Optimized to Death: How Audiences Came to Love Formulas

On the Paradox of Infinite Choice and Creative Stagnation.

When was the last time a genuinely new star achieved stadium-level ubiquity? Billie Eilish in 2019? Before her, you have to go back to Drake (2009) or Taylor Swift (2011, seven years after her debut).

Our dominant musical acts—Swift, Beyoncé, Drake—have been famous for fifteen to twenty years. They are not this decade’s stars; they are last decade’s stars, still holding their positions through accumulated cultural mass rather than continued reinvention.

The pattern extends everywhere. Our highest-grossing films are sequels or extensions of 20th-century IP. TikTok is barely six years old, yet creators already speak wearily of algorithmic demands—hooks within three seconds, emotional beats reverse-engineered from engagement data.

We scroll, watch, listen, and increasingly feel we’ve seen it all before.

This shift is measurable. A 2012 study published in Scientific Reports analyzed the Million Song Dataset and found that timbral variety has decreased, pitch transitions have become more predictable, and loudness has increased. In 2023, nearly 80 percent of top-grossing films were sequels, remakes, or franchise extensions.

What makes this moment distinctive is not that formulas dominate—formulas always have. The sonnet had fourteen lines; the blues had twelve bars; the Hollywood studio system had genre conventions as rigid as any algorithm.

But something fundamental has shifted in our relationship to these formulas. This is what cultural theorist Raymond Williams called our “structures of feeling”—the distinctive ways each era experiences art.

Today’s audiences have become formula-literate in ways our grandparents never were, and this literacy hasn’t liberated us.

Netflix doesn’t just use data to decide what to make; it shares that data with creators, who adjust accordingly. Spotify’s algorithm doesn’t just recommend songs; it shapes how they’re written, mixed, and structured. TikTok doesn’t just distribute videos; it trains a generation of creators in the grammar of virality.

The formulas are no longer hidden. And here’s the oddity: we seem to like it this way. Modern audiences derive genuine pleasure from watching formulas well executed, from anticipating the emotional turn before it arrives, from the satisfaction of prediction confirmed.

We’ve shifted from narrative immersion to what we might call formula consciousness: acute awareness of patterns that transforms consumption from immersion to auditing. We no longer lose ourselves in stories; we track their machinery. We enjoy watching the machine execute with precision.

This meta-awareness can be pleasurable, but once the logic of execution becomes the primary object of enjoyment, imagination is no longer asked to risk disorientation.

Streaming platforms have accelerated this shift, turning formula from craft knowledge into explicit science. Netflix shares data with creators about ideal episode length (47 minutes), viewer drop-off points (eight minutes without a hook), optimal plot turns per hour.

Spotify’s algorithm favors songs between 2:30 and 3:30 with hooks in the first 30 seconds—so intros have disappeared. TikTok trains creators in virality grammar: hook in three seconds, payoff in fifteen, complete arc by sixty.

The formulas are no longer hidden. They’re the product.

And we seem to like it this way. When Marvel movies hit beats on schedule, when pop songs build to inevitable climaxes, we feel not surprise but validation.

We’ve developed optimization pleasure—satisfaction from watching systems execute flawlessly, seeing predictions validated, experiencing emotions on reliable schedule. Prediction is the new catharsis.

This represents a profound shift in what culture is for. Historically, stories helped us make sense of experiences we didn’t understand—grappling with mortality, injustice, transformation. Now stories increasingly deliver predictable emotional experiences as efficiently as possible.

The goal is satisfaction, not insight. Comfort, not challenge. And once culture shifts in this direction, it trains us to prefer the predictable, distrust the disorienting, seek validation over provocation.

The Death of Cultural Rupture

For most of the 20th century, American popular culture experienced genuine shockwaves — moments when something so new arrived that it didn’t just reshape an industry, but rewired the culture itself. These weren’t just trends. They were ruptures—sudden breaks in the cultural order.

