The Future Under Glass

Ghost Modernism, Infinite Archives, and the Flattening of Novelty

Abstract

The twenty-first century has produced an apparent paradox. Cultural production has become more abundant, more accessible and more technologically capable than at any previous point, yet the experience of epochal novelty has weakened. Music, fashion, film, television, graphic design and even consumer technology increasingly circulate inside an environment in which the past is never absent. A teenager can move from 1977

punk to 1997 trip-hop, 2004 fashion photography and a 2026 release in a single scroll. Historical styles do not

return after long periods of obscurity; they remain continuously available, searchable, taggable and

commercially reusable. This essay takes as its point of departure Taraka Larson's early-2010s concept of "Ghost

Modernism" in The Now Age, a remarkably prescient account of a culture haunted by an endlessly accessible

archive. It places Larson alongside Fredric Jameson's account of postmodern pastiche, Simon Reynolds's

retromania and Mark Fisher's "slow cancellation of the future," then extends their arguments into the age of

recommendation algorithms, streaming libraries, franchise economics, microtrends and generative artificial

intelligence. The central claim is not that creativity has ended, nor that culture was inherently better before the

internet. Digitization has expanded access, reduced gatekeeping and enabled enormous quantities of new work.

The more precise problem is that novelty has been flattened: new work is born into immediate comparison with

almost everything that preceded it, while platforms and risk-averse institutions reward recognizable forms,

familiar intellectual property and rapid legibility. The result is a culture rich in objects but often poor in

temporal distance. The essay concludes by examining the analogue and physical-media revival among younger

generations as a possible counter-movement, and argues that the next genuinely distinct cultural era may

depend not on destroying the internet but on recovering scarcity, locality, opacity, patience and the creative

function of forgetting.

Essay structure

1. The strange death of the cultural decade

2. Taraka Larson and the birth of Ghost Modernism

3. Before Ghost Modernism: pastiche, retromania and lost futures

4. The archive becomes an environment

5. Platform time: when trends arrive already exhausted

6. The economics of recognition: why old intellectual property keeps returning

7. More culture, less epoch: the difference between abundance and distinctiveness

8. Generational excavation: when the recent past becomes exotic

9. The return to matter: physical media, third places and deliberate friction

10. The algorithm and the problem of personal taste

11. Artificial intelligence and the archive that speaks back

12. What the stagnation thesis gets wrong

13. What would a genuinely new era require?

14. The right to forgetTHE FUTURE UNDER GLASS

1. The strange death of the cultural decade

A person who grew up in the late twentieth century could once expect decades to announce themselves with

a certain bluntness. The difference between photographs taken in 1968 and 1978, between a television studio in

1982 and one in 1992, or between a pop record released in 1977 and one released in 1987 could be perceived almost

instantly. These distinctions were partly technological: recording equipment, film stocks, broadcast standards,

manufacturing processes and consumer devices changed. They were also institutional and social. A small

number of radio stations, magazines, labels, television networks, fashion houses and retailers created

bottlenecks through which styles passed, while local scenes evolved in relative isolation. Fashion changed slowly

enough for a silhouette to become associated with a period. Music technologies became fashionable, saturated

the culture and were then discarded. Interior design, advertising, typography and consumer electronics acquired

period signatures. Historical memory later exaggerates these differences, but the differences were real enough

that a person could often date an anonymous image within a few years simply by looking at it.

The 2010s and 2020s are harder to read in this way. They certainly have characteristic technologies, from the

smartphone and wireless earbuds to electric vehicles, subscription interfaces and generative artificial

intelligence. Yet their aesthetic boundaries are porous. A new restaurant can be dressed like 1974, a music video

like 1999, an album sleeve like 1983 and a fashion campaign like 2007 without being regarded as unusually

experimental. Streaming interfaces place a record released yesterday beside one made before the listener's parents

were born. A film shot on the most advanced digital cameras may simulate degraded 16mm stock. Digital

photography is routinely filtered to resemble instant film, disposable cameras or early compact digicams.

Computer software offers exact replicas of synthesizers whose limitations once defined entire genres. Even

consumer objects that became obsolete only a decade or two ago can return as lifestyle accessories before the

people who first used them have reached old age.

This is not simply nostalgia. Nostalgia is ancient, and cultural revival predates the internet by centuries. The

Victorians revived Gothic architecture; the Renaissance consciously rediscovered classical antiquity; twentieth-

century fashion repeatedly borrowed from previous periods. The contemporary difference is infrastructural.

Earlier revivals took place against a background of partial forgetting. The past had to be recovered from books,

records, libraries, archives, private collections, specialist shops, museums, photographs, oral memory and

physical travel. Retrieval demanded time. Much remained inaccessible. The internet transformed that

background condition. The past is no longer primarily behind us. It is around us.

The result is a peculiar condition in which cultural chronology still exists but has lost some of its authority.

The date attached to a song, film, photograph or style has become metadata rather than a barrier to encounter.

Historical periods now coexist in the same feed, playlist, search result and mood board. A person can spend the

morning listening to Japanese city pop from 1982, watch a newly released series designed to evoke 1996, buy a

newly manufactured cassette player at lunch, and encounter an algorithmically generated image combining Art

Nouveau, 1970s science fiction and 2000s web graphics by dinner. Each object retains a date, but the user's

experience is simultaneous.

This essay calls the resulting condition the flattening of novelty. The phrase does not mean that nothing

new is created. That claim would be both melodramatic and empirically weak. Digitization has lowered barriers

to production and distribution, allowing more musicians, filmmakers, writers, designers and independent

studios to reach audiences than older gatekeeping systems permitted. Economic research on digitization has

repeatedly found substantial growth in the number of creative products available, including work that would

not previously have survived the economics of physical distribution.[6][18] The flattening concerns something

subtler: the diminished experiential distance between new and old. Novelty can still occur, but it struggles to

dominate the temporal field in the way earlier innovations sometimes did. Every new object arrives inside an

archive that can instantly name its precedents.

The distinction matters because cultural originality is never the creation of something from nothing. Artists

have always borrowed, transformed, misremembered, stolen, translated and recombined. The important

historical question is how those processes occur. A musician in Manchester in 1978 might have known a handful

of German electronic records, Jamaican dub singles and American funk releases, but not possessed instant access

to every adjacent scene, interview, studio photograph and production breakdown. Incomplete knowledge

created mutation. A designer who had seen only photocopies of a distant style reproduced it incorrectly, and theTHE FUTURE UNDER GLASS

error became a new visual language. Local scenes evolved through latency. Rumour mattered. Scarcity created

depth because repeated exposure to a limited number of objects made them symbolically dense.

The internet has not abolished these processes, but it changes their default environment. Search removes

ignorance. Streaming removes scarcity. Social platforms remove distance. Recommendation systems remove

much of the effort involved in discovery. Generative systems increasingly remove even the need to locate a

historical reference manually, because the archive can be synthesized on command. This produces abundance on

an extraordinary scale, but abundance alters the conditions under which distinct cultural eras become legible.

The most interesting question, then, is not whether the present is creative. It plainly is. The question is why

the present so often feels temporally mixed, why cultural forms are revived so rapidly, why major entertainment

institutions invest so heavily in existing intellectual property, why microtrends become exhausted almost as

soon as they are named, and why younger people increasingly seek out the physical limitations that earlier

generations spent decades trying to escape. These phenomena are connected by a common structure: the archive

has ceased to be a repository we occasionally consult and has become the atmosphere in which culture is made.

2. Taraka Larson and the birth of Ghost Modernism

Taraka Larson's The Now Age is an unusually useful starting point because it described this atmosphere

before it had become fully ordinary. Written in the early 2010s, the manifesto was part art theory, part mystical

system, part philosophy of music and part attempt to recover symbolic intensity from a culture Larson believed

was becoming trapped in repetition. Its language is deliberately metaphysical, but the central historical

observation is concrete. Larson argued that the aesthetic impulse of the new millennium was becoming less

engaged in direct dialogue with the present and increasingly "haunted" by the icons and vernacular of the past.

She named this condition Ghost Modernism.[1]

The image is precise. A ghost is not simply an old thing. It is an old thing that remains active in the present

without fully returning to life. Larson described a sonic landscape littered with the remains of earlier

movements, made newly accessible by technology. The World Wide Web, in her account, had sprayed the linear

trajectory of time across a common surface, giving users the ability to experience multiple eras simultaneously

without committing fully to any of them. Music therefore became an expanding archaeology of references.

Styles once bound to historical situations could be extracted as surfaces, recombined and recirculated. Larson

called the result "zombie aesthetics," a culture in which forms survived while the original symbolic force that

generated them weakened.[1]

Her argument is more discriminating than a simple attack on revivalism. The Now Age explicitly

distinguishes imitation from what Larson calls Channeling or the Return to the Source. The difference is not

whether an artist uses the past, because every artist does. The difference is what the encounter with the past

produces. Channeling draws strength from historical material in order to intensify contact with the present.

Zombie aesthetics use the past to escape the present, reproducing signs whose force has already been spent. In

Larson's vocabulary, a genuine return to source does not restore an old form unchanged. It converts inheritance

into renewed symbolic power.[1]

This distinction prevents the argument from collapsing into generational conservatism. It would be easy to

say that the music of one's youth was alive and later music is derivative, or that physical media are authentic

while streaming is artificial. The Now Age is more demanding. A vinyl record can be consumed as a fetish just as

superficially as a digital file, while a sample taken from a decades-old recording can become part of a genuinely

new language. What matters is depth of relation. The problem is not the archive itself. The problem is an

archive used as a catalogue of detachable surfaces.

