Meteorologists strive to predict the weather. Network scientists develop complex algorithms to predict the spread of disease. Might it also be possible to predict the emergence of scientific discoveries? If the answer is “yes,” what are the benefits—and pitfalls—of the ability to do so?