Look at the pace of change in earlier eras:

1955–1965: Elvis appears on The Ed Sullivan Show in 1956 and provokes moral panic. The Beatles arrive in 1964 and redraw the map of global youth culture. Bob Dylan goes electric in 1965 and upends the boundaries of folk music. Within a decade, the cultural landscape looks unrecognizably different.

1966–1977: The Beatles reinvent what a pop album can be with Rubber Soul and Sgt. Pepper’s. Hollywood is disrupted by films like Bonnie and Clyde and Easy Rider. Punk explodes in 1976–77, attacking rock music’s bloat with raw minimalism. Again — the mainstream does not adapt, it is overthrown.

1978–1991: Hip-hop rises from the Bronx and becomes a global force (Sugarhill Gang, Run-DMC, N.W.A). MTV launches in 1981 and redefines how music is even seen. Nirvana’s Nevermind (1991) doesn’t just sell — it ends an era and begins another.

1992–2007: Grunge, gangsta rap, riot grrrl, and Britpop each make legitimate mainstream breakthrough. Napster (1999) blows up the economics of music. YouTube arrives in 2005 and changes access to culture forever.

2008–2019: Then something changes. Streaming consolidates power. Algorithms begin determining what rises. Billie Eilish in 2019 may be the last true “out of nowhere” pop star to achieve global ubiquity.

2020–present: TikTok accelerates trends — but produces few lasting artists. Culture feels faster, but not deeper. AI begins generating content that is indistinguishable from formulaic human output.

The Pattern Behind Every Cultural Breakthrough

Every one of those earlier ruptures had the same basic structure:

▪ They came from the outside — from subcultures, not corporate strategy.

▪ They shocked — and were often rejected by — industry gatekeepers at first.

▪ They reached the masses through uncontrolled channels: college radio, early MTV, mix tapes, word of mouth, file-sharing networks.

The key point is this: the gatekeeping system was once decentralized and full of gaps.

Enough cracks existed for the unexpected to slip through.

Today, those cracks are sealed.

Distribution is now centralized under platforms that do not operate on curiosity or taste — but on data and predictability.

Spotify doesn’t hire DJs. It runs recommendation algorithms. TikTok has no editors. Only engagement-optimizing code. Netflix doesn’t back vision. It relies upon pattern recognition.

These are not cultural institutions. They are behavioral optimization systems. And systems like that can’t afford to be surprised.

Why This Matters

For the first time, we are facing the possibility that cultural revolution is structurally impossible — not because artists aren’t trying, but because the infrastructure won’t let it scale.

This aligns with what cultural critic Fredric Jameson called pastiche — imitation stripped of irony. Stranger Things doesn’t critique the past. It recreates it as mood. Likewise, much of today’s music doesn’t sound like the 1980s — it sounds like longing for the 1980s.

Mark Fisher called this condition capitalist realism — the idea that we can no longer imagine a future that is truly different. Culture becomes backward-looking not out of nostalgia, but out of imaginative exhaustion. His word for this “life after the future” was hauntology — the feeling that we are living among the ghosts of possibilities we no longer believe could actually arrive.

Why the Old System Worked (And Why the New One Can’t)

Why could the constrained studio system produce Casablanca and Double Indemnity? Why could the Brill Building and Motown—literal factories—produce lasting works that contemporary machinery cannot?

Navigable vs. Algorithmic Constraints

Hollywood’s Production Code said you couldn’t show certain things but didn’t dictate how to make people feel. Billy Wilder couldn’t show explicit sexuality in Double Indemnity, so he invented a new cinematic language of suggestion.

Today’s constraints are diffuse and algorithmic—Netflix gives you data about when demographics drop off, which plotlines test well. You’re not navigating artistic rules; you’re optimizing for infinite, shape-shifting metrics.

The old constraints left space for mastery; the new ones demand compliance.

