Abstract:  Galaxies are a complex cosmic stew of stars, gas, dust, and black holes; as a result, their inferred properties are often loosely constrained and exhibit significant degeneracies. This means that the accuracy of simple properties like total stellar mass and recent star formation rate (SFR) depends on the accuracy of many second-order properties, including star formation histories (SFHs), stellar metallicities, dust properties, and many others. The high-dimensional models required for accurate measurements have been stymied until recently due to the substantial computational and statistical requirements. Here, I introduce Prospector, a new Bayesian inference tool designed specifically to deal with these complexities. I re-assess inferences of galaxy stellar masses and SFRs in the distant universe using a comprehensive new 14-parameter physical model, finding that galaxies are ~60% more massive and have ~40% lower star formation rates than implied by classic techniques. These new inferences lower the observed cosmic star formation rate density and increase the observed buildup of stellar mass, finally bringing these two metrics into agreement at the factor-of-two level at 0.5 < z < 2.5. I finish by highlighting how solving this problem unlocks a treasure trove of fundamental and newly answerable questions surrounding the formation and evolution of galaxies in the era of JWST, LSST, ELTs, and beyond.