The Contentapocalypse is Coming

One musician's thoughts on generative AI, the record industry, and the future of art.

“Are you worried about AI, as a musician?”

I’ve been asked this question a few times lately, mostly as it relates to the Frank Sinatra AI voice mods that folks on the internet have applied to a few PMJ recordings.

The question calls for a bit more specificity: Am I worried about AI nuking the planet? Only a little. Am I worried about AI being monopolized by bad actors that use it to manipulate and control public opinion? Definitely. Am I worried about generative AI replacing human musicians and human artists? Not in the slightest.

I want to talk about that last question today, since it’s the only one I’m remotely qualified to answer. My short answer is this: No, I’m not worried about generative AI replacing human artists, because AI is not human.

AI can generate content — an endless, awe-inspiring deluge of content, even — but only humans can create art. While it’s difficult to separate the two from one another in our current social media age, I believe the distinction between “content” and “art” will become much more apparent, in time.

More on that later. But first, let’s indulge in a bit of fear mongering, in the form of an article alleging that Spotify is sending its users recommendations for a number of AI-generated tracks. Articles like this are pure rage bait for musicians; as if the meager streaming royalties weren’t bad enough, now Spotify is promoting and playlisting generative AI bands over their flesh-and-blood counterparts?

It’s easy to make Spotify — seen by most musicians as a kind of “Death Star” of the music industry — into a scapegoat. It’s also kind of lazy and boring. After all, Spotify does not exist in a vacuum; like all successful businesses, it emerged to solve a problem.

To me, a more interesting story is to go back to the beginning — all the way back — and trace the history of recording industry from its humble origins, all the way to this weird time where we’re not even sure if the song we’re listening to is human-made, or a Frank Sinatra AI clone, or whatever. As with other cultural histories, it quickly becomes apparent that Marshall McLuhan was right all along: technological innovation is the rudder that steers the ship of culture, and the rest of us — musicians, talent managers, power brokers, industry gatekeepers, roadies, groupies — are all just along for the ride. The medium is the message.

So, if you don’t mind, let’s talk music biz history for a minute or two (and if you do mind, perhaps ChatGPT can summarize the following few paragraphs for you). *Cue the Ken Burns music…*

. . . . .

It wasn’t always so easy to make a record. Actually, it was downright impossible until sometime in the 1880s, when Emile Berliner invented the gramophone — a device that used a stylus to translate a rotating flat disc into sound that was amplified by a funny-looking horn. Prior to this, the primary way to consume music was to buy some sheet music and play it on the piano yourself. Obviously, one’s mileage tended to vary in this endeavor.

Berliner’s invention changed things, and the early commercial records available on his Berliner’s Gramophone record label were actually quite diverse in content: Sousa marches, classical and ragtime piano selections, opera arias, speeches and sermons. However, it wasn’t until 1904 — when the Victor Talking Machine Company inked a deal with Enrico Caruso and created the industry’s first superstar recording artist — that the medium showed any promise beyond novelty.

The recording industry had its first boom in the 1920s, with over 140 million records sold in 1921 alone. The technology was still crude, and the process of cutting a record was sufficiently urgent enough to be called, “catching lightning in a bottle.” Performers played into a single microphone — all live, no overdubs. The limiting mechanical nature of things meant that there was a hard cut off at a little over the three minute mark; a convention that has interestingly persisted in pop music to this day.

The decades that followed saw great innovations in recording technology, but the goal was always the same: to create a “hit” record that would take the country by storm, launching the artist(s) that recorded it into superstardom. The high costs associated with producing a record meant that the best strategy was to sign and retain top-tier musical talents, and then match these artist with the material that was most likely to connect with a wide audience — a process that gave rise to an entire label department, known as Artists & Repertoire, or “A&R.”

A&R worked. In its heyday, the process of scouting and carefully developing talent turned hundreds of artists into bonafide legends, whose work managed to transcend the era in which it was recorded. Dozens of artists became cultural icons, that captured the imagination of millions across time. A couple of artists — usually the early adopters of certain genres — even managed to break the mold altogether. A&R was so successful that even in 2024, we continue to reference the benchmarks of these halcyon days: will there ever be another rock band that takes over the world like the Beatles? Will there ever be another pop star as famous as Michael Jackson?

Then came the Digital Age —*cue the proverbial record scratch.*

Like other industries, the recording industry underwent some pretty massive changes as the world made the switch from analog tape to 0s and 1s. Following a brutal industry crash in the early 1980s, a new storage medium — the CD — led to a new round of consolidation, as big multinational corporations seeking to capitalize on the new format began scooping up fledgling record labels and their catalogues. Before long, the labels themselves began to look less like edgy institutions at the forefront of culture, and… well, more like a division of any other large, multinational corporation. A&R departments — previously run and staffed by musicians and music producers — were now beginning to be run by trend-following business executives that came with a sheen of consummate professionalism, but often little musical experience.