Craft vs. Data Training

The studio system and Motown trained people through apprenticeship. Directors worked up from assistant to editor. Motown’s Funk Brothers developed intuitive musical vocabulary through daily collaboration.

Film theorist David Bordwell documented the “classical Hollywood style”—shared grammar that enabled both communication and individual expression. Contemporary Hollywood trains people to read data. Film schools teach technique, but the actual training ground is “what tested well.”

Data can tell you what worked; it can’t tell you why or how to make something new that works differently. Craft knowledge is generative; data knowledge is iterative.

Risk Distribution

The old system spread risk across many small bets. Studios made hundreds of films yearly; Motown released dozens of singles simultaneously. Hits subsidized experiments, and sometimes experiments became hits.

William Goldman’s “Nobody knows anything” reflected this uncertainty—since you couldn’t predict success, you had to make many different things.

Contemporary systems concentrate risk into fewer, larger bets. Marvel movies cost $250-300 million. When projects are this expensive, you can’t experiment. Paradox: old system was more constrained per project but more experimental across projects.

The new system is less constrained per project but more formulaic across projects.

The Death of the Middle

Studios made A-pictures, B-pictures, and everything in-between. This middle tier was where people learned greatness—Scorsese made Boxcar Bertha before Mean Streets, Spielberg made Duel for TV before Jaws.

Contemporary Hollywood has blockbusters ($200M) and indies ($5M), and nothing between. No $20-40 million space for real creative risks with professional resources.

Without this middle, there’s no training ground and no space for “interesting failures” that teach valuable lessons.

Integration vs. Fragmentation

At Motown, the same musicians played on dozens of records, developing a house sound. Studio contract players worked together repeatedly, developing shorthand. This created dialogue—directors pushed against conventions in conversation with other directors.

The contemporary system is fragmented. Every project is assembled from scratch. Different platforms have different algorithms and audiences. No stable ecosystem, no ongoing conversation, no accumulation of collective knowledge.

Authority and Authorship

The old system had gatekeepers with taste. Irving Thalberg, Darryl Zanuck, and Berry Gordy made aesthetic judgments, not analytics departments interpreting data. You could argue with them, learn from them, and prove them wrong.

Michael Curtiz could convince Jack Warner to try something different in Casablanca. Marvin Gaye could push Berry Gordy to release What’s Going On.

The contemporary system has gatekeepers without taste—or whose “taste” is algorithmically derived. Executives greenlight based on “comparable titles” and “audience demand signals.” When artists disagree, there’s no one to argue with—data is data.

As filmmaker Paul Schrader complains, streaming executives speak in “completion rates” and “viewing minutes,” not aesthetic judgment.

Time Scale vs. Time Pressure

The old system worked fast but thought long. Hawks made His Girl Friday in weeks, but studios were building careers and catalogs that would outlast single hits. Contemporary system works slowly but thinks short term.

Marvel films take years but are designed for opening weekend. Netflix shows take months but are designed to be watched in the first week. When you optimize for short term, you can’t build depth that creates lasting work.

The Monoculture Paradox

The old system was more homogeneous but produced more diversity. Everyone watched the same networks, listened to the same radio, but within that space was wild variety—westerns, musicals, noir, screwball comedies.

As media scholar Siva Vaidhyanathan notes, the contemporary paradox of choice: more options don’t mean more diversity when mediated by algorithms. We have infinite streaming options, infinite playlists, but it increasingly looks and sounds the same—optimized for the same engagement metrics, using the same data-driven logic.

Paradoxically, when everyone was forced into the same room, they had to differentiate aesthetically. Now everyone finds their own room but ends up in algorithmically similar rooms. Cultural analyst Lev Manovich’s computational research confirms: as personalization increases, aggregate diversity decreases.

The old studio system, the Brill Building, and Motown were factories organized around craft mastery constrained by navigable rules in service of aesthetic ambition. Contemporary Hollywood is a factory organized around data optimization constrained by algorithmic imperatives in service of engagement metrics.

