This talk is a collection of my own failure stories when trying to find, use, or extend existing memory models. My intention is to point out situations in which we, as an already established community, fall short and what I expect from the next decade of memory models research.