I recently came across a Quora post, where I saw the term "Imagination Learning". It seems to be based on something called "Imagination Machines" (the link is based on a guy's work profile as of now; subject to change).
The only thing that I could find on Internet about it is this paper: Imagination-Augmented Agents for Deep Reinforcement Learning. (But I'm not sure if it's related to that concept.)
Any ideas on this would be appreciated.