In March 2026, a venture capital-backed company released a music video for the Academy Awards featuring their AI-generated actress, Tilly Norwood. The internet's response was instantaneous: "This makes Rebecca Black's Friday look like a masterpiece." "At least Skynet was honest about trying to erase humanity."
The video was not technically bad. That was the problem.
Audiences were not detecting incompetence. They were detecting the absence of someone who meant something. Technically polished, emotionally hollow. The production values were there. The intentionality was not. And everyone could feel the difference, even people who could not say why.
That gap: between technical execution and felt human presence: is widening at exactly the moment the creative industry is most vulnerable to it. AI is taking the execution. The presence cannot be generated. The question is who builds the infrastructure around what cannot be taken.
Four simultaneous breaks.
Skill became a commodity. Midjourney, Sora, Udio, ElevenLabs: the barrier to competent images, video, music, and voiceover collapsed to near zero. Junior concept artists, session musicians, motion designers, commercial illustrators: roles that took years to master. The output now takes seconds to generate. Clients choose the seconds. This is not about quality. It is about the economic irrelevance of human execution time.
Distribution captured. Cafune's "Tek It" went mega-viral on TikTok in 2022. Over a billion streams. The band signed to Elektra, hustled to build social media accounts, produced a sped-up version of their own song to match the most popular AI remix. In 2024, Elektra dropped them. "That amount of profit wasn't enough to convince anyone of anything." The platform captured the value. The artist captured the exposure. They ended up signing better terms with TikTok's own distribution platform than any label could offer. Every musician today is forced to become a content creator, whether they want to be or not.
Studio economics broken. In 2023, the Writers Guild and SAG-AFTRA struck simultaneously for the first time since 1960: 11,500 writers, 160,000 actors, 191 days. The core issues were not wages in isolation. Streaming had collapsed the residuals model that had sustained the creative middle class for decades. Writers who earned steady income from TV reruns found themselves earning nothing as shows moved to Netflix and never reran. AI was the next step: generate scripts, replicate likenesses, own everything, pay nothing. The strike won partial protections. The underlying economics did not change.
Attention at war with quality. Ted Gioia maps the cultural food chain: art gave way to entertainment, entertainment is giving way to distraction, distraction is being engineered into addiction. The fastest-growing sector of the culture economy is not creativity. It is the dopamine loop. TikTok, Reels, Shorts, algorithmic feeds designed around stimuli that last seconds and must be repeated. The neurochemical reward is the product. The content is the delivery mechanism. AI generates this content at civilizational scale: cheaper, faster, more precisely targeted. Human creators do not compete in this tier. Nobody does. It is not art.
Culture is eating its own past.
Gioia calls it the Shock of the Old. In 2026, the dominant aesthetic theory: though nobody names it: is that novelty is over. Hollywood runs the Hero's Journey on every major film. It is a storytelling framework derived from myths thousands of years old. Major labels do not sign new artists. They acquire back catalogs. Sony paid $1.2 billion for Michael Jackson's catalog in 2024. No label would invest a fraction of that in launching someone new.
"In 2024, musicians are worth more old than young, dead than alive."
When Spotify pays $0.003 per stream, no artist can afford creative risk. When the algorithm rewards familiarity, studios produce familiarity. When AI generates content in any established style at zero cost, the only defensible position is owning the original IP the AI will imitate.
The result: an industry that has stopped believing a future is possible, optimizing for safety by consuming its own history.
The pipeline dries up. The industry generates revenue from the past while eliminating the conditions that would produce the future. Concept artists, junior directors, emerging musicians: the entry paths that once led to major careers are closing. AI replaces the junior roles. Risk aversion kills the investment in new voices. The catalog grows more valuable as the future becomes less fundable.
The cultural diet narrows. Algorithms optimize for engagement. Engagement correlates with familiarity. Familiarity produces more of the same. Kyle Chayka's Filterworld thesis: the algorithm flattens taste across platforms, geographies, and demographics until everything sounds and looks indistinguishable. Not by conspiracy. As the emergent property of a billion individual engagement decisions, each of which chose the familiar over the unfamiliar. The aesthetic monoculture is algorithm-generated.
