He said genetic programming not genetic algorithms. For the former, type and category theory are useful!