Larson developed the point further in a 2012 interview about the manifesto. She argued that digital

abundance encouraged a "horizontal" listening pattern in which people encountered enormous numbers of

bands at a shallow level. The alternative was "vertical" listening: spending sustained time with a band, song or

record, allowing the experience to deepen rather than continuously moving laterally across options.[2] This is one

of the most prescient elements of the project. It anticipates the problem that later became familiar as infinite

scroll, choice overload and algorithmic discovery. The issue is not merely that more material exists. It is that the

interface encourages a particular geometry of attention.THE FUTURE UNDER GLASS

Horizontal culture is broad, searchable and fast. It can produce a remarkably educated listener who knows

hundreds of microgenres, scenes and historical references. Yet that knowledge may become taxonomic. A band

is reduced to a cluster of descriptors. A film is explained through comparisons. A fashion look is named as a

"core." The category arrives before the experience has had time to become private. Larson complained that a

superabundance of music could reduce complex experiences to one-line tags, and the subsequent decade made

that mechanism central to cultural circulation. A platform works best when things can be recognized, labelled

and routed quickly. Ambiguity is expensive; legibility is efficient.

The personal relevance of Larson's work to this essay is also worth stating plainly. The author knows Larson,

and the interest here is not to elevate a friend's manifesto by sentiment but to take seriously an idea that has

aged unusually well. In 2011 or 2012, many of the mechanisms now associated with platform culture were

embryonic. Streaming had not yet become the default mode of music consumption in much of the world.

Short-form video had not become a dominant cultural engine. Generative AI was not a household creative tool.

The smartphone had not yet absorbed so many functions of social and cultural life. Yet the central architecture

Larson identified was already visible: the internet as externalized memory, the coexistence of multiple times, the

weakening distinction between original and copy, and the conversion of historical gestures into endlessly

reusable aesthetic material.[1][2]

There is also a productive tension in The Now Age. Larson is suspicious of linear historical progress, yet she

mourns the loss of the new. She rejects a simple modernist timeline while wanting art to regain the capacity for

ecstasy, symbolic force and genuine presentness. That tension is useful because it avoids a crude demand that

every decade must invent a new haircut or synthesizer sound. A culture can be temporally plural without being

inert. The real loss occurs when pluralism becomes passive recombination, when the historical archive supplies

ready-made identities faster than artists can develop new ones.

This helps explain why Ghost Modernism feels more accurate in 2026 than when it was coined. The archive

is now not only more complete but more active. Recommendation systems continually resurface older material.

Rights-holders reissue and remaster catalogues. studios mine intellectual property across sequels, prequels,

remakes and reboots. Social platforms create cycles in which a song from decades ago can become newly

ubiquitous because of a meme or television placement. Clothing from the recent past can return before its first

owners have discarded it. The dead do not merely remain accessible. They are recommender-system participants.

Larson's image of possession is therefore worth retaining. To be possessed by a ghost is not to consciously

study history. It is to have history act through you without fully recognizing the relation. Contemporary culture

often announces an aesthetic as spontaneous only for its genealogy to be reconstructed almost immediately by

collective online memory. The novelty is not false, but it is compressed by recognition. The present becomes

crowded with explanatory ancestors.

The Now Age ultimately proposes a spiritual answer to an informational problem: slow down, deepen

attention, restore symbolic meaning, and use the past as a source rather than a costume rail.[2] Whether one

accepts Larson's mystical framework is secondary. The cultural diagnosis is strong. A civilization with near-total

archival access must learn how to distinguish memory from possession.

3. Before Ghost Modernism: pastiche, retromania and lost futures

Larson's Ghost Modernism belongs to a larger intellectual lineage that predates the contemporary internet

but becomes more legible through it. Fredric Jameson's account of postmodernism, Simon Reynolds's

Retromania and Mark Fisher's writing on hauntology each describe a weakening of historical time from a

different angle. Taken together with The Now Age, they form a remarkably coherent sequence: pastiche turns

historical style into a reusable language; digital archives make those languages instantly available; retro becomes

normalized; and the future loses its power to appear radically different from the present.

Jameson's 1984 essay "Postmodernism, or, The Cultural Logic of Late Capitalism" identified pastiche as a

defining practice of postmodern culture. Pastiche, in his account, resembles parody stripped of parody's critical

distance. It imitates a dead style without necessarily mocking it. Jameson connected this to a broader weakening

of historicity in which the past becomes a collection of images and stylistic signs rather than a lived sequence of

social worlds. His discussion of the "nostalgia mode" is especially relevant: cultural works increasingly representTHE FUTURE UNDER GLASS

not the past itself but our accumulated images of the past.[3] A 1950s diner becomes red vinyl, chrome, jukebox

typography and particular music cues. The signs are intelligible even to people who never inhabited the period.

The internet did not create this mechanism. It industrialized access to it. Jameson wrote when archival

images still required television schedules, libraries, video rental, cinema repertory, magazines, record collections

and physical reproduction. By the 2000s, search engines, file-sharing networks, digitized archives and later

social video platforms turned historical style into a continuously queryable resource. The past was no longer

merely represented through commercial media. Users could browse it directly, or at least browse digitized traces

of it, at enormous scale.

Simon Reynolds's Retromania, published in 2011, diagnosed this transition from inside popular music.

Reynolds observed an increasingly intense fixation on reissues, reunion tours, anniversary performances, revivals

and historical rediscovery. He linked the sensation to his own internet use, particularly downloading out-of-

print music and trawling YouTube, where he experienced a kind of atemporality in which past and present

became intermingled.[4] The point was not that retro was new. Rock music had always revived earlier styles. What

changed was the density and simultaneity of revival. Instead of one historical period becoming fashionable,

multiple pasts remained active at once.

Reynolds's argument also helps clarify why the internet's democratization of archives can feel both

exhilarating and claustrophobic. For a music obsessive, the ability to locate an obscure record in seconds is a

genuine miracle. Cultural history becomes vastly more available. Lost work can be restored. Forgotten artists

can find new audiences. Marginalized scenes can be documented outside the institutions that once ignored

them. Yet this abundance also removes some of the friction through which taste was formed. The obscure ceases

to be geographically obscure. The difficult-to-find becomes a search result. Knowledge expands, but the

experience of discovery changes.

Mark Fisher pushed the argument from nostalgia toward futurity. Drawing on Jacques Derrida's concept of

hauntology and Franco "Bifo" Berardi's phrase "the slow cancellation of the future," Fisher argued that twenty-

first-century culture increasingly suffered from the disappearance of the future as an aesthetic sensation. In

interviews around Ghosts of My Life, he stressed that retro had become so normalized it was almost invisible.

Earlier postmodern appropriation still felt like a noticeable stylistic operation; later culture simply assumed the

continued availability of everything.[5] The issue was not that contemporary work was identical to older work,

but that it rarely generated the shock of temporal difference that earlier popular modernisms sometimes

achieved.

Fisher's claim is easiest to understand through an imagined experiment. Take a recording made in 1965 and

play it to someone from 1955; then take a recording from 1975 and play it to someone from 1965; continue

through punk, hip-hop, house, techno, jungle and the early digital turn. There are points at which a ten-year

jump produces a striking change in sonic world. Perform the same experiment across many parts of mainstream

culture after 2000 and the distinctions are often subtler. Production becomes cleaner, louder, more

computational and more hybrid, but older stylistic languages remain continuously available. A song released in

2026 can plausibly sound like 1986, 1996, 2006 or a composite of all three without being treated as a novelty in

itself.

This should not be romanticized into a false story of continuous twentieth-century revolution followed by

twenty-first-century stasis. Earlier decades contained vast amounts of conservative, repetitive and formulaic

culture. Many supposed period styles were recognized only retrospectively. The 1980s did not feel like a unified

costume party to people living through them. Moreover, contemporary music contains innovations in

production, distribution, genre hybridization and global exchange that older listeners may fail to perceive

because novelty is often most obvious to the young. A serious account must preserve these qualifications.

Still, Jameson, Reynolds, Fisher and Larson converge on a real structural shift. Jameson describes the

conversion of history into image. Reynolds describes the archive becoming instantly available. Fisher describes

the weakening of temporal difference. Larson describes the resulting forms as ghosts and zombies, emphasizing

the loss of symbolic force. Their vocabularies differ, but all identify the same basic problem: culture's

relationship to the past has become increasingly immediate, while its relationship to the future has become

harder to imagine.

The internet intensifies this because it converts cultural memory from an occasional activity into a

continuous environment. One no longer goes to an archive. The archive arrives in the feed. It is recommendedTHE FUTURE UNDER GLASS

between contemporary objects, remixed into memes, attached to anniversary posts, restored in high definition,

sampled in new tracks, recut into short clips and resurfaced by users whose age may be half that of the material.

This is historical democratization, but it is also historical saturation.

The result is not the end of time. It is the weakening of temporal hierarchy. The old does not withdraw

enough to become fully old. The new does not remain isolated long enough to become fully strange. Every era

becomes a palette.

4. The archive becomes an environment

For most of human history, cultural memory was selective because storage was expensive and retrieval

difficult. Oral traditions depended on repetition and communal memory. Manuscripts required material, labour

and institutions. Printed books expanded preservation but still depended on libraries, private ownership and

physical distribution. Recorded sound and film multiplied the archive again, yet access remained constrained by

stock, geography, price and formats. Even the mass-media twentieth century preserved forgetting. Television

programs disappeared after broadcast; records went out of print; magazines were thrown away; photographs

remained in albums; foreign scenes were difficult to observe from afar. Culture was abundant by historical

standards but still materially lossy.

Digital networks changed the economics of retention. Once a cultural object becomes data, copying

approaches zero marginal cost. Search indexes what storage preserves. Platforms make circulation perpetual. The

consequences are overwhelmingly positive in many domains. Digitization has reduced the cost of producing and

distributing music, television, books and film, contributing to large increases in the amount of creative work

available.[6][18] Independent artists can publish globally without manufacturing stock. Small audiences can sustain

niche work. Historical material that once required specialist knowledge can be encountered by anyone with

access to a network. The long tail is real, and any argument about cultural stagnation that ignores these gains

becomes nostalgia masquerading as analysis.

Yet preservation and access change creativity because creators are also consumers. Every young filmmaker

now grows up with potential access to an enormous film history. Every musician can study production tutorials,

isolated stems, plugin emulations, gear demonstrations and fan reconstructions. Every designer can assemble

thousands of visual references before making a mark. The reference library that once belonged to large studios,

schools or obsessive collectors is distributed across ordinary devices. This represents a radical democratization of

expertise.

