Our initial goals included figuring out unsupervised learning and solving reinforcement learning, while keeping our team small.