In principle, at least, it should be possible to combine evolutionary modeling with a full understanding of the mass loss process that ends the AGB to predict the dependence of the remnant masses on the initial mass and metallicity of a population of stars. A detailed understanding of the mass loss process does indeed lead to a clear set of predictions about the terminal luminosity of AGB stars as a function of initial M and Z, but this does not translate to a prediction of the remnant masses because standard core mass - luminosity relations are demonstrably wrong.