On Art in a World Measured by the Advertiser's Unit
Let us begin with a fact — cold, verified, boring. According to visitor-observation studies at the Louvre, the average time spent looking at the Mona Lisa is fifteen seconds. Fifteen seconds. For comparison: a study conducted at the Metropolitan Museum of Art in New York (Smith & Smith, 2001) found that the median time an ordinary visitor spends in front of an ordinary artwork is seventeen seconds. In other words, the world's most famous painting attracts even less attention than a random painting hanging in a museum.
Now place that number in front of a contemporary indexing algorithm. The algorithm sees the number and the classification is automatic: high bounce rate, low dwell time, insufficient engagement signal. The result is already prepared — this content does not satisfy user intent. Not worth indexing.
The Mona Lisa, on its own digital page, would not have been indexed. Neither would an ordinary painting. Because seventeen seconds also falls below the algorithm's "thin content" threshold. This simple arithmetic tells us one thing: the entirety of art, by the algorithmic standard, is nothing more than low-quality content.
The sentence sounds like a joke at first. But if you sit with it for a second, you realize that a system which treats the world's most famous artwork — and indeed all of art — as "low-quality content" is not an exception. It is the actual operating logic of the digital environment we move through every day. This is not a bug. This is the design.
Forty percent disappointment
The data has another layer. According to a study by the CouponBirds platform, which mined more than eighteen thousand independent visitor reviews of the Mona Lisa, approximately forty percent of the reviews are entirely negative. In online tourist literature the painting now carries the title of "the world's most disappointing masterpiece."
What does that number mean to an algorithm? High negative sentiment, low rating, a pattern of abandonment. Contemporary indexing systems read these signals as markers of low quality. So: visitors grow angry at the painting because it fails to deliver the experience they expected, and the algorithm records that anger as if it were the painting's own fault.
But the fault is not the painting's — it is the expectation's. The visitor arrives in front of the Mona Lisa carrying the expectation that an advertising frame would deliver: be hooked immediately, be impressed at once, be satisfied in seconds. The painting does not do this. The painting was never made to do this. But the visitor records this as an artistic failure — and the algorithm then counts the record.
The expectation built by advertising has become the metric by which art is evaluated.
The advertiser's units
Bounce rate, dwell time, engagement signal, session depth — these concepts share one thing in common: none of them were invented for art. All of them were developed in the early 2000s by the direct-response advertising industry. The goal was simple and honest: to measure how well an ad sells.
What does a good advertisement do? It hooks immediately, stays short, delivers its message in a single pass, drags the user toward the purchase button. That is why the advertising metric asks: did the user stay long, did they click, did they not bounce? These questions are reasonable for an ad. Even clever.
But the same questions, asked of a poem, produce something entirely different. Because a good poem does not hook immediately — most of the time it is not even understood on first reading. A good poem does not drag the user toward a button; it disturbs the user, breaks something inside them, perhaps makes them come back another day. When a good poem is short, it is short because the form demanded it, not because it is shallow.
The advertisement measures; art sows. Their metrics cannot be the same.
But they became the same. Contemporary indexing systems measure art with instruments invented by the advertiser. This is like trying to weigh love on a digital scale. A number will come out — but no one is allowed to ask what that number is the number of.
The dictatorship of user intent
At the heart of current indexing algorithms lies the concept of "user intent" (search intent). The system rests on an assumption: the user knows what they are looking for; good content is the content that gives the user what they already know, by the shortest route. This is the absolutization of user desire.
The entire history of art tells the opposite. The person who began reading Kafka was not looking for Kafka — Kafka found them, changed something inside them, redefined who they were. The visitor who stands for thirty minutes in front of a Rothko, crying, did not arrive with the intent "I would like to look at this painting"; the painting seized their intent. The reader who first opened Ulysses did not have the intent "let me read a seven-hundred-page experimental Irish novel" — the book created that intent.
Art does not respond to user intent. It builds intent. This is why the indexing algorithm is genuinely blind to art — the only thing it can measure is the user's conscious, short-term, verbally articulable intent. What art does is precisely to slip beneath that intent, walk around it from the side, dismantle it.
A reader who exits a poem quickly may have been transformed by that poem. The algorithm cannot see this. The only thing the algorithm sees is the user pressing the back button after three seconds. The algorithm reads: bad content. What actually happened: good poem.
"Information Gain" — the official name of a category error
The latest fashionable concept in contemporary indexing literature is "information gain." The system asks every page: how much new information does this page offer the user, compared to existing pages? The novelty of the page is scored mathematically; the low-scoring ones are pushed to the back.
The assumption beneath this concept is clear: art = information transfer. If you acquire new information by reading a poem, it is a good poem; if you do not, it is a copy. Because you do not gain "new data" by listening to a Bach fugue, Bach is unnecessary.
