How Artificial Intelligence is changing music

As I write this essay, a rather incredible experiment has happened, which combines data science with music. Beethoven’s unfinished 10th symphony has just been completed, 194 years after the composer wrote it, by an Artificial Intelligence program as part of an effort at Rutgers University. This sort of development in the world of creativity is critical for two reasons. One, for arguments about how “machines can take over everything but the creative strain and original thought”. Two, for the promise it seems to hold for the directions in which the creation and preservation of culture can go.

Coming to the first — that of creativity and the hard-to-articulate human factor, purists have often argued that original thinkers and composers — the Carnatic music Trinity or the Western old masters — are irreplaceable and their works will continue to endure hundreds of years. Their music has convinced us of that, surely, by now. Is it that easy for a machine, which uses past data patterns to construct or simulate where the composer might have gone, to create similar compositions and for those to endure too? Time alone will tell. But it is now becoming possible for AI programmes to replicate frequency and data patterns to a phenomenal degree of aesthetic beauty. The creative intelligence factor, one that has long been held as the sole enclave of human superiority over machines, is one that will need further scrutiny and redefinition.

Coming to the second point, certainly this development holds great potential. The harnessing of technology to be able to co-create freely across cultures and geographies (something that got borne out by the pandemic), to use machine learning to add layers or even bits of compositional material to what we do, and the ability to preserve this data for future generations to use — all this just got bigger.

At a frenetic pace

So where does this leave today’s musician? Somehow, during the pandemic, we are seeing the re-emergence of the old rules. There is a desperate scrambling for share-of-voice, and social media is crammed with musicians who are uploading performative and compositional content at a pace that is too frenetic for anyone to consume all of it reasonably. Everyone is in a hurry to convince audiences that they are still performing, are still relevant — and in so doing, they are unfortunately providing us more of the same.

Now, factor in the notion that a machine and an AI program can possibly create something more palatable, and perhaps at a fraction of the financial and emotional cost of having to deal with people? Where would that leave us?

I posit two solutions, and I am happy to argue these with musicians and creators who may think otherwise. For the human creative intelligence to truly achieve its potential, cultural creators will now have to stay humble and agile. They will need to revisit their art in a deeper manner and strive to break their own patterns and search for something inventive that a data prediction mechanism cannot replicate.

Just a week ago, Arvo Part, the great minimalist Russian composer, cropped up in a conversation I had with musician Sid Sriram. In a world that was edging towards ornate orchestration, Part aimed at the wholeness of single notes and an almost monastic reverence for silences in his music. This broke the clutter of the time and continues to teach and inspire. Listen to the ballads — the John Legends and the Adeles will continue to impress even after all the clamour dies down for the same reasons — the simplistic beauty of melody with minimal orchestration and noise.

Unfortunately, we are in the midst of a sound pandemic where clutter is deemed more trendy.

The second solution will be to embrace co-creation and collaboration more freely. With the possibilities that have opened up now, thanks to even musicians in the remotest parts of the world embracing the Internet, this would be a great time to create more work with different flavours, while striving to strike a distinctive note.

I am by no means saying that the era of the individual creator is over. I am actually arguing for the opposite, though with caveats. Unless we understand what it truly means to break predictability, outdo ourselves in our ability to be inventive, an AI program does run the advantage of beating us at this game.

So, yes, creative intelligence is still very much the human advantage. But the bar has just been raised higher.

The author is a pianist and

music educator based in Chennai.