I am curious if anyone here with a background in AI thinks that a deep reinforcement learning algorithm could learn to play a Paradox Grand Strategy game in the near future. The recent work in DeepRTS is making me think that this may be possible, but I am curious what others here think.