Grappling with our desire to understand nature, we construct models of the specific systems that we wish to study. Unsurprisingly, such models are generally highly tailored to the system of interest. But are all these models really that distinct? Or, could there be a universal model that can somehow describe the behavior of any system we could think of? On page 1180 of this issue, De las Cuevas and Cubitt (1) venture out to weave ideas from physics and computer science in an attempt to answer this question for all classical spin models.