It also means that the conditions of creative ignorance are disappearing. Ignorance has obvious costs,

particularly when it excludes people from education or reinforces provincial prejudice. But partial ignorance can

produce stylistic independence. Scenes become distinctive partly because they do not know everything

happening elsewhere. They solve problems locally. They misuse tools. They imitate distant influences

imperfectly. They develop conventions that look unnecessary to outsiders because those conventions emerged

from particular rooms, budgets, climates, technologies and social groups.

Consider the historical importance of error. A copied record deteriorates. A fashion style described second-

hand is reconstructed incorrectly. A regional band hears a genre through a handful of recordings rather than a

canonical playlist. A filmmaker with limited equipment tries to reproduce a visual effect and accidentally

invents another. Such errors are not merely failures of information. They are cultural mutations. Perfect access

reduces some kinds of mutation because the original reference is always available for correction.

The contemporary mood board illustrates the shift. Digital image search and visual bookmarking allow a

creative team to assemble a highly precise cloud of references in minutes. This can produce sophisticated work,

but it also makes historical style unusually reproducible. A room can be designed to communicate "1970s

California," "late-1990s corporate futurism" or "mid-century Italian" with extraordinary accuracy because the

corpus is visible. The style becomes a toolkit. The better the archive, the easier it becomes to imitate the surface.

Larson's term "zombie aesthetics" describes the risk that fidelity to surface increases while contact with the

social conditions that produced the surface decreases.[1] The style survives but the original pressure is gone. Punk

clothing can be manufactured by luxury brands without unemployment, youth clubs, cheap rehearsal rooms or

the antagonistic social ecology from which punk scenes emerged. Early web graphics can be revived without

dial-up speeds, browser limitations or the optimism and awkwardness of a genuinely new medium. A cassetteTHE FUTURE UNDER GLASS

can be purchased as a deliberate inconvenience by someone carrying a device capable of storing millions of

songs. None of these revivals is illegitimate. Their meaning is simply different.

This creates a culture of selectable temporality. Historical styles become options much like fonts or filters.

The user can choose 1969, 1987, 1996 or 2003 according to mood. Streaming interfaces intensify the experience

because chronology is spatially collapsed: a song from 1972 and a song from 2026 occupy identical rectangles on a

screen. The difference between them may be audible, but the interface treats them as equivalent units of

content. The platform's basic ontology is not historical period but availability.

Availability encourages another transformation: the archive becomes anticipatory. Recommendation

systems do not wait for users to seek the past. They surface it. An old song may be inserted into a personalized

playlist because the system predicts compatibility with recent listening. A decades-old film appears beside new

releases because a service has licensed it or because a user watched something related. A short-form platform can

transform an old recording into a contemporary hit by repeatedly attaching it to new videos. The archive is

therefore not static. It competes for current attention.

This competition changes the economics of newness. A new cultural object no longer competes only with

other new objects. It competes with the best surviving work from history, much of it available at the same price

and often through the same subscription. A new album enters a marketplace where a listener can immediately

choose any canonical album from the past sixty years. A new television series competes not only with current

programming but with entire completed libraries. The abundance is wonderful for audiences, but brutal for

producers. Familiar intellectual property gains an advantage because recognition cuts through abundance.

The same logic applies psychologically. Faced with effectively infinite choice, a known object reduces

decision cost. This helps explain why archival abundance can coexist with conservative consumption. The

catalogue is infinite, but attention is finite. Recommendation systems exist because abundance creates a search

problem. Their success depends on predicting what will satisfy quickly. Prediction, by definition, uses the past to

organize the future.

The archive has therefore become both memory and infrastructure. It preserves culture, supplies creative

reference, structures discovery, lowers risk and competes with the present. We inhabit it before we consciously

decide to study it. The question is no longer whether artists will use history. It is how any artist can avoid being

continuously addressed by it.

5. Platform time: when trends arrive already exhausted

The most visible symptom of flattened novelty is the acceleration of the trend cycle. Fashion provides the

clearest examples because clothing is both material and highly legible. Social platforms can now identify, name

and distribute microtrends at extraordinary speed. A set of loosely related garments and references becomes an

"aesthetic," receives a searchable label, is reproduced by creators, translated into retail products and then

declared exhausted, sometimes within a season. Vogue has described the rise and apparent fatigue of hyper-

specific microtrends and the growing desire to move back toward personal style and physical-world context.[12]

The remarkable feature is not merely speed. It is the degree to which a trend can become self-conscious almost

immediately.

Traditional subculture depended partly on delayed recognition. A scene formed around clubs, record shops,

schools, neighbourhoods, magazines or particular cities. Participants developed codes before mass media fully

understood them. By the time journalists named the style, the people inside it might already be moving

elsewhere. The lag created an incubation period in which a language could become dense. Platforms shorten this

lag. A distinctive look can be captured, circulated globally and labelled while it is still forming.

Naming is culturally powerful because it makes a style searchable. Once a label exists, the archive can be

assembled around it. Tutorials explain how to reproduce it. Brands identify the relevant products.

Recommendation systems route the label toward users whose behaviour suggests receptivity. The aesthetic ceases

to belong primarily to a scene and becomes a package. This does not mean the participants are inauthentic. It

means their environment industrializes recognition.

The acceleration produces a strange temporal reversal. Earlier trends often took years to diffuse from

subculture to mainstream retail. Contemporary microtrends can be manufactured for mass purchase before theyTHE FUTURE UNDER GLASS

have developed enough internal complexity to constitute stable subcultures. The category becomes exhausted

because its visible signs are replicated faster than its meanings can deepen. Saturation arrives before maturity.

This helps explain the increasingly common desire to avoid being "too online." In 2026, reporting on

fashion and personal taste described young adults questioning whether their preferences were genuinely theirs or

the result of algorithmic exposure, while explicitly connecting renewed interest in physical magazines, vinyl,

film cameras and offline references to a desire to escape the same digital pools of imagery.[10] That anxiety reveals

a new kind of cultural self-consciousness. People have always been influenced by peers and media. The

contemporary difference is awareness that influence can be computationally personalized, measured and

repeated until a preference feels spontaneous.

Recommendation algorithms complicate the picture because they can both narrow and widen exposure.

Spotify research has found that algorithmically programmed listening is, on average, less diverse than users'

organic listening, and that users who become more diverse over time often shift away from programmed

consumption.[7] Other work, however, shows that recommendation systems can increase aggregate diversity,

expose users to lower-ranked material or increase the variety available to individuals depending on the design of

the system.[8] More recent research on generative playlist tools suggests that AI-driven recommendations can

redistribute attention from top-ranked tracks toward less prominent material under some conditions. The

empirical picture is therefore not a simple filter-bubble story.

The more useful cultural point is that recommendation changes the mode of discovery. A discovery made

through wandering, friendship, accident, local context or sustained research feels different from one delivered

because a system predicts compatibility. Both may introduce genuinely unfamiliar work. Yet the latter is

designed to minimize friction. It attempts to make the unfamiliar acceptable by connecting it to what is already

known.

This is where novelty can become semantically narrow even when inventory diversity increases. A user may

hear hundreds of new songs while remaining within a stable emotional and stylistic corridor. Platforms are

extraordinarily good at producing novelty-of-item: something the user has not previously consumed. They are

less naturally suited to novelty-of-world: something that destabilizes the user's categories and forces a new form

of attention. The second kind is risky because it can generate immediate rejection, which engagement systems

often interpret as failure.

Platform time is therefore optimized for the near future. It predicts the next click, the next track, the next

video, the next object likely to retain attention. Cultural movements, by contrast, sometimes require periods in

which audiences do not immediately understand what they are hearing or seeing. Modernism frequently

depended on difficulty, refusal and delayed acceptance. A system designed to reduce abandonment has little

incentive to cultivate bewilderment.

The compression of trends also affects memory. Because styles are named and archived instantly, they can

become nostalgic almost immediately. Aesthetic categories from five years ago may be revived as if they

belonged to a distant period. Young users can encounter their own childhood interfaces as retro design before

reaching adulthood. The path from present to archive has shortened. The culture metabolizes itself at high

speed.

This may be one reason the 2020s often feel simultaneously frantic and static. The surface changes

constantly: new memes, microgenres, cores, filters, formats and personalities appear every week. Yet the

underlying structure remains familiar. Aesthetic novelty is produced in small rapid bursts while epochal novelty

becomes harder to perceive. The carousel spins quickly, but it rotates around a stable axis.

The system rewards this because constant minor difference sustains attention. A platform does not require a

new cultural era. It requires a continuous supply of content distinguishable enough to prompt another

interaction but familiar enough to be immediately legible. In that sense, microtrend culture is perfectly adapted

to the economics of feeds. It produces the sensation of change without necessarily producing historical rupture.

The flattening of novelty is therefore compatible with extreme velocity. Indeed, speed may be one of its

causes. A genuinely new form needs time to acquire internal rules, institutions, myths, failures and counter-

movements. When every emerging form is instantly observed, reproduced and monetized, it may be denied the

obscurity in which strangeness becomes durable.THE FUTURE UNDER GLASS

6. The economics of recognition: why old intellectual property keeps

returning

If platform culture explains the acceleration of aesthetics, the economics of film and television explains the

persistence of familiar intellectual property. The dominance of sequels, remakes, adaptations and franchises is

often described lazily as evidence that studios have "run out of ideas." That formulation mistakes a risk strategy

for a creative deficiency. The entertainment industry contains no shortage of original scripts, characters or

concepts. The problem is that large-scale financing increasingly rewards recognition.

Theatrical film is a useful case because the financial exposure attached to global releases is enormous. A

studio may spend hundreds of millions of dollars producing and marketing a tentpole. In a market fragmented

by streaming, gaming, social video and home entertainment, persuading audiences to leave home is difficult.

Existing intellectual property provides an immediate communication advantage. A familiar title performs

marketing before the campaign begins. It brings an installed audience, recognizable iconography, search

interest, licensing opportunities and international awareness. In an environment of attention scarcity,

familiarity is an economic asset.

