The complexity of clouds not only puzzles the casual observer but also challenges our understanding of the climate system. The processes that make clouds so hard to capture span a wide range of scales, from the formation of cloud droplets on aerosol particles to shifts of cloud regimes with large-scale conditions. This multiscale nature of clouds translates into persistently poor constraints on the cloud-mediated aerosol forcing and cloud feedbacks. We will focus on intermediate scales, where shallow cloud fields organize into mesoscale patterns. The evolution of these patterns is shaped by interactions with both smaller and larger scales. I will present examples of how this interplay leads to non-linear sensitivities, abrupt changes of patterns and memory effects. A particular focus will be on the role of aerosols in shaping all three. We will close by discussing implications of these results for numerical simulations and the analysis of satellite data.