The circular component of motion of BLR clouds is a simple consequence of gravity, but the turbulent vertical component of motion has to be maintained against dissipation losses, and an outward transfer of angular momentum is necessary to get inflow. We now know that the viscosity needed to drive the outward flow of angular momentum in accretion discs, and hence the inward flow of matter, is the magneto-rotation instability (MRI) (Balbus & Hawley 1991). Over the last decade increases in computing power and the development of more sophisticated programs by several groups have allowed increasingly detailed magneto-hydrodynamic (MHD) simulations of accretion flows (e.g., Hawley & Krolik 2001, Proga 2003, Anninos, Fragile, & Salmonson 2005, Ohsuga et al. 2009, Shafee et al. 2008), and references therein). In these models attention has been focused a lot on the low-density outflows. Because emissivity goes as the square of density, the emission is dominated by the high-density. To my mind what is impressive about every single one of these models is that, despite the different modeling approaches, the velocity fields of the high-density material all match the velocity field inferred for the BLR. I.e., the dominant motion is Keplerian, but there is substantial turbulence, and a significant inflow. This is very clear when one watches movies different groups make of their simulations. I believe that these simulations give us a physical basis for what we have deduced from BLR observations: the BLR is the material accreting onto the black holes. It was indeed noted a long time ago that if the BLR is inflowing it can provide the necessary mass flux for powering the AGN (Padovani & Rafanelli 1988).