A company trying to patent prediction future behavior based on game play using emotive evolution probably knows that the OASIS HumanML project developed the concepts for this over five years ago. The HumanML work relied on several earlier works in games simulations and psychological theory. The list discussed this in depth and documented the designs.
It seems that despite the jeers from gallery and certain TLA consultants, the HumanML project was right.
I am not familiar with the essential claims of the patent submission but I'll be surprised if there is much there that is novel or doesn't rely on prior works. The patent office should be wary of this one.