The Machine is Dreaming

http://googleresearch.blogspot.com/2015/06/inceptionism-going-deeper-into-neural.html
https://photos.google.com/share/AF1QipPX0SCl7OzWilt9LnuQliattX4OUCj_8EP65_cTVnBmS1jnYgsGQAieQUc1VQWdgQ?key=aVBxWjhwSzg2RjJWLWRuVFBBZEN1d205bUdEMnhB
Google gives a neural network acid and asks it to return images. It combines dogs and fish, plants and buildings, birds and insects into a single images to optimize it's memory usage.

Fucking trippy.

>>13
I doubt their algorithm is capable of building a spatial model of the content of a book, then comparing that model against the real world. I.e. if books describes a "large to the north-west", there should be understanding what is "large city", while north-west id declared by tricky context.

Most important books (like math and physics books) rely heavily on visual cues and illustrations. These hierarchical networks are useless without an algorithm to connect and guide them towards some goal.