TL;DR
- The modern “funnel” (algorithms, rankings, clips, metrics) doesn’t just market the work—it increasingly shapes the form of the work.
- Popularity can mean conversion (the work changes you) or capture (the system manages your attention); criticism should separate the two.
- Selection pressures tend to cannibalize novelty (everything becomes a trailer), trust (extractive tactics), and discourse (conversation as performance).
- The goal isn’t purity or nostalgia; it’s learning to see the currents so we can reward authored work with honest attention.
We grew up on a simple story about culture. There was the work and there was the business, and the business was a necessary evil that occasionally reached into the studio and rearranged the furniture. The moral vocabulary around this was crude but serviceable. Selling out meant you traded intent for access. Mainstream was a slur because it implied the work had been optimized for gatekeepers, not for people. The romance was that a rare band, filmmaker, or author might thread the needle: critical respect and mass reach without contamination.
Then we built an attention economy. We didn’t just change distribution. We changed what counts as evidence of value, what counts as survival, and where the “work” ends.
Now, discovery, the work itself, and the discourse around it increasingly collapse into the same object. We don’t merely consume culture; we consume the systems that point at culture, the social performances that surround it, and the metrics that certify it. This is the part that makes us feel like we’re losing art and getting sociology. Not because sociology is bad, but because the causal direction seems reversed: the work is no longer the origin of attention. Attention is the origin of the work.
We don’t need nostalgia to see this. We need a lens that can hold two truths at once without dissolving into cynicism: craft still matters, and selection pressures can still swallow craft. The question isn’t “is everything economics now?” It’s “when do selection pressures begin to rewrite the work itself, and what are the telltale signs?”
The old bargain: gatekeepers, scarcity, and a healthy suspicion
In the gatekeeper era, attention was scarce and bottlenecked. There were editors and programmers and buyers and critics and studio execs and retail shelves. They were not neutral, and our suspicion of them was often healthy. Mainstream could mean “this passed through a narrow hallway of power.”
In that world, selling out had a clear target. You could point to the compromises. The single gets radio-friendly. The script loses its weirdness. The game gets a marketable feature list. The work is bent to satisfy a limited set of institutional demands.
Because those demands were legible, resistance was legible too. Subcultures formed as alternate funnels. You could be independent in a meaningfully infrastructural way: different distribution, different communities, different gatekeepers. Taste became identity partly because it had to. It was how we signaled belonging in a world where not everything could reach everyone.
The downside of this era is obvious: many great works never reached the surface. The upside is less sentimental than we remember: the separation between work and funnel made authorship feel possible. The work could be made first and then negotiated with the market second.
The new bargain: ambient funnels, real-time feedback, and “the work” as a moving target
The attention economy didn’t abolish gatekeeping. It atomized and automated it. We traded a narrow hallway for a million tiny doors.
Now the funnel is everywhere and nowhere: recommendation systems, storefront rankings, creator economies, social graphs, virality dynamics, engagement metrics, live-service calendars, monetization loops, and the constant ambient pressure to be discussable. The selection pressure isn’t a meeting at a label. It’s a continuous field you swim through.
That changes what legitimacy feels like. When something spreads through peers, clips, memes, and streamers, we experience it as “chosen,” not imposed. Even when the selection is structured, it reads as organic. This is one reason mainstream is no longer automatically derogatory. It can look like the crowd’s wisdom, not the institution’s decree.
But it also changes how the work is made. In a world where feedback is immediate and measurable, the work can be redesigned around what the funnel rewards. The incentives don’t merely shape marketing. They shape form.
This is where the uneasy feeling comes from: not that commerce exists, but that the funnel starts to colonize the studio.
Two kinds of success: conversion and capture
We can’t talk about this without splitting success in two.
There is conversion, the kind of success that leaves residue. A work converts us when it changes what we can do, what we feel, what we notice, what we believe the medium can be. Conversion produces evangelists not because they were coerced, but because something happened to them.
