Dubious. I’d like to read their papers and findings.
Sounds slightly like a funding exercise. (although arent they all :)).
1: Games are big ( 6 billion industry etc), we should tap this with research money for potential gaps in the market.
2: Neural networks great buzz word to investors and government bodies. Sounds like something from the terminator or the future.
Unfortunately, I wouldnt hold out on NN being good for this purpose. Typically they need training on some training set, I would be very keen to hear what their training set is for an online FPS, sounds abit too random and dynamic to be any use.