Yves Citton

Foreword

By Edo Segal

I've been building attention machines for twenty years.

That's not what we called them. We called them platforms, experiences, products. We called them engagement tools, discovery engines, personalization layers. But strip the language back to what actually happens — what the code actually does when it meets a human nervous system — and the function is the same: capture a mind's focus, hold it, and convert the holding into value. I was good at it. My companies were good at it. The metrics went up and to the right, and for a long time I did not ask what the metrics measured.

Yves Citton made me ask.

I came to his work sideways — not through philosophy but through a product failure. We were building an AI-driven content system, and something was wrong. Users were spending more time on the platform. Session lengths were climbing. Every engagement metric was green. But the qualitative feedback was miserable. People felt worse after using the product. They described the experience in language that sounded like addiction, not satisfaction. Time spent was up. Value created was down. And I had no framework to explain the gap, because every framework I'd ever used reduced attention to a single quantity — a resource to extract, a number to maximize.

Citton gave me the framework I was missing. His work on the ecology of attention introduced me to an idea so obvious I was embarrassed not to have seen it: attention is not one thing. It comes in modes — deep and floating, individual and joint, focused and diffuse — and those modes serve different functions. You can maximize one while destroying the others. You can flood an environment with stimuli that drive engagement metrics through the roof while systematically eliminating the conditions under which people think deeply, connect meaningfully, or attend to anything together. You can optimize a system for capture and call it success while the ecology collapses underneath.

When generative AI arrived — when I watched the tools I was building begin to produce content at scales that made human creation look like a rounding error — Citton's framework became not merely useful but urgent. The flood was real. I could see it in my own systems. And I could see that every instinct the industry had — more content, more personalization, more engagement, more capture — was exactly wrong. Not wrong for business. Wrong for the species.

This book is my attempt to understand what Citton understood before most of us were paying attention: that the crisis is not distraction. The crisis is monoculture. The homogenization of how we attend, driven by machines that cannot tell the difference between a mind in contemplation and a mind in compulsion. The flood is here. The question is what we build now.

— Edo Segal ^ Opus 4.6

---

About Yves Citton

Yves Citton is a French literary theorist, philosopher, and professor of literature and media at Université Paris 8 Vincennes–Saint-Denis, where he co-directs the EUR ArTeC (Arts, Technologies, and Digital Codings) graduate research program. Born in 1962, he previously spent over two decades at Université Grenoble Alpes, where he taught eighteenth-century French literature before turning his attention to media ecology, attention studies, and the politics of collective perception. He is the author of numerous influential works including *Pour une écologie de l'attention* (2014), translated as *The Ecology of Attention* (Polity Press, 2017), which established him as a leading theorist of attention in the age of digital media. His broader body of work — including *Renverser l'insoutenable* (2012), *Médiarchie* (2017), *Générations collapsonautes* (2020), and *Faire avec* (2021) — spans Spinozist philosophy, the political economy of affects, media archaeology, and ecological thought. Citton is also co-editor of the journal *Multitudes* and a frequent contributor to public debates in the Francophone world on the intersections of culture, technology, and the commons.

---

Not metaphorical silence — the absence-of-noise kind. The kind where you're sitting with a blank page and there's no tool waiting to fill it for you, no assistant offering to help, no algorithm optimizing your next thought. Just you and the emptiness and whatever your mind decides to do with it.

I built things that eat silence for breakfast. I know this. The tools I've spent my career creating, investing in, obsessing over — they are, every one of them, silence eliminators. Every gap in attention is an opportunity. Every empty moment is a problem to be solved. Every pause is latency. I believed this. I still partly believe it. The builder in me sees an unfilled space and thinks: I could put something useful there.

Citton's ecology of attention haunts me because it names the cost of that instinct. Not the obvious cost — distraction, information overload, the things we already have words for. The deeper cost: the modes of attention that only grow in silence. The floating awareness that generates genuine surprise. The joint attention that makes two people into a community. The collective focus that makes democracy something more than a system of competing feeds.

I've felt those modes dying in myself. I wrote about productive addiction earlier in this book — the inability to stop building because the building is genuinely good. What I didn't say clearly enough is what the addiction replaced. It replaced the walks where I thought about nothing and arrived home with an idea I couldn't have planned. It replaced the conversations where two people stared at the same problem long enough for the problem to change shape. It replaced the boredom — that precious, unbearable, irreplaceable boredom — from which every genuinely new thing I've ever thought emerged.

I'm not going to tell you to put down your tools. I haven't put down mine. What I am going to tell you — what writing this book, what sitting with Citton's framework and letting it reshape how I see my own work, has taught me — is that the ecology matters more than the tool. The tool is one plant in a garden. The garden is the thing that has to survive.

Build your tools. Amplify your reach. Use the AI. Use it hard, use it well, use it for things that matter. But protect the silence. Protect the empty hours, the aimless walks, the conversations that go nowhere productive, the staring-out-windows time that looks like waste and is actually the most important thing your mind does. Protect the commons — the shared attention that makes you part of something larger than your feed.

The ecology of attention is not someone else's problem. It is the ground you are standing on. Cultivate it or watch it die.

The choice is yours. It always was.

-- Edo Segal

Chapter 1: The Commons Before the Flood

For roughly three hundred thousand years, Homo sapiens attended to the world without anyone suggesting that attention was a commodity. A mother watching a child near a fire, a hunter tracking the minute disturbance of grass that betrayed a predator's approach, a gathering of elders listening to a story that encoded the survival knowledge of six generations — in each case, attention was not spent but practiced. It was not captured but cultivated. It was not a resource extracted from individuals but a capacity developed within communities. The ecology of attention, to borrow Yves Citton's foundational term, was sustained by the same environments that sustained everything else: the rhythms of day and night, the demands of survival, the social architectures that determined who spoke and who listened, who watched and who performed, who attended and who was attended to.

This is not nostalgia. Citton's framework is not a romantic appeal to some preindustrial golden age when attention was pure and undistracted. The ecology of attention has always been political. Every society in human history has organized attention — determined what deserves collective focus, what can be safely ignored, who has the right to command the gaze of others. Temples were built to direct attention upward. Thrones were elevated to direct attention toward power. Rituals were designed to synchronize attention, to create moments when every member of a community attended to the same thing at the same time — and in that shared attending, experienced themselves as a community. Joint attention, the capacity of multiple minds to focus on a common object, is not merely a cognitive phenomenon. It is the precondition for collective meaning. A society that cannot attend together cannot mean together.

The critical insight Citton offers — the one that reframes everything that follows — is deceptively simple: attention is not a thing inside a skull. It is a relationship between a mind and an environment. Change the environment, and you change the attention. Not merely its object, not merely its duration, but its fundamental character — the mode in which attending occurs. A mind attending in a quiet forest attends differently than a mind attending in a crowded marketplace. A mind attending to a story told by firelight attends differently than a mind attending to a scrolling feed. The difference is not merely one of content. It is ecological. The environment shapes the attending, and the attending shapes the mind, and the mind shapes the environment it builds, and the cycle continues — for three hundred thousand years, at the pace of cultural evolution, and then, in the last three decades, at the pace of technological disruption.

The transformation of attention from a cultivated capacity into an economic resource has a specific history, and that history matters for understanding what artificial intelligence is about to do to the attentional commons. The story begins not with Silicon Valley but with the printing press, the penny newspaper, the radio, and the television — each of which represented a new technology for capturing and directing collective attention at scales previously impossible. Each also represented a new business model: the realization that if you could capture attention, you could sell access to it. The penny press discovered that newspapers could be sold below cost if advertisers would pay for access to readers' eyes. Radio discovered that broadcasts could be given away free if sponsors would pay for access to listeners' ears. Television refined the model into an art form, constructing entire narrative environments — sitcoms, dramas, news broadcasts — whose primary function was not to inform or entertain but to deliver audiences to advertisers in predictable, measurable quantities.

Citton locates the decisive conceptual shift not in any single technology but in the moment when attention itself was reconceived as a scarce resource subject to economic allocation. The intellectual origin is often traced to Herbert Simon's famous 1971 observation: "A wealth of information creates a poverty of attention." Simon's insight was genuine and important. In an information-rich environment, the limiting factor is not information but the capacity to process it. Attention becomes the bottleneck. But notice what happens when this insight is translated from observation into ideology. If attention is scarce, it must be allocated efficiently. If it must be allocated efficiently, it must be measured. If it must be measured, it must be quantified. If it must be quantified, it must be reduced to something measurable — clicks, views, time-on-page, engagement metrics. And in this reduction, something essential about attention is lost: its qualitative diversity.

This is where Citton's contribution becomes indispensable. The attention economy treats attention as a single, fungible resource — a quantity to be maximized. A click is a click. A view is a view. Sixty seconds of engagement is sixty seconds of engagement, whether those seconds are spent in contemplative absorption or irritated scrolling. Citton insists that this quantification is not merely reductive but actively destructive. It destroys attention's ecology by collapsing its diversity. In the same way that reducing a forest to "board feet of timber" erases the ecological relationships that make the forest a living system — the fungal networks, the insect populations, the water cycles, the carbon exchange — reducing attention to "time spent" erases the qualitative differences between attentional modes that make a healthy attentional ecology possible.

Citton identifies at least four distinct modes of attention, each serving different cognitive and social functions. Focused attention — the concentrated, directed, goal-oriented mode — is what most people mean when they say "paying attention." It is the mode that education privileges, that employers demand, that productivity culture celebrates. Floating attention — diffuse, receptive, open to peripheral signals — is the mode from which creative insight often emerges, the mode in which the mind wanders productively, making connections between apparently unrelated domains. Joint attention — the shared focus of two or more minds on a common object — is the foundation of communication, empathy, and collaborative thought. Collective attention — the broader cultural phenomenon of a society directing its shared focus toward common questions, crises, or celebrations — is the substrate of democratic life.

Each mode requires different environmental conditions to flourish. Focused attention requires quiet, bounded tasks, clear goals, and the absence of interruption. Floating attention requires the opposite: open time, minimal structure, the freedom to follow tangents without accountability. Joint attention requires co-presence, shared reference points, and what Citton calls "attentional coordination" — the social practices through which communities align their focus. Collective attention requires shared media, common narratives, and institutions that convene public focus around common objects of concern.

The attention economy, as it has developed through the twentieth and into the twenty-first century, has systematically cultivated one mode while degrading the others. The mode it cultivates is a degenerate form of focused attention — not the deep, sustained focus of a reader absorbed in a difficult text, but the rapid, reactive, stimulus-driven focus of a user responding to notifications, headlines, and algorithmically optimized content. This mode is measurable, monetizable, and addictive. It produces the behavioral signals — clicks, scrolls, shares — that the attention economy's business models require. The modes it degrades are precisely the ones that resist measurement: the floating attention from which creativity emerges, the joint attention from which empathy grows, the collective attention from which democratic deliberation proceeds.

Citton's ecological metaphor is not decorative. It carries analytical weight. An ecology is a system of relationships in which the health of each element depends on the health of all the others. Destroy the floating attention, and focused attention loses the creative material it needs to focus on. Destroy joint attention, and collective attention loses the interpersonal trust that makes shared focus possible. Destroy collective attention, and individual attention loses the shared meanings that make individual thought communicable. The ecology is interconnected. Degrading one mode degrades them all — not immediately, not obviously, but systemically, in the way that removing a single species from a forest does not destroy the forest overnight but sets in motion cascading effects that may take years to become visible.