The first could produce greatness because greatness was sometimes what it was designed to produce. The second produces formula because formula is what it’s designed to produce.

This updates Theodor Adorno and Max Horkheimer’s 1944 “culture industry” critique. They imagined standardization imposed from above. Today’s standardization emerges from below, through collective optimization responding to engagement data.

It’s not imposed; it’s discovered through algorithmic iteration. And we actively participate—through clicks, streams, viewing minutes—in generating the data that creates the cage.

What’s Lost and What Might Be

Defenders offer counterarguments that deserve to be taken seriously:

“You’re just old. Every generation thinks culture is declining.”

This is the most common dismissal, and it has historical precedent. Critics complained that novels would rot young minds, that cinema was inferior to theater, that rock and roll was primitive noise. Every new medium provokes generational anxiety.

But there’s a crucial difference: measurable homogenization in popular music isn’t subjective complaint—it’s documented fact. The Serrà study I cited earlier used computational analysis to track decreasing timbral variety, more predictable pitch transitions, and increased loudness. These are objective measurements, not taste-based judgments.

Moreover, unlike previous generational critiques, today’s young people agree their culture feels stagnant. TikTok creators themselves complain about algorithmic homogeneity. Interviews with Gen Z music fans reveal frustration with Spotify’s repetitive recommendations. The call is coming from inside the house.

“More diverse voices have access than ever before.”

This is true and important. Barriers to entry have fallen dramatically. Anyone can upload to YouTube, Spotify, or TikTok. Independent artists can build audiences without label support. Voices historically excluded from mainstream culture—women, people of color, LGBTQ creators—have platforms that didn’t exist twenty years ago.

But access without distribution infrastructure means visibility remains gatekept, just by different gatekeepers. A thousand flowers blooming in separate algorithmic gardens creates atomization, not cultural dialogue. The question isn’t whether experimental art exists but whether it can achieve mass cultural impact—can it interrupt the mainstream conversation? The answer, increasingly, is no.

Representation has increased even as genuine disruption has decreased. We have more diverse faces in the formulas, but the formulas themselves have calcified. This matters: inclusion within a stagnant system is better than exclusion, but it’s not the same as transformation. The goal should be both representation and innovation, not one or the other.

“Audiences have more choice than ever.”

This is the most seductive defense and the most wrong.

Infinite choice under algorithmic curation produces the paradox of homogeneity-through-personalization. When Netflix recommends “content for you,” it’s optimizing within demonstrated preferences—you get more of what you’ve already liked, refined and refined until you’re in a loop. Choice multiplies; possibilities contract.

Chris Anderson’s influential 2006 book The Long Tail predicted that digital distribution would liberate niche content from the tyranny of hits. The economics of physical shelf space constrained variety; digital’s zero marginal cost would enable infinite variety to flourish. And in raw number of available titles, this came true.

But Anderson didn’t anticipate algorithmic curation. The long tail exists, but no one can find it. Spotify has 100 million songs; 99 percent get fewer than 10 streams. Netflix has thousands of titles; most get negligible viewing. The head has gotten bigger, the tail longer, but the middle—the space where cult artists once built sustainable careers—has hollowed out.

Eli Pariser’s The Filter Bubble (2011) identified the problem: personalized algorithms create “a unique universe of information for each of us” that “fundamentally alters the way we encounter ideas and information.”

We think we have choice, but we’re actually in highly curated bubbles, each optimized for our existing preferences, rarely exposed to anything genuinely surprising.

The appearance of choice masks its absence. When every option is algorithmically similar—just different enough to feel like variety, similar enough to keep you engaged—you’re not choosing between fundamentally different things. You’re choosing between variations on a theme the algorithm knows you’ll tolerate.

The Stakes Transcend Aesthetics

Popular culture has historically been where societies rehearse new ways of being—where the unthinkable becomes thinkable, where marginal identities find representation.

Rock and roll didn’t just sound different; it performed new relationships between Black and white culture, youth and authority.

