Precipitation extremes have impacts across a wide swath of natural and human systems. There has been quite a lot of work to understand how human activity (greenhouse gas emissions) have impacted extremes, but huge uncertainties remain. The representation of cloud processes, and resulting extreme precipitation, in climate model simulations is a huge source of uncertainty. A key uncertainty originates in the strong dependence of cloud processes on model resolution. Understanding, and ultimately improving such resolution dependence, would go a long way toward reducing uncertainty. We describe results from several recent publications that document and explain an existing resolution dependence associated with the intensification of updrafts. Based on this information, and some recent results, we argue that this resolution dependence ultimately can be traced back to the 'scale separation' assumption that is an intrinsic part of contemporary climate model design. We argue that the community should consider a new class of physics parameterization that simulates the spectrum of updrafts across scales ranging from convective to nearly global scales.