In the past two years, our ability to compute scattering amplitudes relevant for hadron collider phenomenology increased dramatically. I will review advances in computational techniques that lead to this breakthrough and discuss a few lessons from recent one-loop computations.