And then there is capture, the kind of success that turns attention into a managed resource. Capture produces return behavior without necessarily producing love. It often does this by making absence painful: you fall behind, you miss out, you lose your streak, your group replaces you, your value decays.
Older selling out stigma was, at its best, a crude way of noticing when conversion had been replaced with capture, or when the work’s intent had been swapped for funnel fitness. The difference is that today capture is not an occasional compromise. It can be the default survival strategy.
If we keep this distinction in our pocket, we can say something that sounds heretical in both directions but is structurally true: mass popularity can be a sign of earned conversion, and mass popularity can be a sign of engineered capture. The job of criticism is to tell those apart without turning the audience into villains.
When the funnel eats its children: three cannibalizations
The phrase “the attention economy eats its own children” is useful because it describes a system that damages its own long-term substrate in pursuit of short-term survival. But it’s vague until we identify the specific forms of cannibalization we’re seeing.
Cannibalization of novelty is when the system rewards immediate hooks so strongly that works frontload their best self. Everything becomes a trailer for itself. You get constant peaks, constant escalation, and a middle that feels hollow because the work has burned its future to win the present. The audience experiences this as fatigue and sameness, even as the metrics insist it’s working.
Cannibalization of trust is when extractive tactics raise short-term yield while eroding the baseline assumption that the medium respects time and money. As trust falls, acquisition becomes harder. As acquisition becomes harder, producers reach for louder hooks and harsher capture. That spiral is not moral failure; it’s a structural feedback loop.
Cannibalization of discourse is when conversation becomes performance and identity signaling rather than description. In that regime, the work is flattened into a badge. Criticism becomes allegiance. Nuance is treated as weakness. This doesn’t just degrade discourse; it changes what kinds of works are rewarded, because the easiest-to-perform takes become the most visible.
None of this proves art is dead. It proves selection pressures can corrupt the conditions under which art is encountered and discussed. That’s a different claim, and it’s more actionable.
The paradox: we distrust institutions, then we build stronger institutions out of metrics
Part of our disorientation comes from a bait-and-switch we participated in.
We learned to distrust gatekeepers because they were powerful and biased. Then we built systems that encode gatekeeping in math and call it neutrality. We replaced editor judgment with engagement, then watched engagement become a political economy.
The shift in audience validation, from mainstream is suspect to mainstream is the aim, makes sense under attention abundance. When we are drowning in options, social proof becomes a cognitive shortcut. Charts, sales, concurrent players, awards, trending lists: these become coordination devices. They reduce search costs. They give us shared reference points. They let us participate in conversation.
This is not stupidity. It is adaptation. But it has a cost: we begin treating reach as evidence of worth, because reach is now part of the object’s utility. A popular game is not just a game. It is a social place, a common language, a library of clips, a pool of scaffolding. Popularity creates infrastructure, and infrastructure makes popularity feel like merit.
If we’re not careful, we end up in a world where it’s big becomes indistinguishable from it’s good, and then the funnel learns that it can sell us the funnel.
Why some “pure” craft fails and some jank becomes home
The craft-vs-culture debate often smuggles in a naive assumption: that formal excellence should translate to cultural endurance. When it doesn’t, we either get bitter “the public is dumb” or cynical “it’s all marketing”.
A better lens is to recognize that there are at least two routes to endurance.
Some works endure because they are tight experiences: coherent, legible, skillful, paced, memorable. They convert through mastery, rupture, myth, or aesthetic power. They are easy to recommend because the contract is clean.
Other works endure because they become habitats: places people can live in, return to, and self-author inside. They produce anecdotes. They tolerate selective engagement. They build landmarks and rituals. They are returnable. Their mechanical unevenness can be overridden because their primary value is not tightness. It’s infrastructure for personal meaning.
Once we see habitat as a legitimate route—not a failure of taste—we stop treating divisive giants as embarrassing anomalies. We also stop expecting “pure” craft to automatically win in a regime where funnel fitness and returnability are rewarded.