This is a visible category error. A work of art is not a packet of information. A work of art is an experience — and the same experience, undergone at different times, gives entirely different things. The same poem read once at seventeen and once at forty-seven is two different poems. The same painting looked at across two different years is two different paintings. Art is not a repeatable source of information; it is an object that opens differently each time it is repeated.
But the algorithm recognizes only information. Everything that does not fit information's definition is marked as "lacking added value." So the poem has been officially declared valueless in the new ontology of the digital universe. Because the unit of measurement cannot measure it, it is treated as non-existent.
Soft erasure — the algorithmic horror film
What art faces here is not classical censorship. No bureaucrat sits at a desk and says "this poem is forbidden." No commission judges the work. No book is burned.
What happens is much more insidious: soft erasure. The work continues to exist, the artist continues to write, the texts continue to be uploaded to the internet — but they cannot reach any visible place. In the artist's control panel, a line appears: "Crawled — currently not indexed." A digital Schrödinger condition: the work is both present and absent.
This is the underlying logic of a horror film. The enemy is invisible. There is no moment of pain, no decision, no blow. The whole action is mechanical. The algorithm is not deliberately cruel; it is only doing what it was designed to do. Your absence is the byproduct of the system functioning correctly. The digital version of bureaucratic evil — the algorithmic form of banality.
The advertiser's world — the rule that came into effect silently
Three days ago, on the fifteenth of May, two thousand twenty-six, the spam policy of one of the major indexing platforms was quietly updated. The added sentence, in essence, was this: attempts to manipulate AI-assisted search summaries now count as spam.
The sentence seems trivial. In fact it says a great deal. At the center of the policy are the AI summaries — those automatically generated responses the user encounters on the search results page. How much advertising those summaries actually carry is a separate question. But the logic beneath is this: the system now indexes not merely "good content," but content compatible with its own advertising infrastructure. Anything that falls outside the advertising infrastructure — including art — is, from the system's perspective, either noise or a threat.
The price the artist must pay to be allowed to exist is this: to convert their own text into a noise that is suitable for the advertising environment. Keyword density, header hierarchy, structured-data tags, page speed, internal linking, domain authority… None of these have anything to do with art; all of them have to do with advertiser standards. The artist cannot write alone; without an SEO team, a data analyst, and a technical infrastructure manager behind them, they are not visible in the digital world.
The result can be summed up in a single sentence: to become visible, the artist must stop being an artist.
Historical precedent — and something deeper
The historical precedent for this is the television industry. For fifty years, television lived on advertising revenue; advertising revenue depended on ratings; ratings shaped the programs. In the end, television dramas were structured around commercial breaks, season finales were engineered based on advertiser tracking numbers. Art — to some degree — bent itself to the advertiser's need.
But television was at least honest. The writer knew from the start: I am writing into the gap between ad breaks. The screen even displayed, "We'll be right back after these messages." The seam was visible.
Digital indexing does not do this. Digital indexing tells you: "I am a neutral infrastructure. I am only measuring quality." But the quality it measures is the advertiser's definition of quality. The algorithm is an advertising instrument claiming to be transparent. The digital era did not merely erase the boundary between art and advertisement — it denied that the boundary existed.
What we lose
The Mona Lisa would not have been indexed, yes. But the real horror is not this — the Mona Lisa already exists, it hangs in the museum, it is under protection. The real horror is this:
Among the unindexed, there is a new Mona Lisa that will never hang in any museum.
She is writing right now. She is painting right now. She is publishing right now. And none of you can see her. Because the algorithm did not see her. Because the algorithm did not see her, the major platforms did not recommend her. Because the platforms did not recommend her, no one in the networks linked to her. Because no one linked, no journalist mentioned her. Because no journalist mentioned her, no publisher read her. Because no publisher read her, no book was printed. Because no book was printed, she did not enter history.
But the work itself — pen in hand, in front of the screen — exists.
It is only that the system has marked it: "non-existent."
A call
This essay is not a manifesto — it is a diagnosis. We live in the advertiser's world. The units of measurement of the world we inhabit are the units the advertiser invented. Every time the value of art is measured by these units, a small part of art's value is lost — and the loss accumulates.
To notice what the measuring hand actually is, is to take a step. If today a poet, an aphorist, a micro-story writer is invisible in the digital environment — the reason is not that they are writing badly. The reason is that the measuring hand cannot measure them.
Art cannot be measured by the advertiser's unit. There is no need to remind the artist of this — they already know; they feel it every day, in their own absence. The reminder must go to the measuring hand. And if the measuring hand does not listen — which it does not — then perhaps the task of art has now become precisely this: to remain that which the measuring hand cannot measure.
Fifteen seconds was not enough for the world's most famous painting. But we still see the Mona Lisa.
The question is: what will we do for the thousand Mona Lisas now standing inside the digital silence?
