I will speak about a new research project called 'Evolution as an information dynamic system', which involves collaboration between four universities in the United Kingdom. This is a three year project started this year, 2010, and its aim is to develop new understanding of information dynamics in evolution and biology. In particular, we are going to derive new optimality conditions for some evolutionary operators, such as mutation and recombination. Evolutionary states will be represented by probability measures on the space of genetic sequences, and different operators produce different evolution of the states. We define the optimality conditions for evolution based on the maximisation of utility (or fitness) of information principle. The optimal evolution in this sense achieves the shortest 'information distance', and it can be different from an evolution optimal in another sense, such as the shortest convergence time. We argue that the former achieves a better adaptation of organisms living in a dynam ic environment. I will present several early results related to the optimisation of mutation rate parameter. I will review these results in the light of the classical theories of adaptation (e.g. Fisher's geometric model) and error threshold. Then I will outline some future theoretical and experimental work of the project.