Can machines create art? Google-powered Magenta project aims to find out

Can machines (you know, those dead hunks of nuts and bolts) actively produce what humans consider art using nothing but mathematics?

This is the question that a new research group dubbed Magenta hopes to answer using Google technology as a backbone. Unveiled at Moogfest by Google Brain’s Douglas Eck, the study hopes to understand the extent of computers’ creativity using one of Google’s own technologies — TensorFlow.

The tech, which is central to Google’s DeepMind project as well, uses flow graphs to achieve “scalable machine learning.” If you fell asleep during that sentence, it effectively turns mathematical information into, well, art. At least in this particular implementation.

The team will experiment with music as the medium, and the project hopes to “create a system that could give a listener ‘musical chills’ with entirely new pieces of music, on a regular basis, as they sit listening to computer-generated music from the comfort of their couch at home,” the researchers told Quartz.

The thesis might sound obvious (music is, after all, mathematically based), but machines don’t exactly think like human beings.

“[A]rt and music is not just about generating new pieces,” Eck explains. “It’s also about drawing one’s attention, being surprising, telling an interesting story, knowing what’s interesting in a scene, and so on.”

According to The Next Web, there’s also an app planned to showcase the musical endeavours, and possible genius, of the project to the public, but don’t expect computers to rival Beethoven or Lady Gaga just yet.

Feature image: idugh via Flickr

Andy Walker


Sign up to our newsletter to get the latest in digital insights. sign up

Welcome to Memeburn

Sign up to our newsletter to get the latest in digital insights.