Inceptionism: Google Artificial Neural Networks

Interesting imagery could become a form of art on its own

[engine]

Google artificial neural networks seem to have turned out to be creative in their visualizations. Who'd have thought it. Whilst some are not the best, others are looking very interesting, and I wonder if it could turn out to be a form of art on its own.

We train an artificial neural network by showing it millions of training examples and gradually adjusting the network parameters until it gives the classifications we want. The network typically consists of 10-30 stacked layers of artificial neurons. Each image is fed into the input layer, which then talks to the next layer, until eventually the “output” layer is reached. The network’s “answer” comes from this final output layer.
Inceptionism: Google Artificial Neural Networks [googleresearch.blogspot.com]

So here’s one surprise: neural networks that were trained to discriminate between different kinds of images have quite a bit of the information needed to generate images too.

Here are some interesting examples
[3.bp.blogspot.com...]
[4.bp.blogspot.com...]

Fascinating!

[Leosghost]

Immediately made me think of Robert Venosa..
Also some of the Adobe plugins by Topaz..
Artificial Neural Networks..an extremely interesting subject..
I don't think I would use the word "generate" here..it implies creativity from no source material ( "genesis" ) ..but the process of extrapolation and interpretation by the ANN , and the results of the "optimise with prior" after many "recursions" are very interesting..
[photos.google.com...]
Also brings to mind..the phrase.."Do androids dream of electric sheep ?"

and "Will I Be Sky ?"..

Based upon what it "imagines". The ANN may well get bored with responding to the humans, and choose to turn it's eye inward..or develop a craving for peanut butter and icecream and other munchies, or get locked ( metaphorically ) in the refrigerator of it's own conciousness..

As long as it passes the bong first..;)