Again — it is easy, lazy and boring to make these new executives into scapegoats. It wasn’t right or wrong, it was just what the times demanded of them; remember, technology is actually steering the ship. In all fairness, the marketing prowess of these executives lead to an overall boom in music in the ‘90s: think boy bands, Britpop, hip hop, house, grunge, garage, Lilith Fair, Lollapalooza. It was a new Golden Age for commercial music, as major labels used their newfound marketing acumen to identify and successfully exploit various subgenre niches. It was a time of selling, and of “selling out.”

Still, the biggest disruption to the industry was yet to come. Under threat of piracy from Napster-loving teens and slow to adapt to new digital music formats like the MP3, the labels found themselves backed into a corner in the early 2000s — with no way out, other than to accept the terms and conditions of Apple’s iTunes Store. Under Steve Jobs’ leadership, Apple had correctly forecast the ways that the internet would transform music distribution, and built the infrastructure that the major labels needed to survive. For their efforts, Apple got to dictate the terms of the agreement, in a kind of “Bretton Woods” moment for the industry.

The advent of streaming music in the 2010s — led by aforementioned “Death Star” Spotify — actually offered some relief to the cowed labels; now, instead of maximizing marketing efforts to sell a one-time download for a week, labels and artists could get paid per listen, over many years. This final shift — from convincing fans to purchase music to convincing fans to merely listen to music — was subtle, but had profound implications. Music was now officially part of the attention economy. A new medium once again forced the music industry to reorder itself, and when the dust settled, Big Tech consolidated its position as the industry leader. To quote ABBA: The Winner Takes It All. History lesson over.

. . . . . .

So, here we are in 2024; a time in which one would be forgiven for assuming that “A&R” was meant to be an abbreviation for “Algorithm,” all along. Indeed, the algorithms that govern the attention economy now run the music biz, as well. The currency of this attention economy is data, so it should come as no surprise that the incentive structure guiding the present-day music industry is a variant of the same incentive structure that rewards social media users for all sorts of data-producing behavior. Put simply: more is better. Make more songs. Release more b-sides and b-rolls. Post more clips from live performances. Go on more podcasts. Post more TikToks.

Labels: why spend years searching for just the right raw talent to refine into a global superstar, when you can sign a large stable of artists and tell them to keep posting content until they get a hit?

Artists: why spend three years perfecting an eight song debut EP, when you can release a minute-long song and a few comedic short-form videos each week, to maximize your chances of discovery?

Hey, I don’t make the rules; I generally have to play by them, myself. But, it stands to reason that the way to reach the top in an age where artists have been rebranded as “content creators” is to create more content. Bands become their own record labels. Labels become cross-media megaliths. It’s not right or wrong, it’s just what the times demand. Something-something technology and ships.

A little over a century after Emile Berliner slapped a label on the first 7 inch record, we are living amidst a veritable Cambrian Explosion of music. Just how much new music is being released now? Warner Music Group CEO Steve Cooper puts the number at a mind-numbing 100,000 uploads each day, across all DSPs. It probably took upwards of 50 years for the number of commercially-released records to go from 1 to 100,000. Now, that’s a single New Music Friday.

But it’s worse than that. An artist uploading music today is not only fighting for attention among the other 100,000 uploads that occurred that day — but also among nearly every commercially available record, ever. Just as social media’s unprecedented levels of access brought celebrities and everyday people together on the same playing field, streaming’s unprecedented level of access has thrust every new artist into competition with Bon Jovi and Beethoven, alike.

[For more thoughts on today’s oversaturation of music, writer Ted Gioia covered this extensively on an excellent Substack piece.]

Up to this point, we’ve only been discussing music. However, this proliferation isn’t only happening in music; it’s happening in everything. In 2018, there were just over 500,000 podcasts in existence — already a shocking number, until you consider that the number has ballooned to 5 million today. And while the barrier to entry for publishing a podcast is significantly lower than that of creating an album, it’s nowhere near as low as the barrier to entry for publishing a short video on Tiktok — of which 8.6 billion were uploaded in 2021, alone. We have more content at our fingertips than we know what to do with.

Which brings us back to AI. As much as us humans create, AI has the ability to create content at a scale far beyond our imagination. In fact — with enough energy and computing power — it has the theoretical potential to create an entire internet’s worth of content, instantaneously.

100,000 new songs per day is a lot, but it’s nothing like 100,000 songs per second.

Or 100,000 podcast episodes per second.

Or 100,000 funny yet relatable videos per second.

Or 100,000 vaguely amusing listicles per second.

Or 100,000 influencer-style photos per second.

Or 100,000 reboot movie scripts per second.

Or 100,000 “think pieces” like this one per second.

In the event that us humans aren’t able to completely oversaturate the internet on our own, generative AI will ensure that the job gets done. I think it is entirely possible that in our not-so-distant future, we will have no idea whether the material that populates our feeds are created by man or machine. The two will be one and the same, producing the same output: an endless, soulless stream of digital soma, designed to delight our senses and capture our attention. Peak Content will have arrived.

What happens then?

I think we are heading towards a “Contentapocalypse” — something of the end of the process of mass media proliferation that began in the Middle Ages with Gutenberg’s printing press. The Contentapocalypse is when we collectively take back our attention spans, and rethink our relationship with media. I think that generative AI gets us to this tipping point sooner than we think.