This is the ecology that existed — already stressed, already degraded by decades of commercial enclosure — on the eve of the most powerful attention-shaping technology in human history. The large language models, image generators, and creative AI systems that began entering mainstream use in the early 2020s did not arrive in a pristine attentional environment. They arrived in an ecology already pushed toward monoculture — already dominated by the rapid, reactive, individually targeted mode of attention that served commercial extraction. The question Citton's framework forces is not simply "What will AI do to attention?" but "What will AI do to an attentional ecology that is already depleted?"

The answer requires understanding what AI actually introduces into the attentional environment. Previous technologies of attention capture — television, social media, smartphones — operated primarily on the demand side: they competed for existing attention, finding ever more sophisticated ways to capture and hold the attention that human minds produced. AI operates on the supply side. It does not merely compete for attention. It generates the objects of attention at a scale and speed that makes all previous content production look artisanal. A single person with a large language model can produce more text in a day than a medieval monastery produced in a year. A single person with an image generator can produce more visual content in an hour than a Renaissance workshop produced in a decade. The bottleneck was never attention alone. It was the relative scarcity of compelling content competing for that attention. AI removes that scarcity.

The ecological consequence is a flood. Not a metaphorical flood, not a gradual rise, but a sudden and overwhelming increase in the volume of material competing for placement in the attentional commons. Every concept that Citton developed to analyze the attention economy — the distinction between modes, the ecology of cultivation, the commons of collective focus — now operates in a radically altered environment. The commons is being flooded not by a rising tide of human expression but by an exponential explosion of machine-generated content optimized, by design, for individual capture.

Citton's framework suggests that the relevant question is not whether this content is good or bad, authentic or artificial, creative or derivative. The relevant question is ecological: what does this flood do to the conditions under which different modes of attention can survive? What happens to floating attention when every empty moment can be filled with AI-generated stimulus? What happens to joint attention when algorithmic personalization ensures that no two people encounter the same content? What happens to collective attention when the volume of material competing for cultural focus exceeds any community's capacity to process it together?

The answers are not yet fully visible. The flood is still rising. But Citton's ecological framework provides the conceptual tools to see what is happening beneath the surface — to understand that the crisis is not attention deficit but attention monoculture, not distraction but the systematic destruction of the attentional diversity that a healthy society requires. The commons existed before the flood. The question is whether it can survive it.

Chapter 2: Modes of Attention and the Monoculture Machine

In 1890, William James wrote what remains the most quoted sentence in the history of attention research: "Everyone knows what attention is." He then spent six hundred pages demonstrating that no one does. The sentence that follows his famous opener is almost never quoted: "It is the taking possession by the mind, in clear and vivid form, of one out of what seem several simultaneously possible objects or trains of thought." James understood, even in the late nineteenth century, that attention was not a single faculty but a selective process — a choosing, a prioritizing, a taking-possession that implied the simultaneous existence of what was not chosen, not prioritized, not possessed.

Yves Citton's contribution, more than a century later, is to insist that James's insight does not go far enough. Attention is not merely selective. It is modal. The mind does not simply choose what to attend to. It chooses how to attend — and the "how" is shaped not primarily by individual will but by the media environment in which the attending occurs. A reader in a quiet library attends in a mode characterized by sustained, deep, linear focus — what Citton calls "loyal attention," a mode of extended engagement with a single object. The same reader, the same brain, the same neural architecture, placed in front of a social media feed, attends in a radically different mode: rapid, scanning, evaluative, stimulus-responsive, constantly shifting. The difference is not willpower. The difference is ecology.

This distinction between attentional modes is Citton's most analytically powerful tool, and it is the one most systematically ignored by the attention economy's architects. The platforms that dominate contemporary media — and the AI systems now being integrated into those platforms — are designed to maximize a single metric: engagement. Engagement is measured in time, in clicks, in shares, in the behavioral signals that indicate attention has been captured. But engagement, as a metric, is mode-blind. It cannot distinguish between the engagement of a reader absorbed in a complex argument and the engagement of a user compulsively refreshing a notification feed. Both register as attention captured. Both generate the data that feeds the optimization engine. But they are fundamentally different modes of attending, with fundamentally different consequences for the person attending and for the collective attentional ecology.

Citton's taxonomy of attentional modes provides the vocabulary that the attention economy lacks. Consider, first, the distinction between what he terms "deep attention" and "hyper attention." Deep attention is characterized by sustained focus on a single object over extended time — the mode of the reader, the contemplative, the scientist working through a proof, the artist absorbed in a canvas. It is cognitively expensive, environmentally demanding, and increasingly rare. Hyper attention is characterized by rapid switching between multiple objects, a preference for high-stimulus environments, and a low tolerance for the boredom that deep attention often requires as its entry cost. Hyper attention is cognitively cheaper, environmentally ubiquitous, and increasingly dominant.

The critical point is that these are not merely different speeds of the same process. They are different processes. Deep attention and hyper attention activate different neural networks, produce different cognitive outcomes, and serve different social functions. Deep attention enables the kind of sustained analytical and creative work on which complex problem-solving depends. Hyper attention enables rapid environmental scanning, quick pattern recognition, and efficient processing of large volumes of low-complexity information. A healthy attentional ecology, Citton argues, requires both — just as a healthy biological ecology requires both the slow-growing oak and the fast-reproducing grass. The crisis is not that hyper attention exists but that hyper attention is becoming the only mode that the media environment supports.

This is where artificial intelligence enters the ecological analysis with particular force. AI-generated content is, by its nature, optimized for the hyper-attentive mode. Large language models produce text that is fluent, coherent, and immediately accessible — text that reduces the cognitive friction that deep attention requires as fuel. Image generators produce visuals that are striking, polished, and instantly legible — images that reward rapid scanning rather than sustained looking. The content is not bad. Much of it is genuinely impressive. But it is ecologically homogeneous. It is the attentional equivalent of planting an entire continent with a single high-yield crop: productive in the short term, devastating in the long term, because it eliminates the diversity on which the system's resilience depends.

Citton's framework suggests that the relevant question about AI-generated content is not "Is it as good as human-made content?" but "What mode of attention does it cultivate?" An AI-generated essay that is clear, well-organized, and immediately persuasive may be, by any individual quality metric, superior to a human essay that is rough, digressive, and difficult. But the rough, digressive, difficult essay makes demands on the reader's attention that the smooth, optimized essay does not. It requires the reader to slow down, to re-read, to puzzle, to hold ambiguity open rather than closing it prematurely. It requires, in other words, deep attention. And in requiring deep attention, it exercises and sustains the reader's capacity for deep attention — a capacity that, like any living thing, atrophies without use.

The AI-optimized essay, by contrast, slides through the reader's consciousness like water through a pipe: efficiently, pleasurably, and without residue. The reader processes it in hyper-attentive mode — scanning, absorbing, moving on. The information is transferred. The engagement metric is satisfied. But the reader's capacity for deep attention has not been exercised. It has been bypassed. And over time, across millions of such encounters, across an entire population consuming billions of such texts, the collective capacity for deep attention degrades — not because anyone chose to destroy it but because the attentional ecology no longer sustains it. The crop has displaced the forest.

Consider next Citton's concept of floating attention — the diffuse, receptive, apparently unfocused mode of awareness that operates at the periphery of consciousness. Floating attention is the mode of the daydreamer, the flâneur, the mind wandering between tasks. It is the mode that creatives and scientists consistently identify as the source of their most important insights: the shower thought, the dream connection, the idea that arrives unbidden during a walk. Floating attention is cognitively productive precisely because it is not goal-directed. It allows the mind to make connections between domains that focused attention, with its narrowed beam, cannot reach.

Floating attention requires a specific environmental condition: emptiness. It requires gaps in the stream of stimulation — moments when no notification arrives, no content is offered, no task demands completion. It requires, in Citton's language, "attentional fallow" — periods when the mind lies unworked, not because it is lazy but because the most fertile cognitive processes occur below the surface of conscious effort, in the dark soil of unfocused awareness.

AI systems are, by design, emptiness eliminators. The large language model is always available, always responsive, always ready with a suggestion, a completion, a variation, an option. The image generator fills every visual gap. The code assistant anticipates every next step. The creative AI offers itself as a perpetual companion, eliminating the loneliness and uncertainty of creative work — and, in eliminating those discomforts, eliminating the conditions under which floating attention operates. The mind that is never bored is a mind that never floats. And a mind that never floats is a mind that has lost access to the cognitive mode from which genuine novelty emerges.

The third mode Citton analyzes — joint attention — is perhaps the most politically consequential and the most threatened. Joint attention, in its simplest form, is two or more people looking at the same thing at the same time and knowing that the other is looking. It is the foundation of communication: before we can discuss an object, we must attend to it together. It is the foundation of empathy: before we can feel what another feels, we must see what they see. It is the foundation of democratic deliberation: before we can argue about a policy, we must attend to the same facts, the same evidence, the same conditions.

Joint attention requires shared objects of focus — common referents that multiple minds can orient toward simultaneously. In pre-digital media environments, these shared objects were provided by the very limitations of the medium: everyone in a village heard the same town crier, everyone in a nation watched the same evening news broadcast, everyone in a culture read the same bestselling novel. The content was often biased, often manipulated, often inadequate. But it was shared. And in being shared, it created the conditions for joint attention — the conditions under which disagreement itself was possible, because disagreement presupposes a common object about which to disagree.

Algorithmic personalization began the fragmentation of joint attention. AI-powered content generation accelerates it to a qualitatively new degree. When every user's feed is uniquely curated, when every recommendation is individually optimized, when AI can generate infinite variations of content tailored to individual preferences, the shared objects of attention dissolve. Not violently, not dramatically, but gradually, as the common referents that made joint attention possible are replaced by individually optimized streams that are by definition unshared. Each person is more attended to by the algorithm — more precisely served, more accurately targeted — and less capable of attending jointly with others. The individualization of attention is the destruction of joint attention. The optimization of the part is the degradation of the whole.

Citton's ecological framework reveals the pattern beneath these individual observations. The AI-saturated media environment is a monoculture machine. Not because it produces identical content — it produces, in fact, infinite variety — but because it cultivates a single mode of attending. The hyper-attentive, individually targeted, stimulus-responsive, evaluative mode that generates the behavioral data on which the entire system depends. Every other mode — the deep, the floating, the joint, the collective — is crowded out not by suppression but by displacement, the way an invasive species does not kill the native species directly but simply occupies every available niche until the natives have nowhere left to grow.

The monoculture machine is not a conspiracy. No one designed it to destroy attentional diversity. It emerged from the convergence of legitimate goals — efficiency, personalization, productivity, user satisfaction — pursued without ecological awareness. The engineers optimizing engagement metrics are not villains. They are farmers maximizing yield without understanding soil science. They are producing more with every passing quarter. And the ground beneath them is dying.

What Citton's framework demands is not a rejection of AI but an ecological literacy about attention — a recognition that the health of the attentional commons depends on diversity of modes, just as the health of a biological ecosystem depends on diversity of species. Protecting that diversity requires understanding what each mode needs to survive: the quiet that deep attention demands, the emptiness that floating attention requires, the shared objects that joint attention depends on, the common focus that collective attention presupposes. Every AI system, every platform design, every creative tool shapes these conditions. Every design choice cultivates or degrades. The monoculture machine runs because its builders do not see the ecology they are destroying.

Seeing it is the first step. Citton's work provides the eyes.

Chapter 3: The Flood — AI and the Tragedy of the Attentional Commons

In 1968, Garrett Hardin published "The Tragedy of the Commons" in Science, describing a pasture open to all, in which each herder, acting rationally in their own interest, adds one more cow, then one more, then one more — until the pasture is destroyed by the very rationality that each individual herder exercised. The tragedy is not caused by malice or stupidity. It is caused by a structural disconnect: the benefit of adding one cow accrues entirely to the individual herder, while the cost of the degradation is distributed across all herders. Rational individual action produces collective catastrophe.