The 2025 box office illustrates the pattern. The highest-grossing North American releases included Zootopia

2, A Minecraft Movie, Lilo & Stitch, Avatar: Fire and Ash, Superman, Wicked: For Good, Jurassic World

Rebirth, The Fantastic Four: First Steps and How to Train Your Dragon.[14] Most were sequels, adaptations,

remakes or extensions of existing brands. Industry analysis published for 2026 noted that the theatrical market

from 2016 through 2024 was dominated by franchise and adapted properties, although 2025 also produced

encouraging original successes such as Sinners and Weapons.[13] The coexistence of these facts matters. Originality

has not vanished, but familiarity occupies the safest and most expensive tier.

A broader dataset assembled in 2026 suggested a long-term decline in the representation and box-office

share of original screenplays among the fifty highest-grossing North American films, with existing intellectual

property taking a much larger portion of the top market than in the 1990s.[13] Such datasets require careful

classification because "original" can be defined in different ways, and an adaptation may be artistically radical

while an ostensibly original film may be deeply derivative. Nonetheless, the business logic is visible without

needing a perfect metric: studios repeatedly allocate their largest budgets to properties whose names already

circulate in the cultural archive.

Streaming complicates the story because the total quantity of original programming has expanded

dramatically. Research on digitization and the "random long tail" argues that lower distribution costs can

generate more creative products and allow unpredictable high-quality work to emerge.[18] Parrot Analytics has

also noted that the proportion of newly premiered titles directly based on existing IP is not as overwhelming as

public perception sometimes suggests, particularly when the entire supply of film and television is counted

rather than only theatrical blockbusters.[15] This is an essential corrective. The culture is not literally composed

only of remakes. The concentration occurs most visibly in the high-attention, high-budget centre.

That centre matters symbolically because blockbusters are collective cultural events. When the largest

screens, marketing campaigns, toys, theme parks and media conversations are dominated by long-running

brands, the public face of entertainment becomes archival. New work may flourish at the edges while the centre

remains populated by inherited universes. The result is a two-speed culture: extraordinary proliferation below,

conservative investment above.

The same logic appears in television through spin-offs, legacy sequels and adaptations, though streaming has

also produced many new properties. It appears in gaming through remakes, remasters, sequels and long-lived

service franchises. It appears in publishing through celebrity brands and established series. It appears in music

through catalogue acquisition, deluxe reissues and anniversary tours. The particular economics differ, but the

recurring incentive is recognizability under conditions of abundance.

This is where Ghost Modernism intersects with capitalism more directly. The archive is valuable because it

reduces uncertainty. A familiar property is a ghost with measurable awareness. Its prior life supplies data.

Executives know who recognizes it, where it has performed, how it can be merchandised and which

demographics may respond. An entirely new idea lacks this history. The old property therefore functions as a

financial instrument whose cultural memory can be converted into lower perceived risk.THE FUTURE UNDER GLASS

The danger is not simply repetition. Repetition can produce excellent art. Some sequels deepen themes, some

remakes reinterpret old material, and some adaptations introduce texts to audiences who would never encounter

them otherwise. The deeper risk is ecological. When institutions repeatedly reward familiarity at the highest

budget levels, creators learn that novelty must arrive disguised as recognition. An original project is pitched as a

combination of existing hits. Marketing explains the unknown through the known. New worlds are encouraged

to behave like established brands before they have earned the density of history.

The franchise model can also change narrative form. A story becomes less a bounded work than a node in an

expandable system. Endings become openings. Supporting characters become future properties. Backstory

becomes prequel material. A successful object is not merely enjoyed; it is mined. This industrializes what the

internet does culturally: it resists disappearance. Nothing successful is allowed to end cleanly because the archive

contains unrealized future revenue.

This produces what might be called compulsory afterlife. The cultural object is kept alive because

recognition has economic value. Characters return after decades. cancelled series are revived. films are rebooted

for viewers who remember the original and simultaneously introduced to viewers born long afterward. The

distinction between nostalgia market and youth market collapses because the same property can address both

groups differently.

The situation can feel like a loop because the industry's most visible output continually validates memory.

Yet the loop is not proof that audiences reject the new. The success of original films when they receive sufficient

support shows the opposite.[13] The more plausible explanation is institutional risk management. The culture has

not lost the capacity to generate new intellectual property. Large institutions have become reluctant to bet their

most expensive resources on it.

That distinction matters politically and aesthetically. If the problem were exhausted imagination, little

could be done. If the problem is an incentive structure, then different financing, distribution and exhibition

systems could change it. A new era does not require humanity to invent creativity again. It may require

institutions willing to tolerate uncertainty long enough for unfamiliar things to become familiar.

7. More culture, less epoch: the difference between abundance and

distinctiveness

The strongest objection to any argument about cultural stagnation is empirical abundance. There are more

albums, books, series, films, games, podcasts, illustrations, photographs and videos being produced than any

individual could consume. Digital tools have reduced production costs and gatekeeping. Independent creators

can distribute globally. Research by economist Joel Waldfogel and colleagues has repeatedly argued that

digitization expanded the supply of creative work and generated substantial consumer benefit, including high-

quality products that older industrial systems might never have financed or distributed.[6][18] If novelty were dying,

why would cultural production be exploding?

The answer requires separating quantity, diversity and epochal distinctiveness. These are not the same

variable. A system can produce more works, support more niches and contain more stylistic variety while still

making it harder for any single style to define the period. Indeed, abundance may itself dissolve the dominance

required for a coherent era to become visible.

The twentieth-century "decade" was partly an artefact of concentrated media. When millions of people

watched the same television networks, heard the same radio rotation, shopped in similar chains and

encountered the same magazines, cultural changes could become synchronized. A new sound or silhouette

spread through a relatively narrow pipeline and therefore seemed to transform the atmosphere. Fragmentation

weakens that synchrony. Millions of people can now inhabit different aesthetic worlds simultaneously. One

person's 2026 is hyperpop and generative video; another's is classic rock, heritage clothing and physical media;

another's is Korean pop, anime, gaming streams and 2000s nostalgia. The period may be highly creative but

impossible to summarize with a single look.

This is an important correction to the nostalgia of people who remember the twentieth century. The feeling

that decades used to be more distinct may partly reflect the fact that media institutions forced more peopleTHE FUTURE UNDER GLASS

through common channels. Greater diversity can make an era feel less coherent precisely because fewer people

share the same cultural clock. A pluralistic culture may sacrifice some legibility while gaining freedom.

Yet fragmentation alone does not explain the sense of repetition. The niches themselves often draw heavily

on archives. Digital abundance multiplies the number of combinations while keeping historical reference

continuously visible. This produces a culture that can be astonishingly diverse at the level of individual objects

and strangely familiar at the level of stylistic grammar. The number of possible collages approaches infinity, but

the source images remain recognizable.

Music recommendation research captures a similar ambiguity. Streaming adoption has been associated with

increases in the quantity and diversity of music people consume, particularly because experimentation becomes

cheaper when every additional track costs nothing.[6] At the same time, research on Spotify has found that

algorithmically programmed listening can be less diverse than organic listening for many users.[7] A 2025

Scientific Reports study complicated the filter-bubble narrative further by showing that algorithmic curation

can generate different diversity effects depending on scale and measurement.[8] There is no single verdict because

"diversity" can mean many things: number of artists, genre distance, popularity distribution, national origin,

semantic novelty or aggregate variety across the population.

The cultural question is different again. A person can listen to a highly diverse playlist containing Nigerian

Afrobeats, Japanese ambient music, Brazilian funk, American hip-hop and Scandinavian metal, yet experience

each as an equally accessible tile within the same platform. Global access increases diversity of content while

standardizing the mode of encounter. The interface becomes a universal container.

This distinction between content diversity and experiential diversity is central to the flattening thesis. The

internet can expose people to cultures that older media systems ignored, a profound gain. Yet the experience of

discovery is increasingly mediated by the same few interaction patterns: scroll, tap, swipe, like, skip, save, follow.

The world becomes culturally larger while the interface becomes culturally smaller.

Something similar happens to creative production. Software democratizes sophisticated techniques by

packaging them into presets, templates and plugins. A bedroom producer can access sounds once requiring an

elite studio. A small design team can produce effects once reserved for large agencies. This is a genuine

redistribution of power. But when millions use the same software ecosystems, marketplaces and tutorial

channels, production methods can converge. Distinctive local limitations are replaced by global defaults.

The flattening of novelty therefore should not be measured by asking whether the present contains new

work. It should be measured by asking how easily the new can establish a world around itself. Does a new form

produce institutions, spaces, rituals, technologies and modes of attention that make it feel historically specific?

Or does it arrive as another option inside existing infrastructure?

Consider the difference between a genre and a playlist category. A genre historically may arise from clubs,

scenes, equipment, economic conditions, racial and class histories, dress codes, local slang, labels, radio shows

and physical spaces. A playlist category can extract the sonic traits while leaving much of that world behind. The

music still matters, but the ecology is thinner. The platform is excellent at distributing tracks and poor at

reproducing the social density from which movements emerge.

The same can be said of film. A wave such as Italian neorealism, the French New Wave or New Hollywood

involved technologies, institutions, politics, criticism and shared artistic antagonisms. Contemporary cinema

contains many brilliant films, but global streaming can absorb them as individual titles without requiring

viewers to inhabit a movement. The masterpiece survives while the movement becomes optional context.

This may be why the culture can feel simultaneously richer and flatter. Richer because the archive and

present are vast. Flatter because differences coexist on the same plane of availability. The hierarchy between eras,

scenes and media weakens. Everything becomes content, which is another way of saying everything becomes

comparable.

A useful analogy is geography. Air travel and digital communication do not erase cities, languages or local

cultures, but they reduce the friction between them. Airports begin to resemble one another. Global chains

spread. Local difference remains but exists within shared systems. Cultural digitization works similarly. It does

not destroy difference; it changes the background against which difference is experienced.

The challenge, then, is not to restore a narrow twentieth-century monoculture merely so future historians

can identify 2026 by hairstyle. The challenge is to preserve conditions in which genuinely new forms canTHE FUTURE UNDER GLASS

develop enough depth to become more than combinations. Abundance should be compatible with

distinctiveness. Diversity should not require permanent atemporality. A culture can contain many worlds

without forcing them all into the same temporal room.