AI services the nostalgia economy at scale within this decade: generating new Marvel storylines, new Star Wars content, new brand campaigns in established styles, at zero creative wages. The studio that owns the IP wins indefinitely. The human who created the original receives nothing beyond the initial payment. The creative economy bifurcates: those who own the past, and everyone else.
The model was trained on your life's work.
Rhythm and Hues Studios filed for Chapter 11 bankruptcy in February 2013, while their artists were still finishing Life of Pi. The film won the Academy Award for Best Visual Effects. The VFX supervisor was played off the stage mid-speech by the orchestra. Win the Oscar. File for bankruptcy the same week. The studio captures the value. The artists capture the credit.
AI adds a new chapter. The models replacing concept artists were trained on concept artists' work. Getty Images built its own AI image generator trained on its licensed photography catalog: images created by photographers who received a percentage of licensing fees. The AI now competes with those photographers, who receive nothing for having trained it. The New York Times sued OpenAI. Getty sued Stability AI. The Authors Guild sued OpenAI. Every case is asking the same question: does training an AI on copyrighted work constitute infringement? And if yes, what is owed to every writer, artist, and musician whose work was consumed to build a machine that now undercuts them?
No ruling has been reached. The cases move slowly. The displacement does not wait for the courts.
The creative commons: the accumulated output of human civilization: was consumed without consent or compensation to build commercial products. This is the creative industry's version of enclosure. The commons taken, privatized, monetized. The people whose work trained the models receive nothing while the companies that built the models generate billions. This is not abstract. It is the specific economic mechanism that explains why a concept artist with fifteen years of experience is being replaced by a model trained on fifteen years of similar artists' work.
Provenance becomes commercially significant. When AI-generated content saturates every channel, documented human origin becomes a differentiator. "Human-made," "trained on licensed data only," "no generative AI" become labels that command premiums. The story of how something was made, by whom, under what conditions: this is what the AI cannot generate. The certificate of origin becomes the product.
Cultural sovereignty becomes a political and economic question. Communities whose aesthetic traditions were scraped to train Western AI models now receive AI-generated approximations of their own culture back. The homogenization risk is not that everything becomes American. It is that everything becomes the AI's best guess at everything, flattened through the same training corpus, losing the specific human context that gave it meaning.
The counter-signal is structural, not sentimental.
While the distraction machine expands and the nostalgia economy hollows out, something is moving in the opposite direction. Not nostalgia. A market accurately detecting scarcity.
The Taylor Swift number is not anomalous. It is proof of a thesis. The Eras Tour grossed more than any tour in history not because of the songs: those are free on Spotify: but because of the physical presence of a specific person taking a specific risk on a specific night. The imperfection was the point. The fact that it was irreducibly her, in that room, was the product. The experience could not be replicated. That is exactly why people paid $500 for a ticket.
Vinyl follows the same logic at a smaller scale. You take the record out. You put the needle down. You sit with the album as a complete thing, in sequence, at the pace the artist intended. The ritual is the experience. And the ritual resists algorithmic optimization entirely.
The premium on live, physical, and irreproducible experiences rises as AI saturates every other channel. Ticket prices grow. Handmade goods command premiums. Events organized around human presence and shared risk become the high-value tier of the creative economy. This is not anti-technology sentiment. It is rational economic behavior in response to genuine scarcity.
The creative economy bifurcates sharply. One track: AI-generated content at infinite scale, zero marginal cost, captured by platforms and IP holders. The other: irreducibly human experiences at premium prices. The middle, the working creative professional sustained by technical skill, shrinks. What remains is the idea and the presence. Execution becomes infrastructure, not identity.
The creative studio that positions itself on Tier 3: irreproducible, human-presence-anchored, provenance-documented: captures the premium market while the commodity market floods with AI output. The question for every creative organization is which tier they are building for, and whether they know the answer before the bifurcation completes.
Benjamin's aura, updated for 2026.
Walter Benjamin wrote about the aura in 1935: the unique presence of a work in time and space, the trace of the hand and the intention that made it. Mechanical reproduction destroys the aura. The photograph is not the painting. The copy is infinitely reproducible. The original's presence is not.
AI does something Benjamin could not have anticipated. It does not reproduce the original. It generates content that never had an original. No hand. No intention. No moment in which a specific person decided this specific thing was worth making.
