Read a neat paper recently called Dissolving the Fermi Paradox that points out we’ve been doing the math wrong. The basic idea is that given the uncertainty to the parameters in Drake-like equations (e.g. number of habitable planets) we should actually be calculating distributions rather than averages. When you do this it turns out to not be terribly unlikely we are alone.

One of those things that seems obvious in hindsight but surprising that no one wrote this down before.