Yves Citton's treatment of attention as a commons makes Hardin's tragedy newly relevant — and newly terrifying. If collective attention is a commons, a shared resource on which democratic life, cultural coherence, and social trust depend, then the introduction of AI-powered content generation into this commons represents something unprecedented in the history of media: the sudden, radical, and potentially irreversible expansion of every individual's capacity to graze.

Before AI, the attentional commons was grazed primarily by professional content producers — journalists, artists, writers, filmmakers, advertisers — whose output, while enormous, was bounded by human productive capacity. A journalist could write a few articles a day. A novelist could produce a book a year. A studio could release a few dozen films annually. The commons was crowded, certainly. The attention economy had already created intense competition for collective focus. But the competition was, in a sense, artisanal. Each piece of content represented a significant investment of human time, skill, and intention, and this investment created a natural limit on supply.

AI removes this limit. A single person with access to a large language model can produce text at a rate that would have required, just five years ago, a team of dozens. A single person with an image generator can flood visual platforms with material that would have required a studio of artists. A single person — or, more precisely, a single automated system requiring no person at all — can generate, publish, and distribute content continuously, indefinitely, without rest, without cost, without the biological constraints that previously set a floor beneath the supply side of the attention economy.

The ecological consequence, in Citton's framework, is not merely more competition for attention. It is a qualitative transformation of the commons itself. When the volume of content competing for placement in collective attention exceeds any community's capacity to process it, something fundamental changes. The commons does not merely become more crowded. It becomes unparsable. The signal-to-noise ratio collapses not because the noise is louder but because the total volume of signal and noise together exceeds any collective capacity for filtration. The commons, in ecological terms, is eutrophied — choked by an overabundance of the very nutrient (content, information, stimulation) that it requires in moderate quantities to thrive.

Eutrophication is the precise ecological analogy. In a lake, eutrophication occurs when an excess of nutrients — typically nitrogen and phosphorus from agricultural runoff — causes explosive growth of algae. The algae bloom on the surface, blocking sunlight from reaching deeper waters. Aquatic plants beneath the surface die. As they decompose, they consume oxygen. Fish suffocate. The lake, from the outside, looks green and lush — more alive than ever. From the inside, it is dead. The explosion of surface growth has killed the deep ecology that the lake's health depended on.

Citton's attention ecology suggests the same dynamic at work in the AI-flooded attentional commons. The surface is more active than ever. More content is being produced, more engagement is being generated, more metrics are climbing. But beneath the surface — in the deep ecology of sustained collective focus, shared cultural reference, and joint attention — the oxygen is disappearing. The bloom of AI-generated content captures surface attention with extraordinary efficiency. Each piece is optimized, personalized, immediately engaging. And each piece, in capturing individual attention, withdraws that attention from the collective pool. The commons is not being consumed by a few large grazers. It is being consumed by an infinite number of individually optimized micro-grazers, each extracting a tiny portion of the shared resource, and the aggregate extraction exceeds what the commons can sustain.

The tragedy unfolds along three dimensions that Citton's framework illuminates with particular clarity. The first is temporal. Collective attention operates on timescales that individual attention does not. A society's capacity to attend collectively to a crisis — a pandemic, a democratic election, a moral reckoning — requires that the crisis remain in collective focus long enough for deliberation to occur, positions to form, and consensus (or principled disagreement) to emerge. This process takes weeks, months, sometimes years. It requires what Citton calls "attentional durability" — the capacity of a shared object of focus to maintain its place in collective attention against the competition of newer, fresher stimuli.

AI-generated content accelerates the replacement cycle of collective attention. When the volume of new material entering the commons increases by orders of magnitude, each object's dwell time in collective focus decreases correspondingly. The crisis that might have held national attention for a month now holds it for a week. The cultural conversation that might have developed over a season exhausts itself in days. Not because people care less but because the environment offers more: more takes, more responses, more variations, more AI-generated analyses and summaries and hot pieces that each capture a fragment of attention and collectively shatter the sustained focus that deliberation requires. The temporal commons — the shared time during which a society attends together — is compressed by the sheer volume of material competing to fill it.

The second dimension is spatial, or more precisely, topological. Collective attention requires convergence — the focusing of many individual attentional streams toward a common point. Citton's concept of joint attention, scaled to the level of a society, describes moments when millions of minds orient toward the same object: a news event, a cultural phenomenon, a shared question. These moments of convergence are not merely informational. They are constitutive. They create the sense of shared reality that makes a society a society rather than a collection of isolated individuals who happen to occupy the same geography.

AI-driven personalization is a divergence engine. Its fundamental operation is the opposite of convergence: it takes a common input — a user base, a content library, a cultural moment — and produces individualized outputs tailored to each person's preferences, history, and predicted engagement patterns. The result is that the same cultural moment is experienced differently by every individual. The same news event is framed differently in every feed. The same question is answered differently by every AI assistant, calibrated to what the individual user wants to hear. The divergence is not absolute — shared referents persist, cultural moments still break through — but the trend line is clear. Each improvement in personalization is a step away from convergence. Each step away from convergence degrades the topological conditions that joint attention requires.

The third dimension is qualitative, and it connects directly to Citton's analysis of attentional modes. The content that dominates an eutrophied attentional commons is not the content that best serves collective attention. It is the content most efficiently optimized for individual capture. These are not the same thing, and the difference matters enormously. Content optimized for individual capture is designed to be immediately engaging, emotionally activating, and rapidly consumable. It rewards hyper attention. Content that serves collective attention is often the opposite: complex, ambiguous, resistant to quick consumption, requiring sustained focus and collective interpretation. It demands deep attention and rewards joint attention. In an eutrophied commons, the former displaces the latter — not because anyone decided to suppress complex content but because the optimization dynamics of the commons favor what captures quickly over what rewards slowly.

Citton frames this displacement as a form of ecological succession in reverse. In a healthy ecosystem, ecological succession moves toward complexity: simple pioneer species give way to more complex communities, which give way to mature ecosystems characterized by diversity, stability, and intricate interdependence. In the AI-flooded attentional commons, the succession runs backward. The complex, diverse, interdependent ecology of attentional modes gives way to simpler, faster, less differentiated forms. The forest becomes grassland. The grassland becomes bare soil. The bare soil, colonized by the fastest-growing species, becomes a monoculture. The end state is not emptiness but a different kind of fullness — a commons saturated with content and devoid of meaning, blooming on the surface and dead in the depths.

The structural parallel to the original tragedy of the commons is precise. Each individual actor — each person using AI to produce content, each platform deploying AI to generate engagement, each business automating its communication — acts rationally. The content produced is often valuable to its intended audience. The engagement generated is genuine. The communication is efficient. Each cow added to the pasture is, individually, a reasonable decision. But the aggregate effect is the degradation of the shared resource on which all individual acts of communication depend: the capacity for collective attention.

Citton's ecological framework reveals something that purely economic analyses of the attention economy miss: the degradation is not primarily about quantity. It is about diversity. A pasture can sustain a large number of cows if the pasture's ecology is diverse — if different species of grass grow at different rates, if the soil's microbiome remains healthy, if the water cycle is intact. The pasture fails not when the number of cows exceeds some absolute limit but when the grazing pattern destroys the ecological diversity on which the pasture's regenerative capacity depends. Analogously, the attentional commons can sustain enormous volumes of content if the ecology of attention remains diverse — if deep, floating, joint, and collective attention all continue to function. The commons fails not when content volume exceeds some absolute limit but when the content's optimization for a single attentional mode destroys the modal diversity on which the commons' regenerative capacity depends.

This reframing has practical consequences. If the problem were merely quantitative — too much content — the solution would be filtration: better algorithms, smarter curation, more effective gatekeeping. If the problem is ecological — the wrong kind of content cultivating the wrong mode of attention — then filtration is necessary but insufficient. The solution must also include cultivation: the active creation and protection of environments that sustain the attentional modes that the optimization engine degrades. This means preserving spaces of emptiness in which floating attention can operate. It means creating shared objects of focus around which joint attention can form. It means protecting temporal durations during which collective attention can develop. It means, in Citton's terms, treating the attentional commons not as a market to be optimized but as an ecology to be tended.

The flood is already here. The AI systems generating content at superhuman speed are already operating. The eutrophication of the attentional commons is already underway. What Citton's framework offers is not a way to stop the flood — that ship has sailed, that dam has broken — but a way to understand what the flood is destroying and, therefore, what must be actively rebuilt. The commons before the flood was already degraded. The commons during the flood is being transformed. The question that matters now is what the commons after the flood will look like — and whether anyone is tending the ecology that will determine its answer.

Chapter 4: The Smooth and the Fertile — Attention, Aesthetics, and the Death of Surprise

Somewhere in the first decade of AI-generated imagery, a strange consensus emerged. The images were beautiful. They were technically accomplished, compositionally balanced, chromatically harmonious, and immediately striking. They could be produced in seconds, in infinite variety, at no marginal cost. And they were, in some way that was easier to feel than to articulate, dead on arrival.

The deadness was not a flaw. It was a feature — not in the cynical sense of a design choice disguised as a limitation, but in the deeper sense of a quality that emerged necessarily from the optimization process that produced the images. An AI image generator trained on millions of human-made images learns, with extraordinary precision, what visual features correlate with human approval: clean composition, vivid color, dramatic lighting, emotional clarity. It learns to produce images that satisfy. And in learning to satisfy, it learns to eliminate the one quality that Yves Citton's attention ecology identifies as essential to the health of the attentional commons: the capacity to surprise.

Surprise, in Citton's framework, is not a pleasant bonus that attention occasionally encounters. It is the mechanism by which attention renews itself. A mind that is never surprised is a mind whose attentional patterns have calcified — fixed in their ruts, running along grooves worn smooth by repetition, incapable of the reorientation that genuine learning requires. Surprise is the attentional equivalent of disturbance in an ecological system: the storm that topples old trees and creates openings in the canopy, the fire that clears deadwood and releases nutrients locked in accumulated debris, the flood that reshapes the riverbed and creates new habitats. Without disturbance, an ecosystem stagnates. Without surprise, an attentional ecology stagnates. Both become closed systems, recycling their own products, trending toward entropy.

The aesthetic quality that Citton's ecological analysis identifies in AI-generated content — and that the philosopher Byung-Chul Han diagnoses with surgical precision as "smoothness" — is the visual signature of an attentional monoculture. Smooth content is content from which all friction has been removed: all roughness, all resistance, all the qualities that slow consumption and demand active engagement. The smooth image slides through the visual cortex without catching on anything. The smooth text glides through comprehension without requiring re-reading. The smooth melody satisfies the ear without disturbing it. Each encounter is pleasant, effortless, immediately gratifying. And each encounter leaves the attentional ecology exactly as it found it — undisturbed, unrenewed, unmoved.

Citton's analysis of this smoothness differs from Han's in a crucial respect. Where Han locates the problem primarily in the aesthetic object — in the qualities of the smooth thing itself — Citton locates it in the ecology of attention within which the object circulates. A smooth object encountered occasionally, within a diverse attentional environment that also includes rough, difficult, and surprising objects, does no ecological harm. It is one species among many. The problem arises when smooth objects dominate the environment — when the optimization dynamics of AI-driven content production ensure that smooth objects proliferate while rough objects are marginalized, not by censorship but by competitive displacement. The smooth objects capture attention more efficiently. They generate higher engagement metrics. They are favored by algorithms trained to maximize those metrics. And so they multiply, crowding out the rough, the difficult, the surprising, the resistant — until the attentional environment is a monoculture of smoothness, and the capacity for surprise has been, quite literally, selected out.

