I have seen the big mountain

I first heard “The Big Music” when I was teenager, which is probably the only time I would have been receptive to it. If you don’t know it, it’s an unapologetically bombastic rock song from the early 1980s by the Waterboys, before main Waterboy Mike Scott’s full Celtic turn later in the decade. It sort of sounds like “Cortez the Killer” reimagined as an inspirational anthem. The song’s first line bluntly announced its ambitions: “I have heard the big music and I’ll never be the same.” And neither would you, presumably, if you were listening properly. It offered the basic ideals of Romantic poetry — strive for transfiguring personal experience, attune oneself to the sublime majesty of nature, reject the mediocrity and compromises of society — without the difficulties of poetic interpretation. Instead the music seemed to offer a direct pipeline to transcendence.

Not that it ever made me feel transcendent: I tended to relate not to the striving but to the song’s implicit promise of failure. I heard “The Big Music” but not the big music, and I am pretty much the same now as I was then. The only time I hear “The Big Music” these days is when I am running, because I put the track on the iPod Mini I use when I am exercising. Not only is the song nostalgic for me, but its motivational tone is suitably ironic. After Scott sings “I have seen the big mountain, and I swear I’m almost there,” the backup singers immediately reprimand him with “You’ll never get there. You’ll never get there.” He might see the crescent but not the whole of the moon, as Scott would put it in a later song (which I also loved as a teen). The next line — “I will always climb the mountain” — is broken with a pregnant pause before “always,” so it sounds like Scott is petulantly disagreeing with them: “I will!” This comes across not like an expression of determination but more a kind of childish stubbornness — less an incipient triumph of the will than a desperate display of willfulness. It’s hard to say whether that was intentional; I was always more than ready to project my own ambivalence onto the song. Even as a kid I had my doubts about the wisdom of mountain climbing.

The Big Music was eventually floated by critics as a genre name to apply not only to the Waterboys but to early U2, late Simple Minds, Big Country, and the like — earnest, expansive guitar music from Scotland and Ireland that served as the antithesis of the era’s synth pop. (“That techno rock you guys listen to is gutless!”) In Rip It Up, Simon Reynolds uses words like “elemental” and “panoramic” to describe it, arguing that it “gave the quest for ‘indefinable glory’ a vaguely military or messianic aura.” The genre tended to posit listening to music itself as a spiritual quest, a means of fulfilling a personal destiny or achieving an identity, even as the songs seemed designed to dissolve into the “indefinable.” I never saw the video for “The Big Music” in the 1980s, but watching it now, it’s uncanny how on the nose it is for every cliche ever offered about the style. My favorite moment is when Scott is strumming his 12-string acoustic guitar while standing in front of a massive waterfall that certainly would have drowned out whatever sounds he was making. You don’t even have to hear the big music when you can feel it inside you.

I assume there is some sort of Big Music playlist on Spotify somewhere, but it seems somewhat against the spirit of the genre to stream it through a service that has done so much to transform music into data, making something ineffable into something programmable. Here’s an automated playlist for your journey of self-discovery. Or rather, Spotify seems to co-opt the ruse of the Big Music for itself: Instead of being able to perpetually climb the mountain simply by consuming the majestic-sounding song, letting it do the work of sublime transcendence for you, you can instead have a streaming-service algorithm accomplish that, guiding you effortlessly above the mundane and the quotidian to the panoramic vista, like Casper Friedrich’s wanderer above the sea of fog.

Daniel Ek, the CEO of Spotify, was widely criticized last week after posting this strange myopic claim (to set up an even stranger claim about the resurgence of Marcus Aurelius): “Today, with the cost of creating content being close to zero, people can share an incredible amount of content.” Apparently he hasn’t watched the video of “The Big Music” and seen what Scott had to pay in heart and soul. (Could Stoicism have made that?) And apparently he gave zero thought to how musicians manage to eat or pay rent. Presumably he was thinking about the “cost of creating content” as a matter of the tools to record it and the cost of distributing it, not of the life experience that would go into to being in a position to make music anyone else would want to hear.

Ryan Broderick interpreted Ek’s comment as part of what he called the “CEO passive content obsession,” when executives assume that “at a certain audience size, … those people are locked in and will consume anything you throw at them,” no matter how poorly or indifferently made it is. Of course, entertainment company CEOs also consider it their fiduciary duty to acquire “content” as cheaply as possible and to see to it that it is generally devalued as a whole so that entertainers have less leverage. As Broderick points out, “This is why executives love AI.”

In his view, the endgame is for monopolistic streaming platforms to be filled with so much low-quality user-generated, AI-produced material that those same users will have no choice but to keep paying them for access to the algorithmic tools to help sort through it all:

What’s not being spoken out loud by all of the new “infinite AI-generated Netflix” startups or whatever popping up every week is that the AI slop is just the baseline. They want us to to pay to subscribe to their background noise app and they want us to populate it for them, now that they’ve made, as Spotify’s CEO put it, “the cost of creating content close to zero.” And then they’ll want us to pay more to ask the AI slot machine to generate something we actually want.

