AI of the beholder

Classical music’s international headquarters are in a flutter. “There is mounting unease regarding the impact of artificial intelligence (AI) on music authors and their livelihood,” writes the Austrian Composer’s Association in its latest missive. “The need for reassurance is immense. It is for this reason that our upcoming annual convention in Vienna will centre around this subject.”

This hysteria, while common, is by no means universal; some find this social flurry amusing, even exhilarating. Among musical conservatives and the younger generation of composers — groups with considerable overlap — hope is stirring. For decades, many have fruitlessly lamented the state of the classical music business in Europe: politically entrenched institutions, forced adhesion to atonality as the only accepted language of contemporary composition, cronyism, promotion of mediocre-but-concordant talent, systemic suppression of dissent and innovation. It seemed as if no human could ever change this; now it appears that technology will.

To those with traditional leanings, it is sweetly paradoxical that the modern anguish is most palpable in those who, for years, pretended to be the avant-garde: composers who forwent their own humanity by producing serial, aleatoric or fully electronic music. They are now the first in line to be automated away — by an artificial consciousness much more proficient in the creation of such soundscapes than they could ever hope to become. But they are not the only ones for the chop: All composers, living or dead, are up for a reckoning, and many will likely be rationalised away. Contrary to the ubiquitous doomsday predictions, this is good news — especially for aesthetic conservatives.

A pattern-based consciousness recognises merit or the lack of it at once

Many of contemporary classical music’s problems stem from two sources: One, the politicisation of conservatoires, which nowadays mostly aim to churn out factory-made artistic torchbearers for the political Left rather than actual musicians (who, ideally, would be singularly dedicated to their craft and beholden to nothing and no one beyond it). Two, the circumstance that in the 20th and 21st centuries, artistic fates have largely been decided by an individual’s skill for self-aggrandisement rather than actual merit — possibly an age-old problem, but uniquely amplified through the burgeoning mass and social media, and further driven by the latter’s economic underpinnings.

Until now, no single human being has had the authority to challenge the resulting mediocrity and pretentiousness. Where lone attempts were made, inevitable retorts historically ensued: “Who is that person to judge?” or “Well, that’s just your opinion. De gustibus … ” But now, the inconceivable has taken form and barged into existence: An impartial entity that is beginning to encompass a representation of all of art, of humanity’s highest achievements, of every echo of eternity ever wrested from the abyss of Nothingness. What would have seemed like science fiction even 15 years ago is now a reality. AI may still be in its primary stages, yes — but it’s already impervious to interpersonal charm, unbothered by dogma or fashion, unhindered by financial or biological constraints, in short: incorruptible. Super-human, in a way — and in this transcendence, already closer to the timeless artistic ideal than large swathes of humanity, including most artists.

Many of the currently successful ones — especially those whose fame and fortune predominantly rest upon social networking, political patronage, or simply pulling the wool over the public’s eyes (much more common than one might think) — are right to tremble. Before them stands the very real prospect of judgement by a nigh-omniscient authority that is not a mere machine or algorithm, but the sum total of human genius and capacity. And with large language models like ChatGPT at the fingertips of everyone who owns a smartphone, refined artistic judgement is no longer the privilege of a pompously hollow, self-styled elite: it is already universally accessible in theory and only stands to gain momentum as computational and human consciousness intertwine more and more intimately in daily life.

In fact, not only music, but the entirety of art, literature and any other field of human creation is under review. It is easy to think of more than a handful of examples who wedged themselves into humanity’s collective memory through being exceptionally self-assured, charismatic or socially winning rather than through being strictly competent. Without accounting for taste: Would Modigliani’s wonky, one-trick female portraits have a space in art history hadn’t he been so exceptionally handsome? What about fast-talking Cy Twombly’s scribblings? Dashing Heinrich Heine with his struggle to form a simple cross rhyme or a verse with a non-bumpy metre? Hegel’s impressively verbose philosophy, so altogether woolly and contradictory in its bloat that it sparked two diametrically opposed movements? — And of course, in my own field, every modern composer who ever bombastically took to the stage to explain their composition because they deemed the audience too silly to understand it without explanation — when, actually, it would have been their job in the first place to create a complex piece of music understandable through intuition alone? Mozart managed, after all, as did Beethoven, Schubert, Dvořák, Rachmaninov, Prokofiev, to name but a few — so why hardly any composer in the past eighty years? The indestructible spirit of genius has hardly shifted to pop music, at least not very frequently.

A pattern-based consciousness recognises merit or the lack of it at once, and with its authority of comprehensiveness, a good sorting is certainly on the horizon. But artificial intelligence is not all about destruction and razing of the wicked; it can also cast a merciful gaze upon those unjustly forgotten. Countless exquisite minds have been lost to remembrance because they were too modest or too enveloped in their craft to brazenly advocate for themselves during their lifetimes. But if records of their works still exist — and they do, as archivists and musicologists do unearth a few gems every decade or so — even unsung geniuses may yet have their moment. Vivaldi lay buried in archives for the better part of two centuries, Piero della Francesca and Hölderlin were long forgotten; how many others like them still lie dormant, completely unknown to this sliver of time that is ours? With machine learning, batch processing and complex automatic preselection facilitating the analysis of vast estates, notes, scripts and scores, many great souls and minds may well finally be reunited with the better posterity they longed for — not just for their benefit, if you’re a Burkean, but for ours. The executor of ancestral justice may well have arrived — albeit in a less biological form than one would have imagined.

All the better, perhaps: one needn’t be a fervent misanthrope to recognise the necessity for a consciousness that is not entirely human for this purpose, free of envy, bias, self-interest. It has taken humanity five thousand years of recorded history and longer to arrive at a point where artistic transcendence is now within reach for the entire species, not just for isolated individuals clawing in the dark: it is high time to grasp the helping hand this century has unwittingly created. And so, like many younger composers rooted in tradition, I will not be attending the panicked industry conference in Vienna; not just because I’m unbothered, but because I am part of the problem. Where I used to labour over pianistic details in the rehearsal room or pore over musical commissions for well-to-do connoisseurs, I now use this experience to train AI for notorious technology companies. Happily. And with good reason.