For a large number of systems there are probably two principles in action. A principle of efficiency that is essential for enhancing individual survival, growth and reproduction. A principle of surprise that means that the elements or individuals of the system should be kept open to surprise, which is a principle essential for the survival of the species or the group. There is a tension between these two principles. The principle of efficiency leads to more adaptation and specialisation but this is not always the best strategy. As we have seen some models can de thought using these two principles. Suppose an economy that specialises in a given sector or product. This will increase its competitiveness and growth. But unless this specialisation is compatible with the permanence of less specialised firms, that is with firms that can survive even behaving with less knowledge or always keeping open to alternative environments, the specialisation may be dangerous because is vulnerable to surprise. In biology some features of the evolution of species may perhaps also be interpreted using this framework. Economists have traditionally considered uncertainty as a trouble. And in a sense they are right. Dealing with uncertainty means decreasing efficiency. However I am not sure that trying to reduce uncertainty and adapting behaviour to a less uncertain environment is always the best strategy. This paper gives some indication that in some cases this strategy may enhance the performance of the individuals but may endanger the group and finally the individual himself. Perhaps it is not always rational to be rational.