How long can it take for GAN (Generative adversarial networks), given its current state of research with regard to its state of the art development, to yield more efficient results in terms of NLP performances, while the major advantages of NLP may be improved by quantum computing?