Hits: 1021

Human musicians reach a human heart.
What heart can a machine reach?

 

by Eric Keller, December 2017

Traditionally, music has been the unique resort of humans.

But a few years ago, computers started to invade this territory: melodies have been composed by machines that imitate the styles of Bach, Beethoven or Mozart (see many examples on YouTube). At the present time and to ears trained in classical music, these pieces still sound strange and they are evidently full of stylistic errors. But this is only a matter of time. The next generation of AI wizards can – and will probably – do much better yet. The moment will soon arrive when even a trained specialist of Bach's oeuvres will have to listen very closely and for an extended period of time in order to tell if a given piece of music was authored by The John Sebastian or by an AI-imitation Bach programme.

This development raises anxiety all around. Music is big business and feeds many people. In tiny Switzerland, the professional association of music-producing persons (SUISA) numbers over 36'000 members. In the U.S., young and talented musicians are paying tens of thousands of dollars every year to attend top-level musical schools (Juilliard, Berklee, USC, UCLA, Oberlin, etc.). What if AI-imitation starts munching away (yet more) at the increasingly rare well-paying jobs in the music business? Will the dollars, euros, yens and yuans just keep rolling when machines start spewing out Bach, Vivaldi, or at the press of a click, Enya or Lady Gaga?

This question hits me every time I hear yet another synthesis-produced Vivaldi's “Seasons” in the elevator or the supermarket. The next time, it might only sound like Vivaldi and it might come out of a “deep learning” computer at a cost of cents per hour. Gone are the orchestras, gone are the back stage sound engineers willing to work odd hours, gone is even the experienced music producer who could read a score and knew when to raise a certain mike to provide a better profile. Left is one computer operator with a fixed script of programs to rattle off at certain times somewhere, anywhere in the world. And voilà, a programme originating in LA might play – in real time – in a dental office in Barcelona in Spain, earning someone somewhere cents on the hour. You don't believe it? I do, because I've heard such a programme myself, in a dental office in Barcelona, already two years ago.

Who makes music, how, and for whom?

All this forces us to rethink radically the whole equation of music. I am human and I assume that you are, and we all want music that is alive, that moves us and that might at times even be novel.

I've thought this through, and I am no longer worried. I think that all kinds of music will not only stay, but that it will flourish. The essential condition is that it is made by humans for humans. The way I see it, music succeeds to the degree that it connects with human needs, emotions and interests.

Consider what fills us with awe:

More examples abound, and all point to this essential point: Music is alive and strong when it reaches other humans. Play, sing and compose your way into your public's heart, and you will succeed, as long as there are still humans around. Do it really well, and you'll bring joy to hundreds, thousands and perhaps millions of other human beings.

What more can you ask for?

Money? A living? Yes, that's yet another chapter that needs to be well thought through. But it will come much more easily if you do the first thing first: reach other human hearts with your music.

 

 

1 http://www.foxbusiness.com/features/2015/05/08/4-character-strengths-sustain-u2s-success.html