8. Generational excavation: when the recent past becomes exotic

One of the most revealing cultural developments of the mid-2020s is the speed with which the recent past

has become exotic to people born after it. The 1990s and early 2000s are now sufficiently distant from teenagers

and young adults to function as historical environments, yet sufficiently recent to remain legible. The result is a

form of generational excavation in which the material culture of Millennials and late Gen X is rediscovered by

people who did not experience it as ordinary life.

This differs from autobiographical nostalgia. Someone born in 1988 may associate a CD wallet, a CRT

television, a video rental shop, a dial-up modem, a shopping-centre food court or a late-1990s coffeehouse

interior with childhood. Someone born in 2008 encounters the same objects as evidence of another mode of

existence. The attraction can therefore be stronger in certain respects because it is not burdened by the

inconveniences that accompanied the original technology. The young observer can select the romance without

having waited for a modem to connect or carried a folder of scratched discs on a bus.

Yet the appeal is not reducible to style. Pre-smartphone life contains structural features that have become

conspicuous through absence. Media was often material. Music occupied shelves. Photographs had to be

developed or transferred. A person could leave home without being continuously reachable. Television schedules

created shared timing. Video shops and record stores imposed finite choice. Shopping centres functioned as

social spaces rather than merely logistics nodes. A bedroom displayed taste because cultural consumption

required objects: posters, books, games, magazines, discs, tapes, photographs, ticket stubs and devices.

The smartphone collapses much of that biography into a black rectangle. The same object contains camera,

music library, map, television, radio, newspaper, letters, diary, games, tickets and social life. From the

standpoint of efficiency this is extraordinary. From the standpoint of material culture it produces a kind of

visual silence. Two bedrooms may contain radically different digital lives while appearing materially similar.

This helps explain the fascination with objects whose functions have been absorbed by phones. Film cameras,

compact digicams, MP3 players, wired headphones, CDs, cassettes and printed magazines offer not only retro

appearance but separation of functions. Each object does less. The limitation gives it identity. A camera that

only takes photographs does not interrupt the act with messages. A CD contains one album rather than every

album. A magazine has an end. A DVD remains available even if a licence expires. The object imposes

boundaries that the network removed.

Recent survey and reporting evidence supports the broader analogue turn. Young consumers have shown

renewed interest in physical music, film photography and older devices, often citing authenticity, ownership or

relief from digital saturation rather than straightforward nostalgia.[11] Pew Research Center data also show that

heavy internet use remains pervasive among teenagers while a substantial share increasingly say they spend too

much time on social media, and many report attempts to cut back.[9] These findings should not be converted into

a fantasy that Gen Z is abandoning the internet. Nearly all teenagers remain daily users. The analogue revival is

a countercurrent inside digital life, not an exodus from it.

Its symbolism nevertheless matters. A generation born into abundance is experimenting with scarcity

voluntarily. This is historically unusual. Previous consumer technologies were sold primarily by promising

convenience: more songs, faster access, lighter devices, larger libraries, fewer cables, instant photography,

unlimited storage. The revival of older formats reverses the value proposition. Slowness, limitation and friction

become desirable because they distinguish an activity from the general stream of computation.

This is why the return of a 1990s aesthetic such as Global Village Coffeehouse can produce more than

decorative nostalgia. Warm colours, dark timber, murals, handmade-looking graphics, mismatched textures

and physical clutter signify an era before the contemporary minimalist platform aesthetic became dominant.

Younger observers may read such spaces as evidence that commercial interiors once tried to possess a visible

personality. The accuracy of that impression can be debated, and 1990s chain environments were as commercial

as contemporary ones, but the contrast itself is culturally productive. The past becomes a vocabulary for

criticizing the present.THE FUTURE UNDER GLASS

Generational excavation also changes the status of Millennials. A generation that still experiences itself as

contemporary begins to watch its childhood treated as archival material. Objects that were once disposable

become collectible. Interfaces that once looked banal become design references. The transformation can feel

absurdly premature because the people who used the objects are only in their thirties and forties. Yet digital

culture accelerates archival distance. Images of the past remain continuously available, ready to be reorganized

into aesthetic categories long before the people represented in them are old.

The phrase "Millennial core" captures part of this process, but the more interesting phenomenon is temporal

tourism. Younger users can visit a version of the late twentieth century through curated fragments: camcorder

footage, catalogue scans, old commercials, mall photography, archived websites, television clips, family

photographs and restored games. The period becomes explorable as a dataset. It is simultaneously intimate and

artificial.

The danger is obvious. The historical world can be aestheticized until its politics, exclusions, boredom and

inconveniences disappear. The 1990s become chunky computers, warm cafés and carefree children rather than

recession, war, discrimination and unequal access to technology. Every nostalgia mode performs selective

editing. Jameson's warning applies: we risk relating to history through images of history rather than history

itself.[3]

But younger people's attraction to the era can still reveal genuine dissatisfaction with present conditions. The

longing for a world with fewer notifications, more third places, physical ownership and clearer boundaries

between activities is not invalid merely because the remembered period was imperfect. Nostalgia often functions

diagnostically. It identifies what people feel is missing, even when it misidentifies where that missing thing once

existed.

The question is whether generational excavation can become more than costume. If younger people recover

the social functions of older forms rather than only their surfaces, the revival may become inventive. A record

shop is interesting not because vinyl sleeves photograph well but because it can produce repeated local

encounters. A film camera is interesting not only because grain looks nostalgic but because delayed results

change attention. A printed magazine matters not only as an object but because its finite editorial sequence

resists endless personalized feeds.

The past becomes useful when it teaches constraints rather than merely supplying aesthetics. In Larson's

terms, this would be Channeling rather than zombie imitation: returning to a source in order to intensify the

present.[1]

9. The return to matter: physical media, third places and deliberate

friction

The renewed interest in physical media is one of the clearest signs that digital convenience has reached a

point where some consumers actively seek counterweights. Reporting in 2026 found people returning to CDs,

DVDs, vinyl and cassettes for reasons that included ownership, privacy, artist support, independence from

algorithms and the pleasure of deliberate interaction.[11] The motivations are revealing because they are not

exclusively nostalgic. Many describe a desire to control selection rather than have content continuously

presented to them.

Physical media changes behaviour through friction. To play a record, one chooses a specific object, removes

it from a sleeve, places it on a turntable and usually listens for a longer interval than a single streamed track. A

CD requires less ritual but still narrows the field to a finite sequence. A DVD turns a film into something owned

rather than temporarily licensed. These formats do not guarantee attentive consumption, and digital listeners

can of course choose albums deliberately. The point is architectural: the older medium builds certain limits into

the default experience.

Streaming reverses this architecture. The library is effectively infinite, switching is immediate and the

marginal cost of trying another item is zero. These features increase discovery and consumer welfare.[6] They also

make abandonment effortless. The listener can skip before a song establishes itself. The viewer can browse for

longer than they watch. The abundance creates a paradoxical form of scarcity: the scarce resource becomes

commitment.THE FUTURE UNDER GLASS

Larson's contrast between horizontal and vertical listening is useful here.[2] Physical media does not

automatically create verticality, but it can make vertical attention easier by reducing the number of visible

alternatives. The record on the turntable temporarily becomes the world. The streaming service, by contrast,

continuously reminds the listener that another world is one tap away.

This principle extends beyond media. The renewed discussion of "third places" reflects concern that social life

has also been optimized toward frictionless private consumption. A third place, in the classic sociological sense,

is neither home nor workplace but a recurring public environment in which people can spend time and

encounter others: cafés, pubs, libraries, parks, community centres, bookstores and similar settings. Digital

services can replace many practical reasons to visit such places. Shopping arrives at home, entertainment streams

privately, banking occurs through apps, work can be remote and friendship can be maintained through

messaging. Each convenience removes a journey, but journeys create incidental contact.

The desire for more physical social space among younger people therefore belongs to the same cultural

movement as the analogue revival. It is a response to a world in which too many activities share the same device

and location. When work, leisure, socializing, dating, shopping, reading and entertainment all occur through a

phone or laptop, the boundaries that once distinguished them become psychological rather than physical. A

café, library or record store gives an activity a place.

This matters for cultural novelty because scenes need places. Many historically influential movements were

not invented by isolated geniuses but by recurring groups who occupied specific venues, shops, studios, schools

and neighbourhoods. Physical proximity creates accidental collaboration and disagreement. People encounter

material they did not select. Taste is influenced by the person behind the counter, the band on before one's own,

the poster on a wall, the conversation overheard at a table. Such discovery is not personalized and therefore can

be inefficient. Its inefficiency is part of its value.

Online communities can also generate powerful scenes, particularly for people excluded from local

environments. They should not be treated as inferior substitutes. The internet has allowed geographically

dispersed minorities and niche creators to find one another in ways that transformed lives and artistic

production. The more useful argument is ecological: culture benefits from multiple forms of association. When

one infrastructural model dominates, its biases become cultural defaults.

Deliberate friction is therefore not an anti-technology doctrine. It is a design principle. A healthy cultural

environment might preserve some activities that cannot be instantly optimized. Phone-free performances,

limited physical editions, local workshops, independent radio, public reading rooms, analogue photography,

community cinemas and long-form publications all create temporal boundaries. Their value lies partly in

refusing the universal interface.

The contemporary fascination with offline life can easily become another online aesthetic, and this is the

irony the movement cannot escape. A phone-free café becomes famous through photographs circulated online.

A cassette revival depends on internet shops and social media discovery. A club that bans recording may still

market itself through clips. The network is extraordinarily good at absorbing its own critique.

This does not make resistance meaningless. Cultural systems are rarely escaped absolutely. The goal is not

purity but altered ratios. A person can use the internet to discover a book and then spend ten uninterrupted

hours reading it. A community can organize online and meet physically. A musician can distribute digitally

while composing with tools that demand slow attention. The archive can lead toward depth rather than endless

browsing.

