Biological membranes are composed of (among other things) hundreds of different lipids, which are believed to segregate into fluid rafts, which may be relevant to processes like virus assembly. I'll talk about the spherical cow version of cells, synthetic membranes with three components (saturated and unsaturated lipids and cholesterol), which also segregate into two fluid phases. Membranes are also particularly interesting from a physical standpoint because they have both two- and three-dimensional hydrodynamic behavior ("quasi-2D"), with many strange features, such as diffusion coefficients of membrane rafts being effectively independent of their size. These quirks are characteristic of many interfacial fluids, and also appear in thin layers of liquid crystals and protein films at the air-water interface. I'll show some continuum stochastic simulations of membrane domains and phase separation, discuss new ways of measuring membrane viscosity, and suggest why some well-known dynamical scaling laws can change their exponents or even break down for phase separation in membranes. If there's time, I will also discuss how the dynamics of protein diffusion can be altered by coupling to the lipid membrane.