Tuesday, August 15, 2017

Music Machine


Reading this article from AAAS, it suddenly occurs to me that the common argument about machines vs. humans in music is flawed. As I note in my last post about Liza Lim’s “How Forests Think,” we naturally hear the sounds of the forests as music – it is a human response. We turn the sound into music because we’re musical creatures that come from the forest. That small insight, applied here, provides the key to understanding our future role with technology and the arts.

Music isn’t necessarily in the making – deciding which notes and rhythms go together – but in the human response to organized sound, the enjoyment and participation of in the making and listening. That makes AI seriously less threatening. Will machines make music equal to that of humans in creativity, complexity, even spiritual depth? Sure! Will it matter? Nope.

Because, after all, what’s the point?

We ultimately will be able to teach computers how to respond to music like we do. And they will be capable of making music on par with a human composer or performer. But the musical experience itself is human and the need for humans to make music and to appreciate the music that is made, will never, ever go away. Machines are just us – an extension of us and an extension of that musical impulse.

This idea actually extends a slim hope for the economics of music – a weakly floating board to cling to amid the shipwreck of the music industry. As far back as Marshall McLuhan, students of culture have noted that it is only when an object’s use becomes obsolete that we, as a culture, begin to fully appreciate it. The value of paintings and appreciation for the skill of the illustrator increased with the advent of the photograph. Tape and LPs became highly prized when that had been eclipsed by CDs and digital downloads and streaming. We are seeing a rather sudden appreciation of the physical, printed book, separate from the books we read online (the scores of George Crumb and the rise of the graphic novel are just two examples that spring to mind). Once its usefulness is eclipsed, the human achievement of the old medium becomes the thing itself, a source of admiration and appreciation.

That’s where we’re headed with the use of live performing musicians for entertainment. Already it has become a kind of status symbol – only the poor wedding will have no live music. Musicians now can make a living as cover bands, some focusing on famous old acts like Led Zeppelin and the Beatles, but some just cranking out the tunes at parties, human juke boxes – whether imitating artists or rendering them in a more personal style, doesn’t matter, so long as they’re not recordings but flesh and blood, communicating the rhythm.

It is common, too, to see live performance transcriptions of works never designed for live performance, like Frank Zappa’s “G-Spot Tornado” at the 1990s Yellow Shark concerts, or the Beatles’ “Revolution 9” by Alarm Will Sound, or the endless MIDI scores composed for video games performed by orchestras. The subtext there is clear: music is better with living human beings making the sound.

Not better because humans do it better; simply better because human.

It will be no surprise when a future sophisticated robot (like Data in “Star Trek: The Next Generation”) can play Paganini perfectly. But because it’s no surprise, it will also be less valuable than a human achieving the same goal. A player piano can execute rhythms and counterpoints impossible for a human to play. The only reason to care is because a human dreamed up those impossible compositions.

Music is part of our evolution into the creatures we are now and technology is part of that. But the connection between our music and our biology is deep, and it exists now with or without technological aid, offering us immediate union with one another and with the world. We will carry that forward, come what may.

Let’s admit that the Turing Test is merely a landmark on the horizon, and that there will come a day when we literally can’t tell the difference between humans and machines, or the part of us that is human and the part that is machine – no matter how well we know the hybrid. That day is a long way off, but when it comes, the confusion will be best expressed in a song.

--C.
August 15, 2017
theandofone.blospot.com

No comments:

Post a Comment