The word "friction" is often negative in technology because good interfaces minimize unnecessary effort. Yet

some forms of effort are constitutive rather than wasteful. Walking through a library shelf exposes adjacent

books. Waiting for a photograph changes anticipation. Learning an instrument builds bodily knowledge that

cannot be downloaded. Repairing an object creates familiarity with its construction. Remembering a route

strengthens spatial understanding. These are not inefficiencies surrounding the experience. They are part of the

experience.

A culture that wants novelty may need similar friction. It may need spaces in which ideas can develop

without immediate global comparison, where creators cannot instantly check whether somebody has done

something similar, and where small scenes can become strange before they become visible. The physical-mediaTHE FUTURE UNDER GLASS

revival is interesting because it suggests people are already experimenting, consciously or not, with the

conditions that make such depth possible.

The analogue turn therefore should not be understood as a desire to reverse history. Few people genuinely

want to restore the inconveniences of 1995 wholesale. What they want are selective properties that digital

abundance weakened: ownership, privacy, boundedness, tactility, locality and attention. Those properties can be

incorporated into future systems without pretending the internet never happened.

10. The algorithm and the problem of personal taste

Personal taste once developed through an unstable mixture of family, friends, geography, class, education,

local institutions, mass media and chance. None of this was autonomous. Radio programmers, critics, record

labels, advertisers and retailers shaped what people encountered. The mythology of a completely self-created

taste is therefore false. What changes under platform culture is the visibility and precision of mediation.

Recommendation systems learn from behaviour. They observe what a user watches, skips, saves, replays,

searches for or purchases, then compare those signals with patterns across large populations. The result can feel

intimate because the system often predicts preference accurately. Yet prediction introduces a recursive problem.

The platform recommends based on past behaviour; the user consumes the recommendation; the new

consumption becomes evidence for future recommendations. Preference is both measured and produced.

This loop complicates the meaning of authenticity. A person may genuinely love a style discovered through

an algorithm. The fact that a system introduced it does not make the response false. But when the same

recommendation architecture repeatedly presents similar material, the user can reasonably wonder whether

taste is being narrowed by exposure. Recent cultural reporting has captured this anxiety directly, with young

people asking whether their preferences are personal or simply reflections of the same online reference pools.[10]

Empirical research resists simple conclusions. Spotify studies have found that algorithmically driven

listening can reduce within-user diversity compared with organic listening, while other experiments show

personalized recommendations can increase aggregate diversity or be redesigned to introduce more varied

material.[7][8] The outcome depends on what the algorithm optimizes, how diversity is measured and how users

behave outside recommendations. Algorithms are not inherently homogenizing machines; they are incentive

systems implemented through particular objectives.

The cultural concern is that the dominant objective is usually engagement. Platforms benefit when users

remain active. Recommending something radically alien may be culturally valuable but commercially risky if

the user leaves. The system therefore has reasons to explore near the edge of preference rather than far beyond it.

It can surprise without disorienting.

This distinction mirrors the difference between innovation and novelty-of-item. A recommendation

engine can expose a listener to thousands of songs they have never heard. It may be excellent at discovery. Yet

because the selections are predicted from an existing taste profile, the discoveries can form a smooth extension

of identity. The user becomes more knowledgeable about themselves rather than being forced to become

someone else.

Historically, cultural transformation often came from involuntary exposure. A friend controlled the car

stereo. A sibling owned the television. A school library stocked strange books. A support act played before the

band one came to see. A radio station mixed material according to somebody else's priorities. Local scarcity

meant accepting imperfect options. These environments could be annoying and exclusionary, but they also

produced collision.

Personalization reduces collision. The interface becomes increasingly customized, which can make culture

more satisfying and less communal at the same time. Two people using the same service may inhabit entirely

different catalogues. The shared mainstream weakens. This is liberating for niche interests but makes collective

cultural epochs harder to form. A decade once sounded a certain way partly because mass media imposed shared

repetition. Personalized culture may be more individually diverse while socially asynchronous.

The paradox is that personalization can coexist with visual sameness. Millions of users receive individualized

feeds drawn from the same global pools. Aesthetic motifs that perform well can therefore spread extremelyTHE FUTURE UNDER GLASS

quickly, even if each person's sequence differs. The culture feels personalized at the level of feed and standardized

at the level of platform grammar.

The flattening of novelty is intensified by metrics. Creators can see what performs. They learn which hooks

retain attention, which thumbnails attract clicks, which formats increase watch time, which tags bring

discovery. Feedback that once arrived slowly through sales, reviews or word of mouth becomes immediate and

quantitative. This can help creators understand audiences, but it also encourages optimization.

Optimization is not censorship. Nobody needs to order artists to converge. If thousands independently adjust

work toward the same measurable signals, convergence can emerge from rational behaviour. A song becomes

shorter because completion matters. A video begins with a hook because retention drops quickly. A title

becomes more explicit because search matters. A visual style becomes high-contrast because it reads on a phone.

The platform does not dictate the aesthetic; the incentive landscape selects it.

This is another reason apparently radical cultural material can feel structurally familiar. The content may

vary wildly while the delivery obeys common rules. The internet expands expressive possibility but concentrates

distribution through a small number of attention architectures.

A future culture seeking greater novelty would therefore need alternative discovery systems, not merely more

content. Serendipity could be designed deliberately. Platforms could reward sustained engagement with difficult

material rather than only immediate response. Public institutions could support human curation. Users could

choose modes that prioritize distance from existing taste. Local recommendation could matter alongside global

prediction. The goal would not be to abolish algorithms but to stop treating frictionless relevance as the highest

cultural good.

Taste needs continuity, but it also needs interruption. A system that only knows what we already like may

become excellent at preserving identity and poor at provoking transformation.

11. Artificial intelligence and the archive that speaks back

Generative artificial intelligence extends Ghost Modernism into a new phase because the archive no longer

merely surrounds the creator. It can now answer, propose, imitate and synthesize. A user can ask for an image

combining several historical styles, a song with specified genre traits, a screenplay premise resembling known

franchises or a design derived from a textual mood. The reference library becomes conversational.

This development should be understood without mystical exaggeration. Generative systems do not literally

contain every cultural object, nor do they retrieve the past in a straightforward database operation. They learn

statistical patterns from training data and produce new outputs according to those learned structures.

Nonetheless, their cultural relationship to history is fundamentally archival. Their competence comes from

extracting patterns across enormous quantities of prior human production.

That makes them powerful tools for individual creativity. A person without formal illustration training can

visualize an idea. A writer can generate alternatives, critique structure or explore research questions. A musician

can experiment with arrangement. A small studio can prototype rapidly. Such tools lower barriers much as

earlier digital technologies did. It would be inconsistent to celebrate the democratizing effects of inexpensive

recording software while dismissing every new generative tool simply because it automates a later stage of

production.

The concern lies at the collective level. Research published in 2026 has begun to detect a homogenizing

effect in human-AI co-creation. A systematic review and meta-analysis of nineteen studies found a small but

statistically significant reduction in diversity across outputs associated with generative AI use, even when

individual creative performance improved.[16] Another 2026 study reported that AI assistance could increase

creative performance while content homogeneity rose, with some effects persisting after users stopped relying on

the tool.[17] Earlier experiments comparing human and large-language-model writing also suggested that AI-

assisted outputs can contribute less new diversity as collections grow.

These findings are early and should not be treated as a final verdict. Models change quickly, prompts differ,

creative tasks vary, and human collaboration with AI is still developing. It is possible to design systems

specifically to maximize divergence, novelty or random exploration. The 2026 research on generative musicTHE FUTURE UNDER GLASS

recommendations, for example, suggests AI can also broaden attention by lowering search friction and

surfacing less prominent material. Technology does not possess a single cultural destiny.

Still, the homogenization result makes intuitive sense. If many people use the same models, and those

models are trained on overlapping bodies of cultural material, shared statistical tendencies can enter otherwise

unrelated work. The individual may become more productive while the population becomes more similar. This

is the difference between local creativity and collective diversity.

The phenomenon mirrors recommendation systems. A recommender predicts what a user may like based on

prior patterns. A generative model predicts plausible continuations or outputs based on prior patterns. Both are

extraordinarily capable because culture contains structure. Both can also reinforce the most statistically legible

parts of that structure unless deliberately pushed toward the edges.

This gives new force to Larson's warning about imitation of imitation.[1] Generative systems are increasingly

trained on internet corpora that already contain decades of remix, revival, reference and derivative production.

Future models may ingest outputs produced partly by earlier models. The possibility arises of recursive culture:

statistical representations of cultural representations feeding later statistical representations. The ghost acquires

descendants that have never encountered the original.

Again, this need not produce inevitable decay. Human culture has always been recursive. Folk songs mutate

through repeated performance. myths are retold. genres build conventions. apprentices imitate masters. The

important variable is contact with reality. New experiences, communities, technologies, environments and

social conflicts continually introduce information that cannot be generated merely by recombining old media.

Art remains capable of novelty because people continue to live.

The danger emerges when generative systems substitute for that contact. If creators increasingly ask models

what a detective office should look like, what a fantasy kingdom should sound like, how a luxury brand should

write or what a science-fiction city should contain, inherited conventions may become self-reinforcing. The

model returns the statistical memory of representation rather than the inconvenient specificity of the world.

This is analogous to Jameson's nostalgia mode, but automated.[3] Instead of studying a historical period, a

creator can summon a representation of period-ness. The result may be visually convincing while historically

shallow. The tool is optimized to know what "looks like" something. Symbolic force becomes separable from

lived source.

A generative future therefore makes vertical attention more valuable, not less. The creator who studies a

subject deeply, visits places, learns techniques, speaks to people, works with materials and develops private

obsessions gains information that generic models cannot easily supply. Embodied research becomes a

competitive advantage against statistical familiarity.

This suggests an unexpected role for AI in the revival of originality. As generic competence becomes cheap,

specificity becomes precious. If anyone can generate a competent retro poster, the meaningful poster may be the

one rooted in an actual local history, hand process, unusual constraint or personal symbolic system. Automation

can raise the floor while making idiosyncrasy more visible.

