How a Neural Network Can Generate Reasonable Pictures

The idea is to take billions of images, say from the web, for which corresponding captions have been provided. Then one progressively tweaks the parameters of the neural net to make it reproduce these captions when it's fed the corresponding images. At every step, we see in effect what the caption would be. And we try to go in a direction that leads towards a picture with the caption we want. Or in other words, progressively try to get a picture that's of what we want. If everything works correctly, the final pictures will all be of what one asked for - a cat and a party hat.