3. Machine Art

In this chapter, we introduce some of the concepts that enable deep learning models to seemingly create art, an idea that may be paradoxical to some. The University of California, Berkeley, philosopher Alva Noë, for one, opined, “Art can help us frame a better picture of our human nature.”1 If this is true, how can machines create art? Or put differently, are the creations that emerge from these machines, in fact, art? Another interpretation—and one we like best—is that these creations are indeed art and that programmers are artists wielding deep learning models as brushes. We’re not the only ones who view these works as bona fide artistry: generative adversarial network (GAN)-produced paintings have been snapped up to the tune ...

Get Deep Learning Illustrated: A Visual, Interactive Guide to Artificial Intelligence now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.