The cultural future will therefore depend less on whether AI is used than on how institutions reward its use.

If platforms value volume, speed and immediate recognizability, generative systems may flood the culture with

polished ghosts. If audiences and institutions reward depth, research, local specificity and formal risk, the same

tools can become components of genuinely new forms.

The archive has begun to speak. The next problem is teaching ourselves not to mistake fluency for

revelation.

12. What the stagnation thesis gets wrong

The argument developed so far can become too elegant. Once one learns to see ghosts, everything begins to

look haunted. Every sequel confirms franchise stagnation, every retro fashion confirms atemporality, every

recommendation confirms algorithmic control and every physical-media revival confirms digital exhaustion. A

responsible account must resist that self-sealing structure.THE FUTURE UNDER GLASS

First, the internet has generated forms that would have been difficult or impossible under earlier media

conditions. Online video developed its own editing rhythms, parasocial conventions, reaction formats,

livestream cultures and participatory genres. Video games have become persistent social worlds. Internet-native

music scenes can form across continents rather than cities. Memes constitute a highly compressed vernacular

form of collective authorship. Virtual production, interactive storytelling and creator economies have altered

not just distribution but the form of cultural work.

Second, the apparent lack of new eras may be a perceptual problem produced by age. People are most

sensitive to cultural distinctions during adolescence and early adulthood, when tastes help form identity. Later

they can experience contemporary variations as minor because they no longer inhabit the social contexts that

make differences meaningful. A forty-year-old who hears two youth microgenres as nearly identical may be

repeating the complaint of older listeners in every previous era. Historical humility is necessary.

Third, archival access can stimulate originality rather than suppress it. A creator with access to global

traditions can combine influences that earlier geographical barriers kept separate. Digital distribution has

increased the supply of independent work and reduced reliance on major gatekeepers.[6][18] The result may be a

culture whose innovations are distributed across more niches rather than concentrated into a few movements

visible to everyone. What looks like stagnation from the centre may be experimentation at the edges.

Fourth, nostalgia itself can be inventive. Some of the most influential twentieth-century art was

retrospective. Hip-hop sampling turned recorded history into a new instrument. Post-punk borrowed dub, funk,

disco and avant-garde techniques. House and techno reconfigured existing electronic and dance traditions.

Cinema has always remade genres. Architecture continually reinterprets historical forms. There is no stable line

separating reference from originality.

Larson's own distinction between zombie aesthetics and Channeling is useful precisely because it

acknowledges this.[1] The past can be a source of power. The question is whether a work merely reproduces

recognizable signs or transforms them through contact with a present need. That distinction is qualitative and

difficult to measure, but it prevents the argument from becoming a demand for novelty for novelty's sake.

Fifth, the economics of original intellectual property are not uniformly bleak. The 2025 theatrical market

included notable original successes, and industry reporting in 2026 explicitly described renewed interest in fresh

material after several originals broke through.[13] Streaming and television continue to launch new properties

even while the most visible franchises dominate headlines. Parrot Analytics data suggest that when all new film

and television premieres are counted, the supply of non-IP-based work is larger than public discourse often

implies.[15] The cultural centre may be conservative, but it is not sealed.

Sixth, algorithms are not intrinsically homogenizing. Research results vary. Personalized systems can

narrow within-user diversity in some contexts, increase aggregate diversity in others, or be redesigned to surface

less popular material.[7][8] The system's objective function matters. An algorithm optimized for surprise, public

value or exploration would behave differently from one optimized strictly for engagement.

Finally, younger people are not passive victims of platforms. Pew data show heavy use of social media, but

also growing awareness of excessive use and active attempts to cut back.[9] The revival of physical formats, offline

references and personal style can be understood as evidence of adaptive cultural intelligence. People notice when

an environment becomes oppressive and develop rituals to counter it.

These objections do not invalidate the flattening thesis. They refine it. The claim should not be that the

internet ended originality. The stronger claim is that digital abundance changed the ecology in which

originality becomes visible. It removed scarcity, accelerated comparison, increased the value of recognition,

compressed trend cycles and created infrastructures that privilege prediction. At the same time, it expanded

creative participation, cultural access and niche diversity.

Both outcomes can be true because culture is not a single metric. The present can contain more new work

and feel less new. It can be more diverse and more standardized. It can be globally connected and aesthetically

repetitive. It can liberate creators from gatekeepers while making them dependent on platforms. It can preserve

history while weakening historicity.

The apparent contradictions are the phenomenon.THE FUTURE UNDER GLASS

This is why the language of "flattening" is preferable to "death." Novelty has not died. It has lost elevation.

The new no longer stands automatically above or ahead of the old simply because it is recent. It enters a level

field containing the entire searchable past. This can be democratic. It can also be exhausting.

A culture that understands the problem can respond without romantic restoration. There is no need to

destroy archives, abolish streaming or return to a world where access depended on wealth and geography. The

task is to create conditions in which new forms can again accumulate enough distance, difficulty and social

density to feel like worlds rather than references.

13. What would a genuinely new era require?

The fantasy of switching off the internet entirely is useful as a thought experiment because it reveals what

digital culture removed. If the network vanished tomorrow, the immediate consequences would be

economically and socially catastrophic. Communication, banking, logistics, health systems, education,

government, research and ordinary relationships depend on it. A romantic demand for total disconnection

confuses cultural frustration with infrastructure.

Yet imagine only the cultural effects. Local scenes would become more isolated. People would rely more

heavily on physical media, broadcast systems, print and direct social networks. Information would travel

unevenly. Obscure work would become genuinely obscure again. People would misremember things. Rumours

would survive longer. Aesthetic differences between cities might grow. Trends would diffuse slowly. There would

be fewer references in circulation and more repeated exposure to whatever happened to be locally available.

Some of these changes would be losses. Gatekeeping would strengthen. Rural and isolated people would lose

access. Marginalized communities could become harder to find. Independent creators would struggle to reach

global audiences. Cultural monopolies might become stronger rather than weaker. The pre-internet media

world was not a utopia of originality.

But the thought experiment identifies several creative resources worth recovering selectively: opacity,

locality, scarcity, latency and forgetting.

Opacity means not everything must be immediately explainable or searchable. Artists can protect early work

from constant documentation. Scenes can develop private jokes, rituals and language. Audiences can encounter a

work without immediately opening a second screen to identify every reference. Opacity allows meaning to

remain unresolved.

Locality means physical or social context matters. A work can respond to a particular climate, street,

community, room or institution instead of addressing a generalized global audience. Local cultures need not be

isolated; they simply need enough autonomy to develop texture before being absorbed into universal platforms.

Scarcity means attention is not continuously divided among infinite alternatives. It can be voluntary rather

than imposed. A person chooses a limited library for a month, listens to one album repeatedly, reads one

magazine cover to cover or uses a device that performs only one function. Scarcity can create depth without

reproducing historical exclusion.

Latency means ideas are given time before measurement. Not every cultural experiment needs immediate

analytics. Funding systems can support work whose value is uncertain. Venues can program artists who do not

yet have metrics. Studios can tolerate films that are not pre-sold by brand recognition. Schools can teach process

without demanding instant public output.

Forgetting may be the hardest resource to recover because digital systems remember by default. Yet

forgetting is culturally generative. It allows ideas to return transformed rather than perfectly copied. It prevents

every youthful experiment from remaining permanently visible. It creates distance between eras. Privacy

regulation, ephemerality and deliberate archival boundaries could therefore have aesthetic consequences as well

as ethical ones.

These resources point toward a post-digital cultural policy that is neither technophobic nor passive. Public

libraries and cultural institutions can function as curated alternatives to engagement-driven platforms. Local

venues and arts funding can protect scenes before they become commercially legible. Independent cinemas,

community radio, small publishers and physical retailers can create social contexts around works. Schools canTHE FUTURE UNDER GLASS

teach students how algorithms shape discovery and encourage forms of research that move beyond first-page

search results.

Platforms themselves could offer different modes of attention. A music service might include an "outside my

taste" setting designed to maximize distance rather than relevance. A video platform might allow users to switch

off recommendation and navigate through human-curated channels. Streaming services could foreground

complete works rather than endless feeds. Recommendation could be treated as a cultural design problem, not

merely an engagement optimization problem.

Creators can also build constraints into practice. A filmmaker might restrict reference research after a

certain point and work from memory. A musician might use unfamiliar instruments or limit plugin choices. A

designer might begin from local materials instead of online mood boards. A writer might spend periods reading

only primary sources. Constraint has always been a generator of style because it forces decisions.

Most importantly, institutions need to restore the right to fail. Genuine novelty is hard to predict precisely

because it lacks precedent. If a project can be justified only by demonstrating similarity to proven successes, the

system selects against the unfamiliar before audiences encounter it. The economic pressure for recognition is

understandable, but a culture without protected risk becomes archival by design.

There are historical models. Public broadcasters commissioned unusual work because their mandate was not

identical to immediate market return. Arts councils subsidized experimentation. Independent labels tolerated

records that major companies rejected. Repertory cinemas created audiences through programming rather than

demand prediction. Universities and art schools provided spaces where failure could occur without instant

monetization. None of these institutions was perfect, and all could be elitist, but they demonstrate that cultural

novelty is partly an institutional achievement.

The next era may therefore require infrastructure before aesthetics. A new sound alone does not create a

period. It needs venues, communities, criticism, technologies, economies and rituals that allow it to persist. If

every innovation is instantly absorbed into the same platforms, it remains a style option. A genuinely new era

begins when a different way of organizing attention becomes ordinary.

This is why the analogue revival is potentially more interesting than its visual nostalgia suggests. If it

remains a taste for cassettes, digicams and 1990s café graphics, it will become another recyclable aesthetic. If it

evolves into a demand for ownership, third places, offline sociality, slower media, local institutions and human

curation, it could alter the ecology from which new forms emerge.

The task is not to recreate 1997. It is to recover the conditions that allowed 1997 to differ from 1987.

14. The right to forget

The internet solved an ancient problem with extraordinary success: it made culture easier to remember.