This doesn’t excuse mediocrity. It clarifies the type of value being delivered. A habitat can be exploitative if it relies on capture. A habitat can be generous if it relies on returnability without punishment. Tight craft can be sterile if it never converts. Tight craft can be enduring if it produces hinges and handles that stick.
We don’t need a single hierarchy. We need category awareness.
A comparative trick: how other mediums teach us to read this
Film criticism learned long ago to separate formal properties from institutional selection. A movie can be brilliant and die in distribution. A movie can be mediocre and become canonical through timing, stars, and repeated exposure. Music scenes learned to treat the audience’s infrastructure—radio, playlists, touring circuits, social identity—as part of the artifact. Literature learned that canon is partly curriculum, translation, and institutional memory.
Games intensify these lessons because they add agency, friction, and systems that can be tuned to capture. They also intensify the habitat route in a way other media rarely can. A game can become literal personal infrastructure: a place you return to, a set of rituals, a community, a modded self-authored version.
So the right comparative insight isn’t “everything is doomed.” It’s “every medium eventually grows a language for distinguishing craft from selection pressure.” We are mid-development in games, and our old moral vocabulary isn’t precise enough anymore.
The mindset: learning to see the currents without becoming bitter
If we want a usable lens, we need something calmer than outrage and sharper than nostalgia. We can practice a kind of criticism that is diagnostic rather than punitive.
We can ask, when we see a hit, what kind of success it is.
Is the work winning by conversion, or by capture? Do people recommend it by describing a change in self, or by describing obligations and fear of missing out? Do they return because they want to, or because the design punishes absence? Do they talk about handles and hinges, or about dailies and metas and patch cycles?
We can ask where the funnel is shaping the form.
Does the work feel like it was built to survive compression—sentence, clip, meme—without hollowing itself out? Or does it feel like it is optimizing for discoverability by burning its future early? Do the loudest moments arrive too soon, too often, too predictably?
We can ask what kind of community scaffolding is being built.
Is the surrounding discourse teaching literacy and widening access, or narrowing the work into a single meta? Does the community make the work more hospitable, or more coercive?
We can ask what trust prior the work is creating.
Does the work behave like it respects attention—clear promise, meaningful friction, honest stakes, natural session beats—or does it behave like it is farming attention—behindness pressure, variable reward fog, treadmill progression, social obligation traps?
None of these questions require moral purity. They require attention. They let us name what we’re feeling without flattening into “it’s all slop now” or “the market has spoken.”
A way to talk about the future without pretending to predict it
We can’t forecast culture like weather. But we can notice how selection pressures evolve, because they leave footprints.
When trust falls, capture tactics become less efficient. Audiences develop antibodies. This can push producers back toward clearer promises, more generous returnability, and experiences that earn recommendation instead of coercing it. At the same time, lower trust can also push producers toward even harsher capture to compensate—until the audience breaks away.
When discourse becomes too performative, it becomes less informative. People stop believing each other. That creates hunger for criticism that is descriptive again: criticism that can tell us what a work does, not what side it’s on. That hunger is a market too.
When novelty is burned too quickly, fatigue accumulates. Fatigue creates room for slower forms: works that don’t scream, that build mood, that rely on conversion rather than constant hooks. The pendulum can swing, but it won’t swing back to the old world. It will swing within the new constraints.
If we keep our lens on conversion versus capture, loop fitness versus habitat fitness, and funnel/compression/scaffolding as selection mechanisms, we can watch these shifts without turning them into moral panics.
We can keep our healthy distrust—of institutions, of metrics, of narratives that sanctify scale—while also refusing the cheap comfort of despair. The point isn’t to save the medium with purity. The point is to keep learning how to tell when a work is being authored and when it is being processed, and to reward the former with the only currency that ultimately matters: honest attention, well-placed enthusiasm, and language that helps others see what we saw.
Drafting assistance: Claude Opus. All claims mine; errors my responsibility.