the entire universe, like a small patch of ice. In that patch of ice there are few defects (most of the defects are between patches), and in the universe we observe there are few mono-poles.

Bubble Universes. The fact that the visible universe came out of a small patch of the entire universe means that we live in one of the many existing universes. We can think of the original universe as a collection of bubbles and we live in one of the bubbles. Because each bubble at the beginning was not in causal relation wit the others, we will never be able to know about the other universes (because in order to send a light signal from one universe to the other it would require light signals to travel for a time longer than the age of the universe).

During the inflationary era, the universe expanded in size by many billions of times between 10^-35 and 10^-33 sec. after the Big Bang.
Inflation resolves two of the puzzles left by the original Big Bang theory.
The first problem is about the smoothness of the universe. Measurements done today indicate that the cosmic background radiation is smooth over distances that couldn't have been in causal contact. In other words, these regions of space were too far apart from each other so a beam of light would have taken longer than the age of the universe to travel from one place to the other. Because in the inflation scenario the universe expanded so rapidly, a small section of the universe would have grown from a size of 10^-25 cm, much smaller than the nucleus of an atom, to the size of the present universe. In such a case, it is clear that our entire visible universe originates from the same patch.

The second problem to be addressed by inflation is the flatness problem. Measurements done today show that the universe is not curved, but rather flat. In the inflationary universe, this is explained by saying that the size of the visible universe grew so large that it looks flat (imagine that our universe is a small patch on the surface of the balloon - it looks almost flat at that point).

Dark Matter: Astronomers can compute how much luminous matter there is, that is, matter that is visible with telescopes. However, this amount of matter is not enough to explain, for example, the existence and stability of many galaxies. In this case, the motion of stars in the galaxy is not consistent with the amount of mass that is producing light: there must be more mass, by a factor at least 10. So, it is fair to say that most mass in the universe is not seen, and only about 10% or so is visible. The rest is non-luminous material, such as clouds of gas and dust, and planets and stars to dim to see. Even postulating that there is 10 times more non-visible matter than luminous matter, we end up with a total budget of mass of the universe that is
10 to 20 % of the one necessary for having a flat universe. Of course, this extra amount of matter would be required if we believed that the universe is "flat", i.e. just in between a closed universe (endowed with a closed or positive curvature) and an open universe (forever expanding, with a negative curvature). Some astrophysicists have postulated the existence of yet undetected objects (from massive astronomical objects to new elementary particles) to explain this difference. This is still an open problem.