It was not long ago that I had my first suspicion that AI-generated music would soon make its mark as a specific genre representing the aesthetics of our times.
Perhaps, it might even go beyond mere algorithmically-produced renditions of tropes and genre cliches, like background music for listening to absent-mindedly while working or cooking, perhaps AI-generated music could become so popular that its mere ubiquity essentially ushers an entirely new era of how we perceive the concept of music, or even listen to new sounds.
Perhaps AI-generated music could be the entering of a new era, as revolutionary as the first encounter between man and analog computers in the mid-20th century, a revolutionary moment that accelerated vertiginously when humanity moved into its digital era. After all, electronic music could not be possible as an aesthetic statement if the equipment did not contribute to its entire impact, both formally and on all other levels, like production, distribution, etc.
A giveaway clue that AI-generated music is emerging triumphantly and unapologetically is a glut of albums appearing out of nowhere, all seemingly the work of hyper-prolific composers and producers – no information about them can be found, at least via googling their often generic or non-descriptive aliases.
It makes sense to deduce that what we are listening to is an effort produced by the alchemical process of someone prompting one of the many music generation AI software applications that have been available online, mostly free to use, since late 2023.
What ultimately convinces me that this is AI-generated music is not only the unprecedented quantity and frequency of these releases (after all, they could be uploads of an archive created over many years) is that I can't fail to notice the short duration of the tracks found in these albums (all clock under one minute and thirty seconds).
These absurd volume metrics and random temporal limitations, I confidently assume, are clear giveaways of their transhumanist provenance, especially considering both the capabilities and the current limitations of AI generative applications.
As a result of prompts, their quality varies on the sophistication of the artist teasing the algorithm with instructions aiming to inspire what is, for all intents and purposes, a silicon life form.
Even the thumbnail visuals vibrate with an artificially enhanced, tell-tale, artificial, glow, clearly displaying the eye-pleasing graduated shading and harmonic colorways typical of not-so-state-of-the-art AI image generators like DALL-E 2 or Bing Create, a style that (amazingly) already looks dated, kind of vintage, as if the 6th and much more advanced iteration of Midjourney, or even DALL-E 3, haven't already been the industry standard for months.
Artistic periods are accelerating at the speed of light, since only a few years ago, visual trends took decades to mature into recognizable as historical.
As well as the frantic abundance of these releases, another reason characteristic for their algorithmic nature is the radical genre variegation displayed by an artist like Yoci, who seems to have never met a style of music they didn't like, from metal to classical through disco and tango. Even more surprisingly, they seem to be competent in all of them and quite talented in suspiciously many of them, a standard that perversely justifies the chaotic plenitude of styles featuring in the numerous LPs appearing under the Yoci appellation.
From soundtracks applicable to a wide range of filmic moods to instrumental interpretations of disco-ish bops, improbable saloon music presented under the inexplicable title of tango beats, unexpectedly convincing mash-ups of '90s electronica a la Boards of Canada or Broadcast at their most mellow, trip-hop adjacent approximations of ambient drones, the results sometimes achieve a haunted charm, an almost improvisational lightness, that is as surprising as it is interesting.
Far from prosaic or naive, it is inventive and unpredictable music, as it is also indescribably curious and very strange, like listening to an alien life form interpreting Chopin (or Fila Brazilia for that matter), their non-human mediation filtering the data inputs with a layer of detachment, an artifact whose mathematically constructed point of view becomes possible only in retrospect.
The poetic unpredictability of unstable calculations is a notion that springs to mind.
Original in the sense that their aural structure is the unique result formulated by billions of entry points composing the rhizomatic architecture of a complicated data cloud, these tracks manifest an uncanny familiarity to the ear by only retaining the timbre of their sources in terms of acoustics and then calculating their frequencies in the estimation of what is most popularly listened to. It's an aleatory formation, like starlings forming abstract clouds in the sky, indeterminate shapes of synchronized fluttering while preparing to migrate to the next destination, a formal transformation traveling across the geography of imagination.
They can be perceived as the audible outlines of an acoustic version of the visual perception phenomenon called pareidolia, which is the human tendency to visually assign defined outlines of familiar shapes, like profiles or animals or objects, to the random formations observed momentarily on shifting clouds, or even discern human-adjacent images set in the fixed planes of solid natural elements, like a mountain rock range or shifting desert dunes, their undulating peaks perhaps delineating a peculiarly evocative simulation of a human body, resting across the horizon or even noticing how a flower displays colors and spots which might resemble a butterfly, which in turn might be a monarch specimen, featuring the shape of a skull imprinted on its body.
My theory is that this specific power of subliminal recollection is evoked because these tracks originate from text prompts, whose words and syntax are descriptive of genres, a hypothesis strengthened by the fact that the titles also seem to announce or at least intimate some kind of particular mood (Shadows of Suspense, Silver Screen Symphony, etc).
Examined from this illustrative and narrative angle, this music aspires to the copyright-free tracks produced for wide television and radio usage, expressly created for those occasions when music is secondary to the transmission – sports announcements, game shows, newscasts, educational documentaries, etc.
This places Yoci in the category of a library composer, albeit one whose toolkit contains uniquely contemporary powers, namely the capacity to radically and convincingly reconfigure all of recorded music in seconds, a power afforded by an almost Promethean god, a literal semi-God if we recognize that microchips are indeed an evolutionary step for humanity, beckoning us to become one with our silicon destiny.
However, we also need to recognize the recent history of the industrial, content-orientated, need for creating sounds that support visuals in a nonintrusive manner, yet still providing audiovisual producers, broadcast channels, and all kinds of other platforms including social media, with audio that supports specific intentions and non-musical priorities, like discreetly framing spoken words, or even being acoustically friendly to eventual additional audio, like the eventual overlay of field sounds captured in footage for news, police reportage, documentaries or sportscasting.
A personal favorite by Yoci is Tavern Beats. This thematic collection aims to recreate the dulcet tones and bucolic spirit of a band playing in an inn existing to some imaginary Middle Age idyll.
Emotional, organic, and Arcadian in spirit, it is a pleasant surprise, gently seducing the listener with ear-pleasing harmonies, achieved via simple arrangements of faux-medieval simplicity.
The gentle, folksy sound of these most peculiar recordings reminded me of a warped take on late '60s psychedelic folk, perhaps a jam from a West Coast Pop Art Experimental Band sound check session or the backing track of a long-lost Fairport Convention bootleg.
Perhaps I am reading too much into what is ultimately a man-machine interface that can't adhere (yet) perfectly to any specific orders or the demands of exacting users – which also places these recordings in another of my favorite genres, namely experimental music on the verge of becoming, before the novelty ossifies to the stage of specificity that makes it not only recognizable but also a part of the critical cannon, spurring debate and producing discourse, becoming entrenched in the culture.
It is truly the shock of the new, simultaneously thrilling and inexplicable.
Text written by Panagiotis Chatzistefanou, Berlin, March 2024