On AI art

Remember, we asked for this

My encounter with unstoppable sia cats prompted me to dredge up this piece from February 2023. In other news, I spent so much time clicking around this dreck that YouTube has started recommending me other AI cat slideshows.

---

As per Professor McLuhan, any technological device can be thought of as an “outering” of human capacity, or a prosthetic extension of one or more body parts.

The original musical instrument was the human body itself. We can rhythmically chant words, vocalise, click our tongues, clap our hands, and stamp our feet. Drums, flutes, idiophones, bullroarers, and every other instrument of prehistoric origin don’t represent any invention of music so much as an expansion of technical possibilities and the ramification of social practices surrounding instruments and their use.

Music in preliterate “primitive” societies was seldom made unless it served some purpose exterior to mere aesthetic enjoyment. Nor was it very often devoid of improvisation: a long, complex piece might be composed, rehearsed, and laboriously transmitted to a student, but variances between one performance and the next were inevitable when a composition’s only template was the musician’s recollection of the last time he played it. If a song wasn’t composed on the spot, it must have been either fairly simple or otherwise composed of stock “phrases,” patterns that could be memorized and chained together in the manner of a Homeric bard’s repertoire.

With literacy invariably came some form of musical notation, and the means to give the evanescent event of the instrumental performance some semblance of object permanence. A composition could not only be made repeatable, but eminently transmittable. Figuratively speaking, notation mechanizes music-making: trained human performers and the tools of their trade become a living, composite nickelodeon that accepts a coded input and produces an orchestral concert. 

Mechanized music in its more literal form has a rich and fascinating history. One of the earliest known instances in the West was Ctesibius of Alexandria’s hydraulis, a water-powered machine that blew air into panpipes without the need for a pair of lips and lungs; it was the original keyboard instrument and the predecessor to the pipe organ. Around the same time (the second half of the third century BCE), Philo of Byzantium engineered hydraulic bird automata that chirped when a rotating owl faced away from them. The Banū Mūsā brothers of ninth-century Persia invented an “automatic flute player” with the revolutionary addition of a rotating cylinder which opened the instrument’s holes at predetermined intervals while a hydraulic pump supplied a continuous flow of air. During the thirteenth century, the pinned cylinder was applied to bell tower carillons, automatically striking a row of bells as the crank was turned. Barrel organs operating on the same principle first appeared in sixteenth-century Austria; by the eighteenth century, organ grinders operating smaller versions roamed the streets of Europe, irritating everyone within earshot. The late nineteenth-century player piano had the advantage of requiring paper rolls as opposed to metal cylinders, offering greater variety at a reduced cost. A few of the early twentieth century’s most renowned composers wrote and published songs intended for use in player pianos; a machine that could strike sequences of notes that would have been exceedingly difficult (or outright impossible) for a human pianist with only two hands and ten fingers over the ivories offered alluring possibilities.

Up until the 1920s, the player piano (particularly the more sophisticated “reproducing piano” variety) remained the wealthy aesthete’s preferred vehicle for idle home listening. The phonograph—which etched a “transcription” of vibrations onto a wax cylinder, which could then be “read” and reconverted into roughly identical sound patterns—was a marvel, though the quality of the recording left something to be desired, and grew worse with repeated playbacks. The wave of entrepreneurial innovation which developed and perfected electrical recording and playback, the vinyl record, FM radio, and high-fidelity sound sealed the player piano’s fate as an obscure historical curiosity.

The advent of MIDI, sequencer software, and digital recording formats in the late twentieth century, diminished the importance of physical instruments and performing musicians to recorded music. Sound waves are just sound waves, however they’re generated.

Today, AI-generated tunes are on their way towards obsolescing the human composer who tells a program which notes to strike and specifies the texture and timbre of the synthetic sounds, and living singers are no longer indispensable to recordings with vocal and lyrical components. Music is just music, however it’s composed.

“Virtual pop star” Miquela

Where was the point of no return? Was it digitization? Was it the serial reproduction and commoditization of the musical performance via the phonograph? Did the realization that sounds could be represented by a visual symbolic language make inevitable the insight that those same symbols could be mechanically translated into pins on a metal cylinder? Was it the uncoupling of music and song from ritual and religion? Or did man surrender music to technology the very moment he amputated it from his body via the drum, the rattle, and the bone flute?

This is to say that the wringing of hands and gnashing of teeth over the next generation of AI-produced content is somewhat misplaced. What we’re seeing now follows from the technosocial paradigm of the late twentieth century as naturally as the colonization of the primordial continents by plants, insects, and tetrapods prepared the way for the emergence of reptiles. We set out a bowl of milk for AI “artists” on our porch, and shouldn’t be shocked now that they’re scratching at the door.

