The paper argues that if you want to get cost estimates right for large capital investment projects, you need to unlearn conventional cost forecasting, with its century-long record of getting cost estimates wrong. Instead you need to learn behavioral science, because behavior, and not artifacts,...