The selection dynamic is worth examining in detail, because it reveals how AI-generated content degrades the attentional commons even when each individual piece is of high quality. Consider the production of visual art. Before AI, visual production was constrained by human labor. An artist producing a painting invested hours, days, weeks of focused attention in its creation. This investment meant that the artist's idiosyncrasies — their particular way of seeing, their specific motor habits, their unique neurological wiring — were inscribed in the work. The painting bore the marks of a particular embodied mind engaging with particular materials under particular conditions. These marks — the unplanned drip, the imperfect line, the color choice that emerged from a specific palette limitation — were not noise to be eliminated. They were the signatures of genuine encounter, the evidence that a mind had grappled with resistance and been changed by the grappling.

AI-generated imagery eliminates this grappling. The image emerges from statistical pattern-matching across millions of training examples, converging on the visual features most strongly associated with human approval. The result is an image that is, in a precise statistical sense, average — not mediocre, but optimized toward the center of the distribution of human aesthetic preferences. It is an image that satisfies the most people, most of the time, most efficiently. And it is, for exactly that reason, an image that surprises no one. Surprise lives at the edges of the distribution — in the unexpected choice, the idiosyncratic vision, the mistake that opens a new possibility. Optimization drives toward the center. The center, by definition, is where surprise goes to die.

Citton's ecological vocabulary allows this observation to be extended from individual aesthetic objects to the attentional environment as a whole. A single AI-generated image, placed alongside human-made images in a gallery, might be indistinguishable or even superior by conventional aesthetic criteria. But a million AI-generated images, flooding a visual platform, create an environment in which the aesthetic center — the smooth, the optimized, the unsurprising — becomes the norm, and everything that deviates from that norm becomes, perceptually, an anomaly. The rough, the difficult, the surprising become not merely rare but strange — artifacts of an older mode of production, easily identified and easily dismissed as unpolished, amateur, unoptimized.

This perceptual shift is the ecological catastrophe. When the surprising becomes strange rather than stimulating, the attentional ecology has lost its capacity for renewal. The mind encountering rough, resistant content in a smooth environment does not respond with the openness that surprise requires. It responds with impatience, irritation, or incomprehension — the responses of an attentional system that has adapted to smoothness and can no longer process friction productively. The muscle of surprise has atrophied. The ecology that sustained it has been replaced. The attentional monoculture is complete.

Citton's framework connects this aesthetic analysis to a political one through the concept of what might be called attentional fertility — the capacity of an attentional environment to produce genuine novelty, unexpected connections, and transformative insights. Fertility, in biological ecology, depends on diversity. A fertile soil contains billions of microorganisms from thousands of species, engaged in complex chemical exchanges that no single species could perform alone. A fertile attentional environment contains diverse modes of attention engaged in complex cognitive exchanges that no single mode could produce alone. Deep attention provides the sustained focus needed to develop an idea. Floating attention provides the unexpected connections that deep attention cannot reach. Joint attention provides the social testing ground where ideas are refined through encounter with other minds. Collective attention provides the shared context within which individual ideas acquire cultural meaning.

Smoothness degrades fertility by simplifying the ecology. When all content is optimized for the same attentional mode — the rapid, evaluative, individually targeted mode that engagement metrics capture — the other modes languish. Deep attention, starved of objects that reward sustained focus, retreats. Floating attention, crowded out by the continuous stream of optimized stimulation, has no empty space in which to operate. Joint attention, undermined by personalization, loses the shared objects around which it forms. Collective attention, fragmented by the sheer volume and variety of individually targeted content, cannot converge. The ecology simplifies. The fertility declines. The monoculture produces more and more of the same.

The parallel to agricultural monoculture is instructive and precise. Industrial agriculture's great achievement was maximizing yield — producing more food per acre than any previous agricultural system. Its great cost was ecological: the destruction of soil biodiversity, the depletion of topsoil, the dependence on external inputs (fertilizer, pesticide, irrigation) to maintain yields that the degraded soil could no longer support on its own. The monoculture produces abundantly while destroying the conditions of its own long-term productivity. Each harvest extracts from the soil without replenishing. Each season's yield comes at the cost of the next season's fertility.

AI-driven content production follows the same pattern. Each optimization cycle produces content that captures attention more efficiently. Each increase in efficiency attracts more attention to the optimized content and away from the diverse, rough, surprising content that the attentional ecology needs to renew itself. Each shift in the distribution of attention degrades the ecology further, requiring still more optimization to capture attention in an increasingly depleted environment. The system accelerates its own degradation. The monoculture produces more while the soil dies beneath it.

Citton's prescription is not to abandon optimization but to cultivate the conditions that optimization degrades. In agriculture, the parallel movement is regenerative farming: practices that rebuild soil health by restoring diversity, reducing external inputs, and allowing the ecosystem's own processes to regenerate the conditions of fertility. In the attention ecology, the parallel would be what Citton calls "attentional cultivation" — practices, institutions, and environmental designs that actively sustain the modes of attention that the optimization engine displaces.

What would attentional cultivation look like in practice? Citton's work suggests several directions. First, the protection of emptiness — the deliberate preservation of spaces and times in which no content is offered, no notification arrives, no AI system presents options for evaluation. These are not dead zones but fallow fields: environments in which floating attention can operate and creative surprise can germinate. Second, the creation of shared objects — cultural events, civic rituals, common texts, public spaces designed to convene joint attention rather than fragment it. Third, the support of difficulty — the active championing of content that resists immediate consumption, that requires re-reading, re-viewing, sustained engagement, the kind of cognitive effort that deep attention needs as exercise. Fourth, and perhaps most fundamentally, the cultivation of an ecological consciousness about attention itself — an awareness, distributed throughout the population, that how we attend is not a private matter but a public one, not a consumer choice but a civic responsibility, not a resource to be spent but a capacity to be tended.

The smooth aesthetic is seductive because it gives each individual exactly what they want in the moment of encounter. It is ecologically devastating because what each individual wants in the moment of encounter is not what the collective attention ecology needs to sustain itself over time. This is the fundamental tension that Citton's framework reveals: the optimization of individual attentional satisfaction and the health of the collective attentional ecology are, under current technological conditions, in direct conflict. Each person's reasonable desire for content that is immediately engaging, personally relevant, and frictionlessly consumable contributes to the degradation of the shared attentional environment on which meaning, creativity, and democratic life depend.

The death of surprise is not a dramatic event. No one announces it. No metric captures it. It is, rather, a gradual dimming — a slow narrowing of the range of what the collective attentional ecology can produce and sustain. The images become more beautiful and less startling. The texts become more fluent and less challenging. The ideas become more accessible and less transformative. The ecology simplifies. The soil thins. The harvest continues, season after season, each one slightly less nourishing than the last, and no one notices because each individual piece of fruit is perfectly smooth, perfectly ripe, and perfectly indistinguishable from the one that came before.

Citton's work asks whether anyone is willing to prefer the rough fruit — the one with the blemish, the unexpected flavor, the seed that might, planted in the right soil, grow into something no optimization algorithm could have predicted. The answer to that question will determine whether the attentional ecology recovers or completes its collapse. The fruit is on the table. The soil is waiting. The question is who will do the planting.

Chapter 5: The Evaluative Mind and the Death of Dwelling

In the workshops of Renaissance Florence, an apprentice painter might spend three years grinding pigments before being permitted to touch a brush to panel. The grinding was not punishment. It was attentional training — a slow, embodied immersion in the material conditions of the craft that cultivated a specific mode of attending: patient, sensory, intimate with resistance. The apprentice who had spent a thousand hours grinding lapis lazuli into ultramarine understood, in a way no instruction could convey, what that blue could and could not do. The knowledge was not propositional but attentional. It lived in the hands, in the quality of focus that hours of repetitive physical engagement had shaped. When the apprentice finally painted, the attending had already been formed. The ecology of the workshop — its slowness, its hierarchy, its demand for years of apparently unproductive labor — was not an obstacle to creativity. It was the environment in which a particular mode of creative attention could grow.

Yves Citton's framework for understanding attentional modes gains its sharpest diagnostic power when applied to the specific transformation that AI tools introduce into creative practice. The transformation is not primarily about quality of output. It is about mode of attending. AI creative tools — large language models that generate text, image generators that produce visuals, code assistants that write software — do not merely accelerate production. They restructure the attentional ecology of creative work by replacing one dominant mode of attending with another. The mode they displace is what might be called generative attention: the slow, uncertain, groping-in-the-dark mode in which the creator dwells with a problem, lives inside its difficulty, follows paths that lead nowhere, and occasionally, through persistence and luck, arrives somewhere genuinely new. The mode they install is evaluative attention: the rapid, comparative, selection-oriented mode in which the creator scans options, judges quality, and chooses among pre-generated possibilities.

The distinction is not subtle. It is the difference between writing a sentence and choosing between five sentences. Between composing an image and selecting from twenty variations. Between solving a problem and evaluating solutions. Each pair describes a fundamentally different cognitive activity, requiring fundamentally different attentional conditions, producing fundamentally different experiential and developmental outcomes. Citton's ecological vocabulary makes the stakes visible: generative attention and evaluative attention are different species in the attentional ecosystem. They occupy different niches, require different conditions, and serve different functions. A creative ecology that supports both is diverse and resilient. A creative ecology that eliminates one in favor of the other is a monoculture — productive in the short term, fragile and impoverished in the long term.

The mechanism of displacement is not coercion. No one forces the creator to abandon generative attention. The displacement occurs through what Citton might call environmental pressure — the same invisible, structural, seemingly neutral force that causes one species to displace another when habitats change. The AI tool offers itself as a collaborator, and the collaboration is genuine. The text it generates is often good. The images it produces are often striking. The code it writes often works. The creator who engages with these outputs enters evaluative mode naturally, effortlessly, and with immediate reward: the satisfaction of choosing well, of curating effectively, of assembling something polished from high-quality components. The evaluative mode is pleasurable. It produces visible results. It generates the dopamine of completion. And it is, in every measurable sense, more efficient than the generative mode it replaces.

The efficiency is real. This must be acknowledged without flinching, because the argument against evaluative monoculture cannot rest on a denial of its productive power. A writer using a large language model to generate draft paragraphs, then selecting and refining the best ones, can produce polished text at five or ten times the rate of a writer composing from scratch. A designer using an image generator to produce variations, then curating the most compelling options, can develop visual concepts at a pace that would have been inconceivable a decade ago. The output is real. The speed is real. The freed-up time is real. Anyone who dismisses these gains as illusory has not used the tools.

But Citton's ecological framework insists on a question that efficiency metrics cannot answer: what is happening to the generative mode while the evaluative mode flourishes? What is happening to the cognitive capacity for dwelling — for sitting with a problem long enough that it begins to reveal dimensions invisible to the scanning eye? What is happening to the tolerance for uncertainty, for the discomfort of not-yet-knowing, for the long stretches of apparent unproductivity that generative attention requires as its medium?

The answer, emerging from both Citton's theoretical framework and the lived experience of creators working with AI tools, is that generative attention is atrophying. Not dramatically, not overnight, but in the slow, quiet way that any capacity atrophies when it is no longer exercised. The writer who habitually generates first drafts with AI assistance finds, after months of this practice, that the blank page has become more intimidating than it was before — not less. The neural pathways that once carried the writer through the anxiety of the empty screen, through the false starts and dead ends, through the long apprenticeship of sentences that do not work until suddenly one does — those pathways have been bypassed so consistently that they have begun to fade. The writer can still evaluate brilliantly. The writer's generative capacity has dimmed.