Hip-hop created new artistic economies from overlooked materials, gave voice to excluded communities, demonstrated that what dominant culture discarded could become raw material for something new.

When popular culture becomes algorithmically optimized formula, it loses this function. It becomes affirmative—giving us more of what we want—rather than transformative—showing us what we didn’t know we needed. As cultural theorist Stuart Hall argued, popular culture is a site of struggle. But now rebellious energies can’t achieve the scale necessary to interrupt the mainstream.

The margins remain alive—Bandcamp musicians, A24 films, Substack writers, podcast experimentalists. But marginal art can’t perform popular culture’s social function precisely because it’s marginal. It speaks to niches, not publics. It can’t create shared reference points enabling collective meaning-making.

Change probably requires infrastructural transformation: distribution channels not governed by engagement optimization (publicly funded platforms, creator cooperatives); funding models not demanding immediate returns (patient capital, endowments, grants); critical institutions championing difficulty over frictionlessness; audiences willing to accept that surprise might feel like friction.

Whether such infrastructure can be built within platform capitalism—or requires imagining economic relations we can’t currently envision—remains an open question. Fisher might well be right: our inability to imagine cultural futures might be symptomatic of deeper inability to imagine different futures at all. Cultural stagnation is the system working as designed.

Optimized formulas serving engagement metrics mirror optimized returns serving shareholder value. Both are engines of iterative refinement, not generative imagination.

As philosopher Slavoj Žižek argues, contemporary capitalism has colonized imagination itself—we can envision technological progress but not social transformation, dystopian futures but not genuinely different presents. Even our revolutionary fantasies are conservative, imagining return to earlier states rather than advance to something new.

Yet Fisher insisted imagining postcapitalist futures was necessary precisely because capitalism’s requirement of endless growth on a finite planet is a contradiction that will force change. The question is whether that change will be catastrophic or transformative, and whether culture can help us imagine the transformative path.

The future of genuinely transformative popular art probably doesn’t come from the factory. Perhaps the monoculture is over—that shared space where ruptures happened because everyone was watching. Perhaps we’re in a new configuration where “popular culture” means algorithmic surfacing to segmented audiences, not collective events but individualized feeds, not movements but trends.

Or perhaps transformation is already happening in margins but looks different from historical ruptures. Perhaps TikTok microgenres, Discord communities, Substack networks, and Bandcamp scenes are creating new forms of collectivity that don’t scale like 20th-century popular culture but constitute genuine innovation.

What seems clear: algorithmic optimization cannot be the only logic governing cultural production. Art optimized for engagement metrics produces exactly what we have—endless variation within narrow parameters, formulas refined to perfection, satisfaction without surprise. If we want something else, we need systems organized around different values.

This isn’t just a technological problem; it’s political and economic as well. Technology could serve different ends. Algorithms could be designed to surprise rather than confirm, expose rather than reinforce, challenge rather than comfort. But that requires different ownership structures, business models, and metrics of success.

Breaking this loop requires what Herbert Marcuse called “the great refusal”—rejecting optimization logic in favor of play, experimentation, waste. Making art that doesn’t optimize for anything, that refuses metrics, that accepts inefficiency as the price of exploration.

Some of this is happening. Artists working outside platform economy, funding through Patreon or grants, distributing through newsletters, building communities not based on algorithmic recommendation. It’s fragile and small-scale, but it exists.

Whether this can become a model for something larger—whether we can build entire cultural infrastructure on these principles—remains uncertain. But the alternative is accepting that optimization is the only future, formula the final form of culture, surprise something we’ve lost forever.

The human capacity for imagination, for creating something genuinely new, hasn’t disappeared. It’s been channeled, constrained, optimized into narrow forms. But it persists. The question is whether we can build conditions for it to flourish again, or whether we’ll continue refining formulas until we forget what it felt like to be genuinely surprised.

That choice is still ours. For now.