The era of black hole merger observations has finally begun, and modelers are presented with several big challenges. Black hole mergers happen either via binary evolution or via dynamical interaction and capture. I will discuss a potentially important dynamic sub-channel, namely black hole mergers within AGN disks. I will outline a simple 'Drake equation' to parameterize the likely range of rates expected from this channel. I will discuss the production of stellar mass black holes that are overweight compared to their Galactic cousins and I will discuss some of the problems that confront this model, including black hole spins and merger kicks. I will outline how LIGO and LISA can strongly constrain this channel with future observations.