The OpenAI Gym and Universe toolkits are fantastic for testing out and working with reinforcement learning techniques, allowing users to quickly test their algorithms in a variety of environments and games.
I believe they would be a fitting addition to the packages that come preinstalled due to their rather widespread use and useful features, as well as making it easier for novice users, such as myself, to easily transition to using FloydHub.
Would love y'alls thoughts on this
