What can a single image accomplish? When it’s fed to a neural network, it can produce 100,000 additional frames set to music.

It can produce this:

The video above was created by Damien Henry, a Google employee, who used a single image and a motion prediction algorithm to artificially generate every subsequent frame. This blur-filled, 56-minute AI joyride may not be your definition of “art” but it’s certainly an indicator of how adept software is becoming at generating imagery–in this case, not just single images but a (somewhat) coherent sequence of tens of thousands of images.

As the blog Procedural Generation notes, “a year ago this would have been impossible, a year from now the cutting edge results will look quite different.”

Via: DIY Photography

Don’t Miss: This Neural Network Judges the Quality of Your Photos