Music that would once have vanished can be heard instantly. Films can be restored. obscure photographs can

circulate globally. fans can document scenes that institutions ignored. Search allows a person to cross decades in

seconds. Digital memory is one of the great achievements of contemporary civilization.

The unintended consequence is that forgetting, once unavoidable, has become a scarce cultural resource.

Taraka Larson's Ghost Modernism was an early attempt to describe what happens when the past becomes

continuously co-present. Her image of the internet as an externalized memory system now feels less

metaphorical than it did in the early 2010s. The web does not simply contain history. It repeatedly activates it.

Recommendation engines resurrect it. rights-holders remonetize it. users remix it. generative systems

statistically synthesize it. culture therefore lives in an environment of persistent afterlife.[1]

The consequence is not cultural death. The contemporary world produces too much art for that claim to

survive scrutiny. Digitization has enabled new creators, expanded choice and made possible forms of global

collaboration unimaginable under older media systems.[6][18] The stronger diagnosis is that novelty has been

flattened by proximity. The new is born beside everything that came before it.

This changes how eras form. Earlier cultural moments depended partly on disappearance. Technologies

became obsolete and therefore acquired historical distance. scenes dissolved. records went out of print. television

programs became memories. styles became embarrassing enough to be abandoned before they wereTHE FUTURE UNDER GLASS

rediscovered. A new generation encountered the previous one through fragments rather than complete

databases. The gaps mattered.

Today the gap is shorter. A style can become retro while its first practitioners are still active. A franchise can

be rebooted before cultural memory of the previous version has faded. A song can disappear from charts and

return years later through an algorithmic trend. A childhood interface can become an aesthetic reference before

the child reaches middle age. Culture rarely leaves the room.

The result helps explain the feeling of being stuck in a loop. It is not that every new work is a copy. It is that

the loop is structurally available as the path of least resistance. Familiar intellectual property lowers financial

risk. familiar aesthetics lower cognitive risk. recommendations lower discovery risk. generative tools lower

production risk. Each innovation is rational in isolation. Together they can produce a culture that is

extraordinarily efficient at extending what already exists.

The loop also explains why younger people may be drawn to objects associated with earlier digital or pre-

digital life. Their interest is not merely a desire to cosplay their parents' adolescence. The older object promises

boundaries. A CD ends. A film roll runs out. A magazine cannot refresh. A café can be entered without creating

a data trail of every glance. A physical collection belongs somewhere. These limits can feel exotic to people raised

in systems designed around infinite continuation.

There is a danger that the desire for boundaries will itself become aestheticized and sold back through the

same platforms. That is almost certain to happen. "Offline" can become a brand, analogue can become a content

genre and digital detox can become a subscription product. Capitalism has always been capable of packaging

resistance. Yet incorporation does not erase the original need. The question is whether the need produces durable

practices after the trend passes.

The most productive lesson from The Now Age is therefore not that historical borrowing must stop.

Larson's concept of Channeling offers the opposite principle: the past can be approached as a source of energy

rather than a catalogue of costumes.[1] The difference lies in whether history deepens contact with the present. A

revival that restores communal space, sustained attention or symbolic depth may generate something genuinely

new even if its materials are old.

The same principle applies to technology. The answer to algorithmic culture is not necessarily to reject

algorithms but to design them around values other than immediate engagement. The answer to generative AI is

not necessarily abstinence but creative practices that demand specificity, embodiment and contact with reality.

The answer to streaming is not necessarily to abandon streaming but to build rituals of commitment within

abundance.

Novelty has always depended on constraints. A musical instrument has a finite range. A film stock behaves in

particular ways. A building has a site. A scene has a geography. A human body has limited time. Infinite choice

can weaken form because form requires exclusion. To create is to decide what will not be included.

The internet made exclusion psychologically harder by keeping alternatives permanently visible. Every

decision can be compared with thousands of references. Every style has predecessors. Every experiment can be

searched. The creator begins to feel that originality is impossible because the archive can always produce an

analogy. But originality was never the absence of analogy. It was the ability to make inherited material answer a

present condition so strongly that the result acquired its own historical gravity.

That remains possible.

The next genuinely distinct era may already be forming in places that do not yet look important: local

rooms, private group chats, small labels, community workshops, unusual software scenes, schools, independent

games, regional music, children using obsolete devices incorrectly, artists who deliberately refuse metrics, or

creators combining technologies in ways that major platforms have not optimized. Historical novelty often

looks trivial before institutions recognize it.

What may need to change is our ability to leave such things alone long enough to become themselves.

A culture that remembers everything must learn selective amnesia. It must preserve archives while creating

spaces where archives do not dictate the next move. It must allow people to be out of the loop, to misunderstand,

to discover late, to repeat imperfectly, to encounter without explanation and to create without knowing

immediately how the work will be categorized.THE FUTURE UNDER GLASS

The future will not be rescued by pretending the past is gone. It will be rescued by restoring distance from it.

That is the deeper meaning of the flattening of novelty. The problem is not that history has become too

rich. The problem is that history has become too near. We live in a room containing almost every image, sound

and story that preceded us, with the walls continuously rearranging themselves according to what we have

already looked at. The room is astonishing. It is also difficult to leave.

The cultural task of the next generation may be to build doors: not doors back to an imagined analogue

paradise, but doors into spaces where attention can gather, where locality can matter, where institutions can risk

failure, where symbols can recover depth, and where something new can exist for a while before the archive

explains it.

Perhaps the future needs, above all, the right to arrive before it is remembered.

Sources and further reading

The essay uses bracketed source numbers keyed to the references below. Sources were cross-checked against

current public material available in October 2026. These references are selected rather than exhaustive.

[1] Taraka Larson, The Now Age: 'Introduction: The Birth of Ghost Modernism.' https://now-age.org/ghost-modernism/

[2] Vee Hua, 'Prince Rama Band Interview: Utopia of the Now Age (w/ Manifesto),' REDEFINE, 3 February 2012.

https://redefinemag.net/2012/prince-rama-band-interview-utopia-the-now-age-manifesto/

[3] Fredric Jameson, 'Postmodernism, or, The Cultural Logic of Late Capitalism,' New Left Review I/146, July-August 1984.

https://newleftreview.org/issues/i146/articles/fredric-jameson-postmodernism-or-the-cultural-logic-of-late-

capitalism

[4] Simon Reynolds, 'Paperback Q&A: Simon Reynolds on Retromania,' The Guardian, 10 April 2012.

https://www.theguardian.com/books/2012/apr/10/paperback-q-a-simon-reynolds-retromania

[5] Mark Fisher, 'Hauntology, Nostalgia and Lost Futures,' interview with Valerio Mannucci and Valerio Mattioli, 2014,

republished by Blackout. https://my-blackout.com/2019/04/26/mark-fisher-hauntology-nostalgia-and-lost-

futures-interviewed-by-v-mannucci-v-mattioli/

[6] Luis Aguiar and Joel Waldfogel, 'Quality Predictability and the Welfare Benefits from New Products: Evidence from the

Digitization of Recorded Music,' NBER Working Paper 22675, 2016. https://www.nber.org/papers/w22675

[7] Ashton Anderson et al., 'Algorithmic Effects on the Diversity of Consumption on Spotify,' Proceedings of The Web

Conference 2020. https://doi.org/10.1145/3366423.3380281

[8] Dougal Shakespeare, Victor Chareyron and Camille Roth, 'Reframing the filter bubble through diverse scale effects in online

music consumption,' Scientific Reports 15 (2025). https://www.nature.com/articles/s41598-024-75967-0

[9] Pew Research Center, 'Teens, Social Media and AI Chatbots 2025,' 9 December 2025; and 'Teens, Social Media and Mental

Health,' 22 April 2025. https://www.pewresearch.org/internet/2025/12/09/teens-social-media-and-ai-

chatbots-2025/

[10] The Guardian, 'Have I been influenced, or is this actually me? How personal taste fell out of fashion,' 14 June 2026.

https://www.theguardian.com/media/2026/jun/14/have-i-been-influenced-personal-taste-out-of-fashion-

algorithm

[11] Jane Clinton, 'If you can't hold it, you don't own it: why media fans want to escape algorithms with CDs, DVDs and vinyl,'

The Guardian, 2 October 2026. https://www.theguardian.com/media/2026/oct/02/physical-media-fans-streaming-

algorithms-cds-dvds-vinyl

[12] Madeleine Schulz, 'Did Micro-Trends Kill the Trend Cycle?', Vogue, 12 June 2024. https://www.vogue.com/article/did-

micro-trends-kill-the-trend-cycle

[13] NRG / TheWrap, '2026 In the Frame,' industry report, 2026, including analysis of franchise dominance and original-film

successes in 2025. https://www.thewrap.com/wp-content/uploads/2026/01/2026-In-The-Frame_NRG_TheWrap-

Pro.pdf

[14] The Numbers, '2025 Top-Grossing Movies,' current historical box-office chart.

https://www.the-numbers.com/market/2025/top-grossing-movies

[15] Parrot Analytics, 'The IP Illusion: Why Original Content Is Rising But Still Overshadowed,' 14 June 2025.

https://www.parrotanalytics.com/insights/adapted-ip-vs-original-concepts-supply-demand-trends/THE FUTURE UNDER GLASS

[16] 'Does generative AI make us think alike? A systematic review and meta-analysis of homogenisation effects in human-AI co-

creation,' Behaviour & Information Technology, published online 31 August 2026.

https://www.tandfonline.com/doi/full/10.1080/0144929X.2026.2726451

[17] Yiyong Zhou et al., 'Creative scar without generative AI: Individual creativity fails to sustain while homogeneity keeps

climbing,' Technology in Society 84 (2026), 103087. https://doi.org/10.1016/j.techsoc.2025.103087

[18] Joel Waldfogel, 'The Random Long Tail and the Golden Age of Television,' Innovation Policy and the Economy 17 (2017);

and 'Digitization and Its Consequences for Creative-Industry Product and Labor Markets,' NBER / University of Chicago

Press, 2022. https://www.nber.org/books-and-chapters/innovation-policy-and-economy-volume-17/random-

long-tail-and-golden-age-television