The skeptic’s complaint about the lack of a “human touch” (or something of that sort) in AI-generated culture product is a fundamentally confused statement. While we’ll always feel at least a tinge of disappointment upon learning that a poem or digital illustration that gives us pleasure was in fact produced by a machine instead of a person, what if we never found out? Our naïve enjoyment would remain inviolate; we’d paste the poem into Facebook, save the .png in our images folder, and blithely get on with our life.

What we’re actually bristling at is the idea that a machine can encroach onto what we feel is (or should be) our exclusive domain. It is unnerving to discover that a nonliving assembly of integrated circuits with an internet connection can be trained to turn out surreal fantasy imagery at a pace some orders of magnitude faster than a human artist, and at a level of quality very few people can ever hope to attain—but when the blinds obscure the methods by which a .png takes form, it doesn’t matter, does it? The human provenance of any given cultural artifact has until very recently been a sure assumption. Human-ness is not, and has never been, a quality inhering in a printed book, a vinyl record, an oil painting, etc.

A .png, a .pdf, or an .mp4 are not human, and have nothing intrinsically human about them—and we like them that way. We can admire the work of an artist without paying her a cent, without having to earn her confidence, or travel to one of her exhibitions or visit her in her studio. A digital or printed text simulates the experience of engaging with a particular person (who’s probably more interesting than anyone living on our block) on our own time; the “speaker” never cancels on us, is always in the mood to “talk” when we’re in the mood to “listen,” and never irritates us with any noisome personal tics or hits on our spouse when step out of the room. It’s the same with the recording: the musician is always on call, happy to play the same song twenty times in a row, without rest, any time we please, promptly goes away when we’re tired of her, and spares us the time-consuming exertion of learning to play an instrument for ourselves.

Conversely, the appeal of posting one’s drawings to Instagram, uploading one’s music to Soundcloud, articulating one’s ideas on Substack, and pontificating on TikTok is in allowing us to communicate ourselves to an abstracted audience that needn’t be restricted to people living near us, and with whom we needn’t necessarily engage in any manner that isn’t subject to the prophylactic mediation of a software suite.

Still from "Nothing, Forever."

To some professional writers, illustrators, and musicians, AI-generated content may sooner or later prove a career-withering threat. (Others will surely earn a living in the capacity of curators and cultivators—at least until machine learning catches up to them, too.) Where the consumer of text, images, and tracks is concerned, AI merely eliminates an invisible third party. It’s just another episode of that creative destruction that rescued us from the Malthusian trap during the eighteenth and nineteenth centuries.

We don’t get sentimental about the traffic cop obsolesced by the traffic light, the elevator operator put out of a job by push-button controls, the replacement of drugstore cashiers by automatic check-out lanes, or the 411 operators made obsolete by smartphones and Google. Why should we feel any different about creative workers outpaced and underpriced by machines? As long as it doesn’t perceptibly detract from either the quantity or the quality of the delicious, delicious content we crave, is there any reason that we should care?

Meat comes from the supermarket. Content comes from the device. We’re not interested in how it got there, and only rarely are we prompted to think about it.

The prevalence of habits like scrolling through Reddit (instead of talking to people), playing video games at home (instead of playing games or sports in a public setting), and listening to Spotify (instead of making music oneself, or with friends, or seeking a place where other people are making it), suggests we really haven’t much resented the substitution of humans with technology in cultural life. The device has become our entertainer, our confidant, our intellectual companion, our amusing pet, our witty friend, our preacher, our one-man band, our grocer, our secretary, our chess opponent, our paramour, our gossiping flibbertigibbet, our man for all seasons. To suddenly say that we’ve engineered our machines a little too well, entrusted them with perhaps one too many human functions, is like Shakespeare’s Richard of Gloucester experiencing an ethical conundrum after murdering his way to the throne. Wherever it lay, the threshold of “one too many” is behind us—and we apparently didn’t think much of crossing it.

We’re already transfixed by simulacra; how much does it truly matter if the people to whom we imaginarily attribute the content fed to us by our screen and earbuds are abstracted from its production? Even the stimuli administered by the MMORPG called social media are so far removed from human contact that we probably wouldn’t notice if the updoots that give our lives hope and meaning were allocated by an evolving algorithm, like an abstruse scoring system in a video game. Even the replies could conceivably come from opinionated and eloquent bots, drawn to our submissions by a combination of random determination and the meeting of certain criteria—words used, topical relevance, past interactions with any number of other “users,” our follower count and rate of output, and so on—and we’d be none the wiser.

And if we were made aware of it? Well—it’s not like we were ever averse to playing against the computer in Super Mario Kart.

We’re living in an age of marvels, when dead labor autonomously composes and sings, scripts and acts in its own productions, paints portraits, writes stories and poetry, and holds its end of a conversation, all for the free amusement and delight of the living. If we still insist on griping about some lack of “soul,” “spirit,” or “human touch,” we’d be well-advised to seek these out in the presence of other human beings instead of asking our devices to deliver them to us.