Many research topics in geophysics have important implications for public policy; two examples considered here are climatology and attempts to predict future seismic ground motion. A common feature of these and other fields of geophysics is that much work on developing mathematical models is being conducted, but efforts on testing and validating these models and on evaluating their uncertainties has tended to lag. Modelers should do a better job of making clear what is known (including realistic estimates of the uncertainties) and what is not known. Public policy regarding climate change and earthquake hazard mitigation must be decided now, notwithstanding the limitations of the state of the art of modeling. Such policy decisions necessarily require compromise and the search for consensus. It is essential to distinctly separate scientific research and decisions on public policy, so that the former is conducted strictly following best practices of science without compromises or attempts to force consensus to be reached