June 4, 2016 weblog
Google AI team looks to music makers to make some noise
(Tech Xplore)—Where art meets AI—or, more specifically, where AI takes on the challenge of creating "art"—is now in the active stage within the Google project tagged Magenta. And now Magenta, with all its dedication and focus on machine learning, has a melody to show for its efforts.
The 90-second curiosity is the first song created by the project, and it was created via a neural network, given four notes to start, noted CNET.
Thankfully, the tune is easily forgettable, if you can imagine a dumb, not very smart or human-like, machine to emit music, barking out notes with no tints or shadings. The drum sounds are good but those are human overlays.
That, nonetheless, does not seem to be the point. The BBC's Alexander McNamara, Editor, Science Focus, put the song in a perspective that would be welcomed by many Magenta supporters.
"Their first attempt is definitely something, maybe not compelling but it is a start we suppose, and given it had only four notes to play with (the drums were added afterwards) it's not totally awful:"
McNamara also pointed to Google's broader motivations. "Google hope that this will allow developers and musicians to use Magenta to create something new and original with the algorithms."
His comments suggest that the song effort is not harboring delusions of replacing human composers with machine computers, that it is not either/or but and/and, with potential for interesting results.
"They also hope that in the future musicians will be able to play along with music generated by Magenta in a live setting."
Who knows? Marc Schneider, contributing editor, Billboard, similarly had the reflection that "A goal of Magenta is to create an open-source tool to bring together artists and coders looking to make art and music in a collaborative space. "
As part of the initiative, he said, "Google will provide audio and video support, tools for MIDI users and platforms that will make it easier for artists to connect with machine learning models."
This reflects one of the key goals of the Magenta team and that is to build a community of artists as well as machine learning researchers and other techies.
Again, who knows what can come of this? Posting on Wednesday, Douglas Eck, a research scientist working on the Magenta project, commented that "We don't know what artists and musicians will do with these new tools, but we're excited to find out. Look at the history of creative tools. Daguerre and later Eastman didn't imagine what Annie Liebovitz or Richard Avedon would accomplish in photography. Surely Rickenbacker and Gibson didn't have Jimi Hendrix or St. Vincent in mind. We believe that the models that have worked so well in speech recognition, translation and image annotation will seed an exciting new crop of tools for art and music creation."
One thing is certain; the message on the wall is something like Make Some Noise. "Once we have a stable set of tools and models," Eck said, "we'll invite external contributors to check in code to our GitHub. If you're a musician or an artist (or aspire to be one—it's easier than you might think!), we hope you'll try using these tools to make some noise or images or videos… or whatever you like."
© 2016 Tech Xplore