In today's environment, we seem to have little time for either problem identification or solution implementation. What seems to work well instead (and is more fun) is to try small experiments. The idea is not to solve problems, but to try to learn in little steps.