Citton would recognize this as a classic ecological dynamic: habitat destruction through displacement. The generative mode's habitat is difficulty. Its habitat is the blank page, the unsolved problem, the silence that precedes speech. When AI tools fill that silence with options, they do not destroy the habitat violently. They develop it. They build on it. They improve it, in the way that draining a wetland and planting a crop improves the land's productivity — destroying, in the improvement, the ecological functions that the wetland performed. The wetland was not productive in any measurable sense. It filtered water. It absorbed floods. It sustained species that existed nowhere else. Its value was ecological, systemic, visible only when it was gone.

The generative mode of attention is the creative wetland. Its value does not appear in any efficiency metric. It cannot be measured in words per hour or images per session. Its value is developmental: it builds the cognitive and emotional capacities that make genuine creative work possible. The capacity to tolerate uncertainty. The capacity to follow an intuition without knowing where it leads. The capacity to fail, repeatedly, and to learn from failure things that success cannot teach. The capacity to dwell — to remain with a problem not because you have been assigned to it but because something in the problem has captured you, and you cannot leave until you have understood it, and the understanding comes not on your schedule but on its own.

Dwelling is the word that Citton's framework most needs and that the attention economy most thoroughly erases. To dwell is to remain in a place — an intellectual place, an emotional place, a creative place — longer than efficiency requires. To dwell with a difficult passage in a book is to re-read it not because you failed to understand it the first time but because it rewards repeated attention, revealing new dimensions with each encounter. To dwell with a creative problem is to resist the temptation of the first adequate solution and to continue exploring even after something workable has appeared. To dwell is to attend not for the purpose of extracting value but for the purpose of deepening relationship. Dwelling is the mode of attention in which meaning accumulates.

AI tools are structurally hostile to dwelling. Not intentionally — their designers did not set out to destroy the human capacity for sustained creative engagement. But the tools' fundamental affordance is the generation of options, and options are the enemy of dwelling. When a writer is struggling with a paragraph — genuinely struggling, in the productive, generative way that Citton's framework identifies as cognitively essential — the struggle itself is the work. The difficulty is not an obstacle to be overcome but a medium to be inhabited. The writer who dwells in that difficulty long enough will either find a way through that could not have been found quickly, or will discover that the difficulty itself was the signal — that the paragraph should not exist, that the argument needs to go somewhere else entirely, that the struggle was the message.

The AI tool, presented with the same difficulty, offers five clean alternatives in three seconds. Each is fluent. Each resolves the struggle. Each eliminates the dwelling. And each, in eliminating the dwelling, eliminates the cognitive process through which the writer would have learned something that no amount of evaluating pre-generated options can teach: what it feels like to find your own way through difficulty, and how that finding changes both the work and the worker.

Citton's concept of attentional cultivation becomes crucial here. Cultivation, in the ecological sense, is the deliberate shaping of an environment to support the growth of desired species. To cultivate generative attention in an AI-saturated creative environment requires understanding what generative attention needs to survive — and actively providing those conditions, even when they are inefficient, even when they are uncomfortable, even when the AI tool is right there, offering to make the difficulty disappear.

The conditions are specific. Generative attention requires time without predetermined outcome — hours, sometimes days, in which the creator works without knowing whether the work will produce anything usable. It requires tolerance for failure, not as a regrettable byproduct of creative work but as its essential nutrient. It requires silence — not literal silence but informational silence, the absence of suggestions, completions, alternatives, and options that evaluative attention feeds on. It requires, in a word, emptiness. The same emptiness that Citton identifies as essential to floating attention, the same emptiness that the AI-saturated environment systematically eliminates.

The uncomfortable implication is that the most productive creative use of AI tools may require the deliberate, disciplined, countercultural practice of not using them — at least during the generative phase of creative work. This is not a Luddite position. It is an ecological one. The farmer who understands soil science does not refuse to use fertilizer. The farmer uses fertilizer strategically, in the right amounts, at the right times, on the right fields — and leaves other fields fallow, because the farmer understands that long-term productivity depends on practices that look, in the short term, like waste. The creator who understands attentional ecology does not refuse to use AI tools. The creator uses them strategically — for evaluation, for refinement, for the expansion of possibilities that evaluative attention excels at processing — and protects the generative phase from their intrusion, because the creator understands that long-term creative capacity depends on attentional practices that look, in the short term, like inefficiency.

This is harder than it sounds. The AI tool does not knock and wait to be invited. It is integrated into the creative environment itself — embedded in the text editor, the design application, the development environment. It offers suggestions before they are requested. It completes sentences before they are finished. It proposes images before the visual imagination has had time to generate its own. The environmental pressure is constant, ambient, and almost impossible to resist through willpower alone. Citton's framework suggests that the solution is not individual discipline but environmental design — the deliberate construction of creative spaces in which generative attention is protected by the structure of the environment itself, not by the fragile willpower of the individual creator working against the grain of every tool at their disposal.

What this looks like in practice is the subject of ongoing experiment among creators who have recognized what Citton's framework describes. Some maintain separate working environments — one with AI tools, one without. Some impose temporal boundaries — generative mornings without AI, evaluative afternoons with it. Some use AI tools only for tasks they have already completed independently, treating the tool as a second opinion rather than a first draft. Each of these practices is an act of attentional cultivation — a deliberate shaping of the creative ecology to preserve the diversity of modes on which long-term creative health depends.

The Renaissance workshop understood this without needing a theory. The three years of grinding pigments was not wasted time. It was attentional infrastructure — the patient construction of a mode of attending that no shortcut could produce. The apprentice who skipped the grinding and went straight to painting might have produced adequate work. But the apprentice would have been evaluating rather than dwelling, selecting rather than discovering, performing creativity rather than inhabiting it. The workshop's ecology was designed to prevent exactly this substitution. The question Citton's framework poses to the present moment is whether any contemporary creative ecology can be designed with comparable wisdom — and whether the builders of AI tools have any incentive to design it so.

The incentive structure, as it stands, points in the opposite direction. Every AI tool is optimized for adoption, and adoption is measured by use, and use is maximized when the tool is always present, always helpful, always filling the silence that generative attention requires. The monoculture machine runs on the elimination of emptiness. And the death of dwelling proceeds, quietly, in workshops that have forgotten why the grinding mattered.

Chapter 6: Joint Attention and the Dissolution of the Common World

On the evening of July 20, 1969, an estimated 600 million people watched Neil Armstrong step onto the surface of the Moon. The broadcast was grainy, the audio crackled, the image was a ghostly blur of light and shadow. By any contemporary standard of media quality, it was terrible content. But 600 million people attended to it simultaneously, and in that simultaneous attending, something happened that no amount of high-definition, individually optimized, algorithmically targeted content has ever replicated: the formation of a planetary moment of joint attention.

Yves Citton's concept of joint attention — borrowed from developmental psychology, where it describes the foundational capacity of infants and caregivers to attend to the same object together — is extended in his work to encompass the political and cultural dimensions of shared focus. Joint attention, in Citton's expanded usage, is not merely two or more minds looking at the same thing. It is two or more minds knowing that they are looking at the same thing, and knowing that the other knows, and orienting their subsequent thought and action in light of that shared awareness. Joint attention is the cognitive infrastructure of the common world — the world that exists not inside any individual mind but between minds, in the shared space that mutual attending creates.

The common world, in this formulation, is not given. It is constructed — continuously, effortfully, through the practices and technologies that enable joint attention. A town square constructs common world by creating a physical space where people encounter the same sights, the same sounds, the same events. A newspaper constructs common world by creating a shared informational object that readers can discuss, argue about, and reference in conversation. A television broadcast constructs common world by synchronizing attention across geographical distance. Even when these technologies were used for propaganda, even when they distorted reality, even when they served the powerful at the expense of the powerless, they performed this constructive function: they created shared objects of attention around which the common world could crystallize.

Citton's analysis of joint attention reveals something that might otherwise remain invisible: the common world has material conditions. It does not emerge spontaneously from the mere existence of multiple minds. It requires specific media configurations — technologies, institutions, practices — that make shared attending possible. And it can be destroyed, not by any deliberate act of demolition but by the gradual erosion of the conditions that sustain it. The common world does not collapse. It evaporates. It becomes thinner, less substantial, less capable of supporting the weight of democratic deliberation, cultural coherence, and mutual understanding that a functioning society places upon it.

The evaporation has been underway for decades, accelerated by each successive wave of media personalization. Cable television fragmented the broadcast audience. The internet fragmented it further. Social media, with its algorithmic curation of individual feeds, advanced the fragmentation to a degree that would have seemed pathological to media theorists of the mid-twentieth century. But each of these technologies still operated within a constraint that preserved some residual capacity for joint attention: all of them distributed content that humans had created. The supply of content, while growing rapidly, was still bounded by human productive capacity. And within that bounded supply, certain objects — major news events, viral phenomena, culturally dominant entertainments — still achieved sufficient penetration to function as shared referents. People could still argue about the same movie, react to the same scandal, mourn the same tragedy. The common world was thinning, but it had not yet disappeared.

AI-generated content removes the constraint that kept the common world, however thinly, in existence. When every user can be served content generated specifically for them — not selected from a common pool but created on demand, tailored to individual preferences, history, and behavioral patterns — the shared referent dissolves entirely. Not because anyone decided to dissolve it, but because the economics of attention capture favor individual optimization over collective coherence. An AI system that generates a unique article for each reader, calibrated to that reader's interests, reading level, and ideological predispositions, captures more individual attention than a single article written for a general audience. The optimization works. Each individual is better served. And the object around which joint attention might have formed — the common article, the shared text, the thing that two strangers could both have read — ceases to exist.

Citton's framework reveals the paradox at the heart of this optimization: the system that maximizes individual attentional satisfaction destroys collective attentional capacity. Each person is more precisely attended to by the algorithm than at any previous moment in human history. Each person is less capable of attending jointly with others than at any previous moment in human history. The individual experience is enriched. The collective capacity is impoverished. The pasture is green for each cow. The commons is dead.

The consequences of joint attention's dissolution are not abstract. They are visible in the specific pathologies of contemporary public life that Citton's framework illuminates with uncomfortable precision. Democratic deliberation presupposes joint attention: the capacity of citizens to focus on the same problems, consider the same evidence, and argue within a shared framework of reference. When joint attention dissolves, deliberation does not become more difficult. It becomes structurally impossible. Not because people disagree — disagreement is the substance of democratic life — but because they no longer attend to the same reality. They have no common object about which to disagree. Each inhabits an individually optimized informational environment that shares fewer and fewer reference points with anyone else's. The result is not polarization, which implies two poles oriented toward a common center, but fragmentation — the splintering of the attentional commons into individual shards that share no common edge.

Citton's analysis suggests that what is commonly diagnosed as a crisis of trust is actually a crisis of joint attention. People do not distrust institutions because they have been given good reasons to distrust them — though they often have. They distrust institutions because they no longer attend to the same institutional reality. The citizen whose informational environment is constructed entirely by algorithms encounters a version of institutional reality so different from the version encountered by a citizen with a different algorithmic profile that the two citizens might as well be living in different countries. Trust between them is not broken. It was never formed, because trust requires the shared attending from which mutual recognition grows, and shared attending requires common objects of focus that the personalization engine systematically eliminates.

The AI amplification of this dynamic introduces a new and specifically dangerous element: the capacity for synthetic joint attention. An AI system can generate content that simulates the experience of shared focus — creating the appearance of a common world without the substance. Deepfake videos, synthetic news articles, AI-generated social media personas, and algorithmically coordinated campaigns can produce the feeling that "everyone is talking about this" when in fact the "everyone" is partly or wholly manufactured. This synthetic joint attention is more dangerous than the mere absence of joint attention, because it produces the social effects of shared focus — mobilization, emotional contagion, consensus formation — without the epistemic foundation that genuine joint attention provides. People act as though they have attended to the same reality. They have attended to different synthetic realities designed to produce the same behavioral response.

