… to predict the present: an excellent short paper from a few days ago by Choi and Varian. This is incredibly interesting (and as you will read, working pretty well even in its simplest form!); I expect quite some new research to follow up on their call.
