Although it is uncertain, it is necessary to make science useful. Science is only useful if it tells you about some experiment that has not been done; it is not good if it only tells you what just went on.