Citton's ecological vocabulary provides the critical distinction: genuine joint attention is a commons — a shared resource produced by and for the community that sustains it. Synthetic joint attention is an enclosure — a privately controlled simulation of the commons designed to extract value from the behavioral responses it generates. The difference is not visible from inside the experience. The citizen mobilized by synthetic joint attention feels exactly as engaged, exactly as connected, exactly as part of a community of shared concern as the citizen mobilized by genuine joint attention. The difference is structural, ecological, visible only from the systemic perspective that Citton's framework provides.

The dissolution of the common world proceeds along a specific temporal logic that Citton's framework helps to trace. First, the shared objects of attention are replaced by individually optimized objects. Second, the awareness of sharing — the knowledge that others are attending to the same thing — fades. Third, the social practices that depended on shared awareness — conversation, deliberation, collective action — lose their foundation. Fourth, the institutions that organized collective attention — newspapers, broadcast media, public forums — lose their audience and their authority. Fifth, the capacity for joint attention itself atrophies, as the neural and social pathways that supported it are no longer exercised. The dissolution is not a single event but a cascade, each stage enabling the next, and the process is well advanced before most people notice that the common world has become thinner.

What Citton's ecological framework demands, in the face of this dissolution, is not nostalgia for the broadcast era — an era that had its own severe distortions of the attentional commons — but a recognition that joint attention is a commons that requires active maintenance. It does not sustain itself. It does not emerge naturally from the mere coexistence of multiple minds. It requires what Citton calls "attentional institutions" — media structures, social practices, and design principles that create and protect shared objects of focus. Public broadcasting, shared civic spaces, common educational curricula, cultural events that convene collective attention around common objects — these are not luxuries or anachronisms. They are the infrastructure of the common world. And they are under threat not from any ideology that opposes them but from an optimization logic that does not see them, does not value what they produce, and replaces them, relentlessly and without malice, with individually superior alternatives whose collective consequence is the evaporation of shared reality.

The Moon landing was terrible content by every metric the attention economy values. Low resolution, poor audio, no personalization, no interactivity, no algorithmic optimization. Six hundred million people watched it together, and in watching it together, they constituted — however briefly — a common world. The ecology of attention that made that moment possible is not gone. But it is dying, quietly, in the space between one individually optimized feed and the next. And the AI systems now flooding that space with infinite, personalized, superbly engaging content are not building a new common world. They are building six hundred million private ones, each perfectly appointed, each empty of everyone else.

The question Citton's framework leaves open — and it is a genuinely open question, not a rhetorical one — is whether the common world can be reconstructed under conditions of AI abundance, or whether the dissolution is irreversible. The tools exist to build attentional institutions adequate to the new environment: AI systems designed to create shared objects of focus rather than individually optimized ones, platforms that prioritize collective attention alongside individual engagement, design principles that treat joint attention as a value to be measured and protected. The tools exist. The incentives do not. And the tragedy of the commons, as Hardin observed, is precisely that rational individual action — in this case, the rational optimization of individual attention capture — produces collective catastrophe, not because anyone wills the catastrophe but because no one is tasked with preventing it.

The common world was always a construction. It was always imperfect, always partial, always contested. But it existed. Its existence was the precondition for every form of collective life that depends on shared meaning. Citton's work is, at its heart, an argument for understanding what the common world requires, so that those who build the next media environment might build it with the ecological literacy that the builders of the current one so conspicuously lacked.

Chapter 7: Productive Addiction and the Colonization of Inner Silence

There is a particular quality of silence that precedes genuine thought. Not the silence of a quiet room — external silence can be purchased with noise-canceling headphones. The silence in question is internal: the absence of stimulation in the mind's own theater, the moment when no notification beckons, no option presents itself, no voice — human or artificial — offers a suggestion, a completion, a next step. This silence is uncomfortable. It is boring. It is the cognitive equivalent of hunger — a signal that the mind is empty and needs filling. And like hunger, it serves a biological function. Internal silence is the condition under which the mind's autonomous processes — association, integration, incubation, the slow consolidation of fragmentary insights into coherent understanding — do their most important work. To eliminate this silence is to eliminate the conditions under which a particular and irreplaceable kind of cognition occurs.

Yves Citton's ecology of attention identifies this inner silence not as an absence but as a habitat — the specific environmental condition that certain modes of attention require in order to exist. Floating attention, the diffuse and receptive mode from which creative insight characteristically emerges, cannot operate in a stimulation-saturated environment. It requires gaps. Pauses. The unstructured time that productivity culture labels as waste and that Citton's framework identifies as the fallow field in which the soil of future thought regenerates. To colonize this silence — to fill every gap with AI-generated content, every pause with algorithmically surfaced stimulation, every moment of potential boredom with a suggestion or a completion — is to destroy the habitat of a cognitive mode that took hundreds of thousands of years to evolve and that no amount of evaluative processing can replace.

The colonization of inner silence is not a metaphor. It describes a specific and measurable transformation in the attentional environment of the early twenty-first century, accelerated dramatically by the introduction of AI creative tools. The smartphone began the process by ensuring that no moment of waiting, no interval between tasks, no pause in conversation need ever be unfilled. Social media continued it by transforming every idle moment into an opportunity for consumption. AI creative tools complete it by extending the colonization into the last remaining refuge of inner silence: the creative process itself.

This is the context in which productive addiction — the compulsive use of AI tools that generates genuinely valuable output — must be understood. The term captures an experience that millions of knowledge workers and creators have encountered since the mainstreaming of large language models: the discovery that working with AI is not merely useful but compelling in a way that resists cessation. Not because the work is trivial — the opposite. The work is often the most interesting, most productive, most consequential work the person has ever done. The AI tool does not distract from meaningful work. It makes meaningful work so continuously available, so endlessly extensible, so reliably rewarding that the creator cannot find a natural stopping point. The addiction is not to entertainment but to productivity. And this is precisely what makes it so difficult to recognize as a problem.

Citton's framework reframes productive addiction as an ecological crisis. The individual creator's experience is one of abundance — more ideas, more output, more creative possibility than ever before. But the ecological perspective reveals what the individual experience conceals: the abundance is purchased at the cost of attentional diversity. The creator working compulsively with AI tools is operating in a single mode — the evaluative, the productive, the goal-directed — for twelve, fourteen, sixteen hours a day. Every other mode of attention is crowded out. The floating attention that requires idleness has no idleness to inhabit. The joint attention that requires presence with others loses its time slots to yet another productive session. The contemplative attention that requires stillness is overwhelmed by the perpetual motion of generation and evaluation. The creator is more productive than ever. The creator's attentional ecology is a monoculture.

The mechanism of colonization is worth examining in detail, because it operates through a feedback loop that Citton's framework illuminates with particular clarity. The AI tool generates output. The output requires evaluation — a mode of attention that is cognitively engaging, metabolically rewarding, and measurably productive. The evaluation produces refined output, which suggests new directions, which the AI tool explores, generating more output requiring more evaluation. Each cycle takes minutes. Each cycle produces visible results. Each cycle activates the brain's reward circuits in ways that idle contemplation does not. The loop is self-reinforcing: the more the creator evaluates, the more material there is to evaluate, and the more productive the session feels, and the harder it becomes to step away from a process that is generating genuine value with every passing minute.

The genius of the loop — its terrible, structural genius — is that it never presents itself as a problem. Every other form of compulsive technology use eventually produces a moment of self-recognition: the social media user who looks up from the scroll and realizes that two hours have disappeared, the gamer who notices the sun has risen. Productive addiction produces no such moment, because the time is not wasted. The output is real. The value is genuine. The creator who works with AI for sixteen hours and produces a draft, a design, a prototype, a strategy that would previously have taken weeks has not lost time. The creator has gained time. The colonization of inner silence proceeds not against the creator's will but with the creator's enthusiastic consent, because the colonizing force is not distraction but productivity — the very thing the creator values most.

Citton's ecological vocabulary provides the diagnostic tool that the language of individual choice cannot. Choice-based frameworks ask: "Is the creator choosing freely?" And the answer is yes — the creator is choosing, at every moment, to continue working because the work is valuable. Ecological frameworks ask a different question: "What is happening to the habitat?" And the answer is that the habitat of every non-productive mode of attention is being destroyed, not by force but by occupation. The productive mode expands to fill every available hour, and the hours it fills were previously occupied by other modes — the idle, the contemplative, the social, the empty — that served functions invisible to the productivity metric but essential to the long-term health of the attentional ecology.

The parallel to industrial agriculture is exact. Industrial farming is the most productive system of food production ever devised. It feeds billions of people who would otherwise go hungry. Its output is real, its value genuine, its benefits undeniable. And it is systematically destroying the soil on which future food production depends — compacting it, depleting its organic matter, killing the microbial ecosystems that make it fertile — because the productivity metric does not measure soil health. The farmer who maximizes this year's yield at the expense of the soil's long-term fertility is making a rational choice by every available metric. The farmer is also participating in a slow-motion catastrophe visible only from the ecological perspective that the productivity metric was not designed to provide.

Citton's framework suggests that the productively addicted creator is the attentional equivalent of the industrial farmer: maximizing this year's cognitive yield while depleting the attentional soil on which future creativity depends. The depletion is not felt immediately. Creativity does not stop when its attentional preconditions erode. It continues, drawing on reserves accumulated during years of pre-AI creative practice — the years when boredom was unavoidable, when silence was a regular companion, when the mind had time to wander without purpose or direction. Those reserves are finite. They were accumulated under attentional conditions that no longer exist. When they are exhausted, the creator will discover what the farmer discovers when the topsoil is gone: that productivity without ecological sustainability is a loan taken against a future that will eventually foreclose.

The foreclosure takes a specific form that Citton's analysis helps to predict. The creator whose attentional ecology has been reduced to a productive monoculture will find that the evaluative mode, even at peak efficiency, produces diminishing returns. Not because the AI tools become less capable — they will continue to improve — but because the creator's capacity to bring genuine novelty to the evaluative process will have degraded. Evaluation is only as good as the criteria the evaluator brings. And criteria are formed not in the evaluative mode but in the other modes — the floating, the contemplative, the social, the idle — that productive addiction displaces. The creator who has not wandered has nothing new to bring to the selection process. The creator who has not dwelt in difficulty has no feel for what difficulty reveals. The creator who has not been bored has no access to the strange, unexpected connections that boredom's silence makes possible. The evaluative mode, running at full capacity without the input that other modes provide, becomes a machine for selecting among variations of what the creator already knows — an optimization engine that has lost its capacity for surprise.

This is the deeper meaning of attentional monoculture: not that the dominant mode is bad but that the absent modes are irreplaceable. The industrial farm produces enormous quantities of grain. It cannot produce a forest. The productively addicted creator produces enormous quantities of polished output. The creator cannot produce the kind of work that only arises from the attentional conditions that productive addiction has eliminated — the kind of work that surprises even its maker, that arrives from somewhere the maker did not know existed, that bears the unmistakable signature of a mind that has been somewhere no algorithm could suggest going.

Citton's prescription is not abstinence but cultivation. The ecology of attention can be repaired, but not by willpower alone and not by rejecting the tools that have colonized it. It can be repaired by the deliberate construction of environments that protect the endangered modes — environments in which silence is not a bug but a feature, in which boredom is not a problem to be solved but a condition to be sustained, in which the productive mode is bounded by temporal and spatial limits that preserve the habitat of the modes it would otherwise displace.

