Communicating the strengths and limitations of climate modeling to those outside the field of climate science is a formidable challenge. The nuances of scientific language can be lost in the translation to natural language when climate predictions are presented to a general audience. This loss in translation can lead to misinformation and disinformation that hampers a rational response to the climate crisis. Even simple terms like “model,” “data,” and “prediction” have many different meanings depending on the context. Anytime we talk about the future, we are using a model. In climate science, we might think we are dealing with data from the past, but often this is processed data that is produced by analysis models applied to raw data. The word “prediction” can mean a range of things, from unconditional prophecies to conditional projections.