Well, they are different approaches that are best suited for different kind of problems. Genetic algorithms are useful when you can only check how good is a solution with a fitness function. (e.g: evolving a controller that makes a robot walk, measuring just how far it traveled. There are many fascinating examples on YouTube) Reinforcement Learning algorithms can be more useful when you have frequent rewards during a run of the algorithm, in that case you can update your behaviour online more frequently.
RL has become a very important field of AI, with the advent of Deep Learning, and it’s a very interesting topic. Many problems in robotics and other field are well suited for RL, so usually RL is a better approach. (DeepMind and OpenAI are focusing great energy on it)
If you are interested I suggest you to look at Evolutionary Robotics, by Nolfi & Floreano, a very nice book on genetic and evolutionary algorithms.
For RL, there are many introductions online, a good reference book is Reinforcement Learning by Sutton & Barto.