The difficulty is that every incentive in the contemporary attention economy points in the opposite direction. The creator who takes an afternoon to wander without purpose produces nothing measurable. The creator who spends a morning in silence generates no output. The creator who preserves hours for social presence — genuine joint attention with other humans, unmediated by screens or algorithms — sacrifices productive time that the AI tool could have made spectacularly useful. Every act of attentional cultivation is, by the productivity metric, an act of waste. And the ecology of attention will not be saved by a metric that cannot see it.

It will be saved, if it is saved at all, by creators who understand what Citton understands: that the silence is not empty. It is full of everything the productive mode cannot access. And that what grows in silence — slowly, invisibly, on no one's schedule — is the thing that makes all the productive output worth producing in the first place.

Chapter 8: Cultivating the Commons — Toward an Ecology of Attention Worth Inhabiting

In the final decades of the twentieth century, a quiet revolution occurred in the way a small number of farmers thought about soil. For a hundred years, industrial agriculture had treated soil as a medium — an inert substrate whose function was to hold roots and absorb chemicals. Productivity was measured in bushels per acre. Soil health was not measured at all, because no one had conceived of soil as something that could be healthy or sick. It was dirt. You put seeds in it, added fertilizer, and harvested the result. The revolution began when a handful of soil scientists demonstrated that dirt was not inert but alive — a complex ecosystem of bacteria, fungi, nematodes, and organic matter whose health determined, over time, the productivity that farmers assumed was determined by their inputs. The farmer who understood soil as an ecosystem farmed differently. Not less productively, but more sustainably — and often, over the long term, more productively as well, because the yields that industrial methods had extracted were loans against soil capital that healthy practices could renew.

Yves Citton's ecology of attention proposes an equivalent revolution for the information age. The attentional commons — the shared resource of collective human focus — is not an inert medium through which content passes. It is a living system whose health determines, over time, the quality of everything that grows in it: thought, culture, democracy, creativity, mutual understanding. Treat it as inert — as a mere conduit for content, a surface to be farmed for engagement — and it degrades, slowly and invisibly, until the cognitive yields that seemed guaranteed begin to fail. Understand it as an ecology — a complex system of interdependent attentional modes, each requiring specific conditions, each serving functions that the others cannot perform — and entirely new possibilities for cultivation emerge.

The preceding chapters have traced the degradation. Chapters one through four described the flooding of the attentional commons by AI-generated content, the collapse of shared referents, and the temporal fragmentation of collective focus. Chapter five analyzed the displacement of generative attention by evaluative attention in creative practice. Chapter six examined the dissolution of joint attention and the consequent evaporation of the common world. Chapter seven documented the colonization of inner silence through productive addiction. Each chapter described a different dimension of the same underlying process: the transformation of a diverse attentional ecology into a monoculture optimized for individual capture and productivity, at the cost of the collective, contemplative, and creative modes that democratic and cultural life require.

The question that remains is whether the degradation is reversible — and if so, what cultivation would look like in an environment saturated with AI tools whose default operation accelerates the very monoculture that cultivation must resist.

Citton's framework insists that the answer is not prohibition. The ecology of attention will not be restored by banning AI tools any more than a degraded prairie will be restored by banning tractors. The tools are part of the environment now. They will remain. They will improve. They will become more deeply integrated into every domain of cognitive and creative work. Any program of attentional cultivation that begins with "stop using the tools" has already failed, because it mistakes the symptom for the cause. The cause is not the tools themselves but the absence of ecological consciousness in their design and deployment — the failure to understand that every media technology shapes the attentional ecology, and that the shaping can be deliberate, informed, and oriented toward the health of the commons rather than merely toward the efficiency of individual capture.

Cultivation begins with diagnosis, and Citton provides the diagnostic vocabulary. The first question to ask of any AI tool, any platform design, any creative workflow is: what mode of attention does this cultivate? Not merely: does it capture attention? Not merely: does it increase productivity? But: what kind of attending does it foster, and what kind does it displace? A large language model used for brainstorming cultivates a different mode of attention than the same model used for first-draft generation. The brainstorming use invites exploratory, divergent, playful attending — a mode adjacent to the floating attention that creativity requires. The first-draft use invites evaluative, comparative, curatorial attending — a mode that is productive but generatively impoverished. The tool is the same. The attentional ecology it produces depends entirely on how, when, and under what conditions it is used.

This suggests that the basic unit of attentional cultivation is not the tool but the practice — the specific pattern of interaction between mind, tool, and environment that shapes what mode of attending emerges. Citton's framework supports a taxonomy of practices organized by the attentional mode they sustain.

Practices that cultivate generative attention share a common structure: they create conditions for open-ended exploration without predetermined outcomes. Freewriting without AI assistance. Visual sketching before generating images. Problem formulation before solution search. The practice of beginning a creative session with twenty minutes of unaided work — writing, drawing, coding, thinking — before engaging the AI tool establishes the generative mode as the session's foundation and positions the evaluative mode as a supplement rather than a replacement. The time cost is minimal. The ecological benefit is substantial, because it exercises the generative pathways and maintains the mind's familiarity with the productive discomfort of working without assistance.

Practices that cultivate floating attention share a different structure: they create conditions for unfocused, receptive, non-goal-directed awareness. Walking without earbuds. Sitting without screens. Commuting without podcasts. These are not productivity practices. They produce nothing measurable. Their value is entirely ecological — they maintain the habitat of a cognitive mode that cannot survive in the absence of emptiness. Citton's framework reframes these practices not as leisure or self-care but as infrastructure maintenance: the upkeep of the attentional conditions on which creative and cognitive health depend. The creator who walks for thirty minutes without stimulation is not wasting time. The creator is maintaining the attentional soil in which tomorrow's ideas will germinate.

Practices that cultivate joint attention require a more radical departure from the individualized workflows that AI tools encourage. Joint attention requires co-presence — the experience of attending to the same object at the same time as another mind, with mutual awareness of the shared attending. Reading groups, collaborative workshops, shared viewing experiences, face-to-face conversations about common texts or problems — these practices construct the intersubjective space that joint attention requires. They are inefficient by every individual metric. A person reading a book alone reads faster than a person discussing the same book with others. A person generating ideas with an AI tool generates more ideas than a person brainstorming with colleagues. But the solo activities produce individual output. The joint activities produce the common world — the shared referents, the mutual understanding, the collectively held meanings that make individual output communicable and meaningful.

Citton's framework suggests that the most ecologically valuable form of joint attention in the AI age may be the deliberate practice of attending to the same AI-generated content together — reading the same AI-generated text, examining the same AI-generated images, evaluating the same AI-produced proposals, not individually and asynchronously but collectively and in real time. This practice transforms AI output from a solvent of joint attention into an object of joint attention. The content itself may be synthetic. The attending is genuine. And it is the attending, not the content, that constructs the common world.

At the institutional level, attentional cultivation requires what Citton calls "attentional institutions" — organizations, platforms, and social structures designed to protect and promote the modes of attention that the market economy neglects. Public libraries are attentional institutions: they create environments optimized for deep attention, freely accessible, insulated from commercial pressure. Public broadcasting is an attentional institution: it produces content designed to convene collective attention around common objects of concern, without the algorithmic personalization that fragments joint attention. Universities, at their best, are attentional institutions: they create temporal and spatial environments — classrooms, seminars, studios, libraries — in which sustained, shared, generative attention can occur under conditions insulated from the productivity pressures that dominate the rest of the economy.

The AI age does not eliminate the need for attentional institutions. It makes the need urgent. As the commercial media environment becomes increasingly optimized for individual capture and evaluative processing, the attentional institutions that sustain alternative modes become more ecologically essential — the last habitats in which endangered modes of attention can survive. Defunding public libraries, commercializing public broadcasting, and restructuring universities around productivity metrics are, in Citton's framework, acts of ecological destruction as consequential as draining wetlands or clear-cutting forests. They eliminate the institutional habitats that sustain the attentional diversity on which democratic and creative life depend.

The design of AI tools themselves represents perhaps the most consequential frontier of attentional cultivation. Citton's framework implies that the design of a creative AI tool is an ecological intervention — every interface choice, every default setting, every interaction pattern shapes the attentional ecology of every person who uses it. A text editor that displays AI suggestions continuously, in real time, as the writer types, cultivates a specific mode of attending: one of constant evaluation, perpetual comparison, and the habitual subordination of the writer's own generative process to the machine's suggestions. A text editor that displays AI suggestions only on explicit request, after the writer has completed a passage, cultivates a different mode: one in which generative attention is protected by the interface design itself, and evaluative attention is engaged deliberately, at the writer's chosen moment.

The difference seems trivial. It is not. Citton's ecological framework reveals that the aggregate effect of millions of such design choices, compounded across billions of interactions, across years of habitual use, determines the shape of the attentional ecology in which an entire civilization thinks, creates, and governs itself. The interface is not a neutral frame around the content. The interface is the environment. And the environment, as Citton insists throughout his work, is not a backdrop to attention but its primary determinant.

There is a deeper question beneath the practical prescriptions, and Citton's framework does not flinch from it: what kind of attention is worth cultivating? The attention economy's implicit answer is: the kind that produces measurable engagement. The productivity culture's implicit answer is: the kind that generates output. Citton's answer is different and more demanding: the kind that sustains the capacity for collective meaning.

Collective meaning — the shared understanding that allows a society to interpret its experience, deliberate about its future, and recognize its members as participants in a common project — is the product of an attentional ecology, not of any individual mind. It emerges from the interaction of multiple modes of attention, sustained over time, in environments that support their coexistence. Deep attention produces the analytical clarity that meaning requires. Floating attention produces the creative insight that renews it. Joint attention produces the shared reference that makes it communicable. Collective attention produces the cultural coherence that makes it stable. Remove any of these modes, and the meaning does not disappear immediately. It thins. It fragments. It becomes private rather than shared, individual rather than collective, optimized rather than genuine.

The ecology of attention worth inhabiting is one in which all these modes coexist — not in pristine balance, which never existed and never will, but in sufficient diversity that the degradation of one mode does not cascade into the destruction of the others. Citton's vision is not utopian. It does not require the elimination of the attention economy, the banning of AI tools, or the return to some pre-digital attentional Eden that never existed. It requires something simultaneously more modest and more radical: the recognition that attention is an ecology, that ecologies require cultivation, and that cultivation requires the willingness to protect what the market cannot value and preserve what the productivity metric cannot see.

The soil revolution in agriculture did not end industrial farming. It created a parallel tradition — a way of farming that understood the soil as a living system and treated its health as a non-negotiable precondition for sustainable productivity. The attentional revolution that Citton's framework implies would create a parallel tradition in the information age — a way of building, designing, creating, and governing that understands attention as a living system and treats its ecological health as a non-negotiable precondition for everything that attention makes possible.

The tools for this revolution exist. The AI systems now flooding the attentional commons could, with different design principles, become instruments of attentional cultivation rather than extraction. The platforms that currently optimize for individual capture could, with different metrics, optimize for collective coherence. The creative workflows that currently colonize inner silence could, with different practices, preserve the emptiness from which genuine novelty grows. The possibility is real. The technology is sufficient. What is lacking is the ecological consciousness that would make these choices visible — the recognition that the attentional commons is a commons, that it is being degraded, and that its degradation threatens not just productivity or creativity or democracy in isolation but the capacity for shared meaning on which all three depend.

