Widget Image

Artificial Intelligence and art: How Google taught technology to paint

By Benjamin Stecher

The pictures on this page were all designed by a machine learning algorithm written by the team at Google Brain. Art is the latest frontier AI companies are trying to tackle. Slowly but surely AI is becoming better and better at more and more things.

Photo courtesy: Google Deep Dream

The program, called Google Deep Dream, learned how to identify objects by scanning millions of photos pixel by pixel. First, it learned how to distinguish between all the colors and shades, then it scanned for border areas between objects. Over time it learned how to separate one object from another and built up a catalog of every object from every picture it had scanned. It then figured out how to arrange and categorize objects that had similar characteristics and learned how to recreate random composites of those objects. Finally, when prompted it displayed a random set of those pictures over a landscape template.

Google is developing a variety of tools to help speed machine learning along. One of the latest is called AI Experiments, a website where you can play with an assortment of programs that use various machine learning techniques. Warning, by playing you are actually helping the machine get smarter, making you personally culpable when they eventually rise up and take over the world. But don’t worry, it probably won’t be as bad as you think.

Photo courtesy: Google Deep Dream

The Ownership

There’s a question I hesitate to ask. Who owns the rights to these pictures? As far as I know we have a right to claim ownership over the things we design, create, or produce. But what happens if we design something that designs something? Who owns that second thing? This may seem trivial right now, but what happens when AI starts creating software, publishing news stories, creating diagnostic tools, etc.? If the people who wrote the code that created AI also own what their AI creates, well eventually these tech companies are going to make an AI,or have enough different AIs, to make anything imaginable better than any human can. Those few companies would thus, given enough time, own everything.

Photo courtesy: Google Deep Dream

Note: Once the right algorithm is assembled it can take that program just a few weeks, or depending on the skill possibly a few hours or even minutes, to master what takes us a lifetime.


Benjamin Stecher is a Writer for Futurism.

This article was originally published on The World Economic Forum.

Featured image: Fortune.com