This vision of Spotify — an invasive menace that seeks to deskill music consumers and compel them to accept cheaper and ever less substantial content in their pursuit of cruel optimism — contrasts pretty strongly with the view that emerges from Computing Taste, Nick Seaver’s excellent ethnography of streaming company engineers. Seaver generously withholds judgment of the employees of firms like Spotify and strives to understand instead how they can construe what they are doing as beneficial to music listeners. Many of their rationalizations appear to align with what Ek wrote in a post attempting to clarify his obtuse remarks: “In this environment of constant creation, we can identify and ensure that the bold, exciting, world-changing ideas and pieces of art don’t get lost in the noise.” In other words, we must build great recommendation algorithms to rescue great music.

Seaver cites lots of presentations and comments from within the streaming industry that echo that basic premise, that bringing more powerful recommendation algorithms to the masses will bring order to the content wilderness and blaze trails for people to explore and make new discoveries. It’s taken as given that not only is there “too much content” in the world that everyone should feel responsible for, but also that the ideal way of listening to music is “discovery” and that everyone should aspire to “grow” their musical interests to better realize their personal potential. (“I have seen the big mountain!”)

There is less acknowledgement that the scale of streaming platforms is the main contributor to the sense of surfeit, aggregating material from diverse contexts and conditions and making it accessible for frictionless consumption as essentially interchangeable and disposable content. Seaver quotes Eric Drott, who argues that companies like Spotify deliberately develop marketing materials that work to “transfigure plenitude into a form of lack.” The platforms make songs easier to access but also harder to care about, which corresponds with the increasing presence of algorithms built directly into the listening apparatus to do the caring for you. Seaver notes that researchers tended to find that “people’s preferences were ‘noisy’” and not strongly held; streaming services look to give them concrete, trackable substance by disregarding what people say they want or deliberately choose and focusing instead on the data they generate.

The streaming company engineers’ work can be understood as a dogged effort to replace the eliminated context with an ersatz data-based version that their employers can administer and control and profit from. (This is the sort of critique that Seaver is often at pains to disavow and mitigate.) They rationalize this through a kind of paternalistic pastoralism, believing they are tending to the garden of music data and leading their flock to the best feeding pastures. But of course it also means they are intentionally trying to make people into sheep. Seaver likens their practices to the practices of trappers, tracing a “shift in the purpose of algorithmic recommendation, from aiding overwhelmed users to enticing people into becoming users.” Inadvertently or not, the engineers began to muddle the distinction between satisfaction and compulsion. Seaver describes a TikTok-like project that imagined the offering of “no interface” for listeners to choose music as a kind of perfecting of listenership: Its user would be “a member of various statistical distributions, waiting for interactions to actualize an identity out of the potentials predicted by certain known features.”

It feels obligatory here to note the resurgence of collecting vinyl and rhapsodize about the ritual of selecting a record and dropping the needle on it and so forth — all these deliberate (and increasingly expensive) acts of agency to sanctify the listening process and elevate it above “passive content obsession.” Playing albums on a turntable feels like a mode of resistance to being a “user.” In my day, when I was a suburban teenager listening to “The Big Music” on my Walkman, being “into music” meant knowing about bands like the Waterboys; now being into music seems to be more about how you listen to it — it would be about having a Walkman, not some specific cassette. When I happen to tell someone I don’t use Spotify, on the rare occasion that I am in a conversation, I feel the same sort of smugness I would when I used to tell people I didn’t have a TV back when that was true. (Perhaps this is why I am in so few conversations.)

If I heard the big music on Spotify, I would never be the same, not because I would be overpowered by the music’s spiritual purity but because I would be interpellated as a listener on Spotify’s terms, which would be pushing my listening interests in the direction of what makes the company viable. I would become a cog in their cybernetic machine, acclimated to pushy interfaces and autoplays, to having the popularity of music foregrounded in play counts and integrated as a normal part of the listening experience, to the idea that music was just data points in latent multidimensional space, information logged in an endless set of hard-to-navigate menus, databases, and spreadsheets and that there was somehow too much of all of it. Curiosity would be compulsory, a kind of work ethic. I would be trained into “listening to Spotify” instead of listening to specific music, and subtly encouraged to see the point of listening to music as producing more data for Spotify to refine its algorithms and perhaps its eventual generative models. If I wanted to have taste, I needed to be surveilled until I produced enough listening data to become profilable, and then the platform would create a genre of music just for me. I have heard the big music and I will always be the same.

I would be encouraged to see the whole world as an interlocking set of recommendation algorithms, from which there is no escape. No more wide open spaces in which the big music might reverberate. Everything is a trap, which is to say an infrastructure, which is to say a design for living. Seaver quotes a computer science paper that proposes that “life itself is the original recommender system” and “man is the query.” Tech companies have benevolently assembled this original system as a literal network of machines. The meaning of life, the point of having experiences, is so that they can be captured and processed to optimize the overall system and improve the flow of data — recommendation converges with obedience in the service of the master plan, so that what we will enjoy most is doing what we are told.