The Tilly Norwood moment is audiences answering a question they did not know they were asking: was there ever anyone here? The answer was no. They could feel it, even without language for it.
"The tech platforms don't want to find the next Michelangelo or Mozart. They want to create a world of junkies. Because they will be the dealers."
Ted Gioia, "The State of the Culture, 2024," The Honest BrokerThis connects to a deeper truth about the creative industry's current moment. The distraction economy is not hostile to art. It is structurally incompatible with it. You cannot have aura in a three-second scroll. You cannot have presence in an algorithmically optimized feed. The attention economy does not want people to feel something lasting. It wants them to feel something again in thirty seconds.
The paradox: AI was built from human aura. Every brushstroke, every lyric, every line of screenplay in those training datasets carried the trace of someone's intention. The model extracted the patterns and discarded the intention. It learned what creative work looks like without learning why anyone made it. The result is technically competent and humanly absent: exactly what you would expect from a process that consumed the products of meaning without consuming meaning itself.
Three patterns. One event. The nostalgia economy consumes the past because it cannot afford a future. The extraction machine builds AI on human creative labor and pays nothing for it. The counter-signal grows because scarcity of human presence is accelerating. All three are the same structural shift: the creative economy discovering it built its infrastructure on the wrong foundation.
Where the industry is going.
The creative industry is splitting into three distinct tiers. Not gradually. Now.
Reboots, IP, back catalogs, franchise properties. AI services this tier fully within this decade. The IP holder wins. The original creator earns nothing. Dying slowly and very profitably.
Live presence. Risk taken in public. Documented human origin. Something that could have failed and did not. Irreproducible by definition. Premium by scarcity. Growing in direct proportion to AI saturation of Tiers 1 and 2.
The working creative professional who existed in the middle: sustained by technical execution: belongs largely to Tier 2. That tier is being serviced by AI. The displacement is not announced. Job listings disappear quietly. Budgets for junior roles do not return. The craft remains. The livelihood does not. The question for every creative organization is not whether this is happening. It is which tier they are positioned in when it completes.
The scarcity is the opportunity.
Every structural bifurcation creates a gap. This one is widening faster than anyone is building for it.
Provenance becomes the product. "Human-made," "no generative AI," "live," "original": these become premium differentiators within five years. When AI-generated content saturates every channel, the documented story of how something was made, by whom, under what conditions, is what the machine cannot produce. The certificate of human origin becomes the creative brief. The proof of intentionality becomes the value proposition.
Live and irreproducible as the growth tier. Concert revenue at $31B and growing. Immersive experiences, live shows, events where the room itself is the product. These formats cannot be streamed, replicated, or generated. They require presence and the irreducible risk that it might not work. That risk is exactly what makes them worth paying for. Taylor Swift's $2B tour is not a nostalgia act. It is a proof of concept: the thing that cannot be replicated is the thing people will pay most for.
Point of view as the scarce resource. AI can produce technically correct work. It cannot produce the decision about what the work should make people feel and why. That decision requires someone who has spent years caring about something enough to develop judgment around it. Direction, curation, the ability to know what a moment needs before it happens: these are not skills AI acquires. They are what remain when every other skill has been commoditized. The people who have them are becoming rarer and more valuable simultaneously.
Community around the irreducible. The creative professionals displaced from Tier 2 are not disappearing. They are looking for somewhere to go. Community organized around genuine craft, real collaboration, and creative standards that have nothing to do with engagement metrics: this is the infrastructure the market is not yet building. Whoever builds it first owns the relationship with the next generation of the people who will still be making things in twenty years.
The soul is not a skill. The AI proved it by learning every skill and producing something audiences immediately recognized as soulless.
What the soul actually is: the trace of intention. The evidence that a specific person, at a specific moment, decided this was worth making and took the risk of making it. The possibility that it could have been different. The fact that it was not.
The machine that replaced you was built with your work. That is the injustice. But the machine also revealed something it could not have intended: what it took from the creative industry was the execution. What it could not take was the reason anyone did any of it in the first place.
The creative industry that survives is the one built around that reason.
Everything else is the distraction machine.
Build the other thing.