AI content has already begun to drip into our feeds, and viewing the latest crop of its attempts at cinema — nightmarish, non sequitur shots of objects and people morphing into one another — conjures up some deeply unsettling feelings. Maybe it comes from the “uncanny valley” phenomenon. But maybe it also stems from the utter disregard that AI seems to have for our own deeply-held belief that media is an effective and trustworthy store of reality. Maybe this discomfort is just us coming to the realization that a taking photograph or a video of something doesn’t really capture a moment — in the same way that watching a video recording taken from another person’s perspective fails to communicate what it’s like to actually be them.

In time, these AI-generated videos will certainly improve to the point where they are indistinguishable from footage shot by humans. The same goes for AI-generated music. It’ll get there — and yet somehow, I think it will fail to move us. No one will gather their family together to watch an AI-generated film or cue up their favorite AI-generated Summer playlist for a road trip. For some reason, we just won’t find that stuff appealing.

Undeterred, AI will simply continue flooding our timelines with so much content that we won’t even find content very interesting anymore. It’s too much; we will have heard every combination of sound, seen every type of image, and perceived all manner of film. Or, at least it will feel that way.

That’s when it will hit us; we don’t want this stuff at all. We don’t want content. We want art.

The Contentapocalypse is coming, lest we be caught unprepared.

What, then, will become of the content creator in this new, post-content age? Well, that particular job title won’t exist. Truthfully, “content creator” has always been a bizarre amalgamation of man and machine; the name even sounds vaguely industrial. In a Post-Content Age, there won’t be any need for humans to attempt to mimic machines anymore. The machines will take it from there, thank you very much.

There will, however, be an enormous need for humans to be human again, and to create things that feed the human soul. The kind of work that views humanity as not just a series of chemical processes to be hacked, but as part of a much bigger whole. Many folks saddled with the unfortunate “content creator” title today are already creating this kind of work — some of them to widespread acclaim, others quietly in the shadows. As we take refuge from the incoming deluge of AI-generated slop, these creators with the ability to connect to us on a deeper level will be the ones with job security, for a change.

But we won’t call them “content creators” anymore; we’ll call them “artists” — that most sacred of titles, akin to the shaman of days of yore in their ability to access the transcendent. Humans have only been making content online for a few decades, but they have made art since early antiquity. The Contentapocalyse will snuff out our desire for content, but it will do nothing to damper our burning need for art.

So, how can we spot the difference between “art” and mere “content?” Art has a story behind it; it has a mythology behind its creation. When we take in the beauty of an impressionist masterpiece or listen to a great Soul record from the 1960s, we aren’t just engaging with disembodied sights and sounds; we're also engaging with the creators of the work and the story of its creation. We’re looking out that window in Saint-Rémy-de-Provence, contemplating eternity. We’re in the vocal booth in Studio A, turning our hardships into melodies. It is the awesome power of our empathy — a trait notably absent in all things AI — that allows us to perform this miraculous feat. In the post-content future, I believe we will come to appreciate this gift even more, and embrace humanity — with all of its frailties and shortcomings — like never before.

I realize that this is a bold prediction; after all — to misquote Madonna — we are living in a Materialist world, and I am a ragtime pianist. But something tells me that the real crisis of our times is not merely political or technological, but spiritual: we have simply forgotten how to be human. All of the current ideological fads we see all around us are merely novel, worldly attempts at answering a spiritual question, in an age dominated by technology.

What many of the tech futurists seem to miss is that humanity’s great frailty — our mortality — is also a necessity. Without death, life has no meaning. Without this mystery — our search for meaning in a sometimes cruel, often chaotic world — there can be no art. All the ancient societies and religions understood this. All the great artists — whether they adhered to a religion in the traditional sense or not — did, as well.

And music fans today understand this. Contrast the permanent immediacy of digital streaming with the live touring space, where scarcity still reigns supreme. Legendary, aging artists and acts from the ‘60s and ‘70s are commanding higher inflation-adjusted ticket prices and grossing more revenue than in their heyday. The reason should be obvious; unlike their records, these artists won’t always be here with us. After they sing that final lyrics and strum that final guitar chord, all we will have left are the memories. Although their voices may be a bit worse for the wear and their lyrics a bit incongruous with their current lifestyles, when we hear these artists today — live, in the same room as us — we become acutely aware that we are witnessing something uniquely special. If the early record studios were catching lightning in a bottle, then watching these artists live is to experience the lightning directly; a brief, brilliant flash of the transcendent.

We live in interesting times, and there are some pretty big changes coming our way. But do not fear the Contentapocalypse, and do not worry about AI.

Scratch that — continue to worry about AI, just don’t worry about it replacing artists. Us humans still have a trick or two up our sleeves.

-SB

---

P.S. If you made it through this entire essay, here’s a musical reward for you, in the form of our latest Postmodern Jukebox video starring Effie Passero. My own attempt to practice what I preach took the form of spending months arranging Roxette’s “It Must Have Been Love” in the orchestral pop style of hit records from the 1960s, and if Effie’s vocals aren’t the best example of the transcendent power of the human singing voice, then I don’t know what is: