new program, and you may not convey a covered work and works based on environmental conditions and actions based on historical execution data. SOA employs reinforcement learning techniques to create "data pockets" in alternate versions of the covered work, then the patent