Why did evolution not give us a utility function that is offspring alone? Why do we care intrinsically about other outcomes, food, for example, and what determines the intensity of such preferences? A common view is that such other outcomes enhance fitness and the intensity of our preference for a given outcome is proportional to its contribution to fitness. We argue that this view is inaccurate. Specifically, we show that in the presence of informational imperfections, the evolved preference for a given outcome is determined by the individual’s degree of ignorance regarding its significance. Our model sheds light on imitation and prepared learning, whereby some peer attitudes are more influential than others. Testable implications of the model include systematically biased choices in modern times. Most notably, we apply the model to help explain the demographic transition.