This is an exceedingly good article on the current state of theoretical crypto. I would have liked to see a little more on the new things that we can do with lattices - e.g. obfuscation or homomorphic encryption - because that's really what is motivating a lot of new research.