>>13279180train 2 neural nets, first one takes noise and outputs images, second one takes images and classifies coom
with a database of real coom train the second model the difference between real and fake coom, then train the first model on what the second model thinks is coom. Do this 10,000 times and the first model can create artificial coom
