Chapter 2 Section 2.1 Decision Making{0}
Section 2.1: Decision-Making Systems
Artificial Intelligence Learns To Play Music
My article “Virtual Musicians, Real Performances: How Artificial Intelligence Will Change Music” is about a company called Zenph Sound Innovations and how they are developing
complex software that models musical performances based off of old music recordings. In the simplest of terms, this company is designing artificial intelligence software that can listen to recordings of musicians, either living or dead, and analyzing their playing style in order to create a virtual profile of that artist. While the software is still in the very early testing phase and is currently limited to only a few musical instruments, the goal of this company is to be able to license and market “style plug-ins” that would allow you to hear your own music played back as if a famous musician was playing it.
The current technology only allows for Zenph Sound Innovations software to playback existing performances, note for note, yet clearer than the original recording. However, as the AI is further developed and fine tuned, they hope to be able to create original works with any “style profile” that they or a consumer chooses. Zenph’s long-term goals for this software is to be able to model instruments into the software so that a consumer could write a song on a virtual guitar using this software and then filter the song through different “style filters” to make it sound as if any number of famous guitarists are playing the consumers song.
In section 2.1 of our textbook, it talks a little about how Artificial Intelligence is slowly being integrated into our daily lives. The goal of Artificial Intelligence is to have a piece of software recognize input, learn from that input, and then apply what it learned to future inputs, including making decisions or in the case of the article, apply a vast array of musical playing styles to a piece of music. Zenph’s software is a perfect example of Artificial Intelligence. Musical recordings are run through the software which recognizes, identifies, and learns patterns of that particular artists playing style and is then able to apply what it has learned to other pieces of music.
I really thought this article was fascinating because the concept of taking a song that you have written and being able to hear how it would sound as if Jimi Hendrix or Jimmy Page or even Kurt Cobain were playing your music is just so cool. While I am not at all musically talented, I have numerous friends who are incredibly talented and they are always looking to try out new styles or technology to make their sound better or to further develop styles. Many of them like to try and replicate certain famous guitarists’ styles on their songs, but with this technology, it will make it even easier to hear how your song would sound with that added styling. I know that once this technology is perfected, many of my more musically inclined friends will be clamoring to give it a try.
Works cited:
Photo: Retrieved September 1, 2010, from: http://www.technologyreview.com/files/11634/robot_hand_x220.jpg
Van Buskirk, Eliot (2010, March 2) Virtual Musicians, Real Performances: How Artificial Intelligence Will Change Music Retrieved August 31, 2010. From: http://www.wired.com/epicenter/2010/03/virtual-musicians-real-performances/