Citton's work provides this consciousness. Whether it arrives in time is a question the ecology itself will answer. The flood is rising. The commons is thinning. The silence is filling with the hum of a billion individually optimized suggestions. And somewhere beneath the surface, in the deep soil of collective attention where the roots of shared meaning grow, the question remains: will anyone notice what is being lost before the loss becomes irreversible? The ecology of attention is patient. It does not collapse on schedule. It degrades gradually, invisibly, accommodating each new extraction until the day it cannot accommodate one more — and by then, the knowledge of what it was, and what it sustained, and why it mattered, has itself become one of the things the degradation has consumed.

The farmer who understands soil knows that the most important crop is the one you never harvest: the cover crop, the nitrogen-fixer, the green manure plowed back into the earth to feed next year's growth. The builder who understands attention will know something similar: that the most important thought is the one you never publish, the most important silence is the one you never fill, and the most important act of cultivation is the one that produces nothing visible — because the thing it produces is the capacity to produce everything else.

Chapter 9: The Cultivation of Creative Attention

In 2012, the artist David Hockney was asked why he spent hours each morning drawing the same Yorkshire landscape on an iPad, producing images that looked, to the untrained eye, like competent but unremarkable sketches of trees. His answer was precise: "I'm not making pictures. I'm learning to see." The distinction matters. A picture is a product — an object that enters the world and competes for attention. Learning to see is a practice — a cultivation of the attentional capacity that makes genuine pictures possible. Hockney understood something that the attention economy's architects have systematically forgotten: that the value of creative practice lies not primarily in what it produces but in what it develops. The drawing trains the eye. The eye, once trained, sees what untrained eyes cannot. And what it sees becomes the raw material of work that no amount of technical facility can fake.

Yves Citton's ecology of attention provides the theoretical framework for Hockney's intuition. If attention is an ecology — a system of interrelated modes that must be cultivated rather than merely captured — then creative practice is, at its deepest level, a form of attentional cultivation. The painter who spends years learning to see color does not merely acquire a skill. She develops an attentional mode — a way of attending to the visual world that transforms what the world reveals. The writer who spends decades learning to listen to the rhythms of sentences does not merely master a craft. He cultivates an attentional sensitivity to language that produces perceptions unavailable to the uncultivated ear. The musician who practices scales for ten thousand hours does not merely build finger dexterity. She builds an ecology of auditory attention — a capacity for hearing harmonic relationships, rhythmic subtleties, timbral nuances — that constitutes a fundamentally different way of being in the sonic world.

This understanding of creative practice as attentional cultivation has profound implications for the integration of AI into creative work. The implications are not the ones most commonly discussed. The standard debate — whether AI helps or hinders creativity, whether AI-generated work is authentic or derivative, whether AI will replace artists or empower them — operates at the level of production. It asks what AI does to creative output. Citton's framework redirects the question to a deeper level: what does AI do to creative attention? Not what does it produce, but what does it cultivate? Not what objects does it generate, but what modes of attending does it develop or degrade?

The answer is ecologically complex. AI tools can, under certain conditions, enrich the ecology of creative attention. A large language model that offers unexpected associations can stimulate the floating attention from which creative insight emerges. An image generator that produces surprising visual juxtapositions can provoke the perceptual recalibration that genuine seeing requires. A music generation system that creates harmonic combinations outside the artist's habitual vocabulary can expand the auditory attention's range. In each case, the AI operates as what Citton might call an "attentional irritant" — a source of productive surprise that prevents the attentional ecology from settling into comfortable patterns.

But these enriching conditions are not the conditions that AI tools are typically designed to produce. The dominant design paradigm for creative AI is not irritation but facilitation — not surprise but satisfaction, not expansion but efficiency. The AI is designed to understand what the creator wants and to deliver it as quickly and smoothly as possible. The prompt is the desire. The output is the fulfillment. The friction between desire and realization — the friction that constitutes creative practice — is precisely what the tool is engineered to eliminate.

Citton's ecological analysis reveals why this elimination is so dangerous. The friction of creative practice is not an obstacle to creative attention. It is the medium through which creative attention develops. The painter who struggles with a color that will not behave as expected is not wasting time. She is building perceptual capacity — learning to see the color's actual properties rather than its imagined ones. The writer who fights with a sentence that refuses to say what he means is not failing. He is developing linguistic attention — learning to hear what words actually do rather than what he assumed they would do. The musician who practices a passage until her fingers find what her ear already hears is not merely drilling technique. She is building the attentional bridge between auditory imagination and physical realization that makes musical expression possible.

Remove the friction, and the capacity does not develop. This is not a metaphorical claim. It is an ecological one. In the same way that a muscle that is never stressed does not strengthen, and a plant that is never exposed to wind does not develop the root structure to withstand storms, an attentional capacity that is never challenged does not deepen. The AI that eliminates the struggle between intention and realization does not liberate the creator from unnecessary suffering. It deprives the creator of the environmental stress that attentional development requires.

The evidence for this claim is emerging across creative domains, though it remains largely anecdotal because the transformation is too recent for longitudinal studies. Architects report that junior designers who begin their careers with AI visualization tools produce competent renderings more quickly than any previous generation but demonstrate weaker spatial reasoning — the capacity to imagine three-dimensional structures from two-dimensional representations that constitutes the architect's distinctive mode of attention. Software developers report that programmers trained with AI coding assistants write functional code faster but understand systemic architecture less deeply — the capacity to hold an entire codebase in mind that constitutes the engineer's distinctive mode of attention. Music producers report that artists working with AI composition tools generate polished tracks more efficiently but develop weaker harmonic intuition — the capacity to hear where a chord progression wants to go that constitutes the musician's distinctive mode of attention.

In each case, the pattern is the same: the AI accelerates production while decelerating cultivation. The output improves. The attention degrades. The creator becomes more productive and less perceptive — a combination that is, in Citton's ecological terms, the precise signature of an environment being strip-mined rather than sustainably farmed.

The question, then, is whether creative attention can be cultivated in AI-augmented environments, or whether the integration of AI into creative practice inevitably degrades the attentional ecology that genuine creativity requires. Citton's framework suggests that the answer depends not on the AI itself but on how the environment around the AI is designed. The tool is not the ecology. The tool is one element within the ecology. The ecology includes the practices, institutions, rhythms, and social structures that shape how the tool is used — and how, crucially, it is not used.

Consider the difference between two hypothetical creative environments. In the first, a writer works with an AI assistant that is always available, always responsive, and always ready to generate options. The writer's workflow is continuous: write, generate, evaluate, select, revise, generate again. Every moment of hesitation is an opportunity to consult the AI. Every moment of uncertainty is an opportunity to request alternatives. The attentional mode this environment cultivates is evaluative — the rapid scanning and comparing that selecting among options requires. The attentional mode it suppresses is generative — the slow, uncertain, apparently unproductive dwelling from which genuine novelty emerges. The writer becomes an excellent editor and a diminished creator.

In the second environment, the same writer works with the same AI assistant but within a practice structure that includes deliberate periods of AI absence — what might be called "attentional fallow." The writer begins each day with an hour of writing without any AI access: pen on paper, or a blank screen with no connection to any generative tool. During this hour, the writer experiences the full friction of creative practice — the blank page, the false starts, the sentences that do not work, the ideas that refuse to cohere. The writer struggles. The writer fails. The writer's attention, deprived of the evaluative option, is forced into generative mode — forced to produce rather than select, to dwell rather than scan, to create rather than curate. After this hour, the writer turns to the AI and works in collaborative mode — generating, evaluating, refining. But the generative capacity has been exercised. The attentional ecology includes both modes. The soil has been allowed to rest.

Citton's framework suggests that the second environment is not merely preferable but ecologically necessary — that the cultivation of creative attention in an AI-saturated world requires the deliberate construction of environments that protect generative attention from the evaluative mode's encroachment. This is not a rejection of AI. It is a recognition that every ecology requires diversity, that every mode of attention requires specific environmental conditions, and that those conditions must be consciously maintained rather than left to the market's optimization logic. The market will always optimize for the measurable — and generative attention, by its nature, resists measurement. Its value is visible only in retrospect, in the work that emerges from periods that looked, at the time, like unproductive wandering.

The institutional implications of this insight are significant. If creative attention must be cultivated rather than merely captured, then the institutions that shape creative practice — universities, studios, companies, funding bodies, platforms — bear ecological responsibility. They are not merely distributing resources. They are shaping attentional environments. A university that integrates AI into every aspect of creative education without preserving AI-free spaces for generative attention is not merely adopting a new tool. It is restructuring its attentional ecology in ways that may be efficient in the short term and devastating in the long term. A company that measures creative productivity by output volume without attending to the attentional diversity of its creative workers is not merely optimizing. It is strip-mining.

Citton's concept of "attentional commons" becomes particularly urgent here. Individual creative attention exists within a collective ecology. A writer's capacity for generative attention depends not only on her personal practices but on the shared culture of creativity within which she works — the conversations, the mutual provocations, the joint attention to common objects of creative concern that constitute a creative community. When the entire community shifts to AI-augmented production, the collective attentional ecology shifts with it. The conversations change. The shared references change. The mode of joint attention — what creative communities attend to together and how they attend to it — changes. Even the creator who maintains personal practices of attentional cultivation finds herself working within a collective ecology that may no longer support the modes of attention her work requires.

This collective dimension points toward Citton's deepest concern: the relationship between creative attention and democratic life. Citton argues, consistently and with considerable evidence, that the modes of attention required for genuine creative work are the same modes required for genuine democratic participation. Both require the capacity to dwell with ambiguity rather than resolving it prematurely. Both require the capacity to attend to perspectives that challenge rather than confirm existing beliefs. Both require the patience to engage with complexity that cannot be reduced to simple choices. Both require, fundamentally, the capacity for joint attention — the shared focus on common objects that makes collective deliberation possible.

If AI systems degrade these attentional modes in creative practice, they degrade them in democratic practice as well. The creator who loses the capacity for generative attention loses something that matters far beyond the creative domain. She loses a mode of being in the world — a way of attending that is essential not only to making art but to making meaning, not only to producing culture but to sustaining the collective life that culture serves.

The cultivation of creative attention is, therefore, not a luxury concern — not a matter of artistic temperament or aesthetic preference. It is an ecological imperative. In an age when AI can generate any content at any scale, the scarce resource is not production but perception — not the capacity to make things but the capacity to notice things, to attend to things, to dwell with things long enough for their genuine nature to reveal itself. This capacity must be cultivated. And cultivation, as any farmer knows, requires not only planting and watering but also the discipline to leave fields fallow, to resist the temptation to plant every square meter, to trust that the apparently empty ground is doing work that the eye cannot see.

Citton's ecology of attention, applied to the AI-saturated creative landscape, arrives at a conclusion that is simultaneously hopeful and demanding. The ecology can be sustained. The modes of attention that creativity and democracy require can be preserved. But preservation requires design — the conscious construction of environments, practices, and institutions that protect attentional diversity from the monoculture machine. It requires builders who understand that they are not merely building tools but shaping the environments in which attention lives. And it requires a willingness to value what cannot be measured — the slow, the quiet, the apparently empty — because the most important things happening in the ecology of attention are happening in the fallow spaces, in the silences between stimuli, in the unproductive hours that produce nothing visible and everything that matters.

Chapter 10: Building the Ecology We Need

On December 9, 1968, Douglas Engelbart stood before an audience of approximately one thousand computer scientists in San Francisco and performed what has since become known as "The Mother of All Demos." In ninety minutes, he demonstrated the computer mouse, hypertext, video conferencing, collaborative document editing, and a graphical user interface — essentially the entire conceptual architecture of personal computing, thirty years before most of these technologies reached the public. But the most radical aspect of Engelbart's demonstration was not any single technology. It was his framing. He was not, he explained, building tools to make individuals more productive. He was building environments to augment collective intelligence — to create the conditions under which groups of human minds could think together more powerfully than any individual mind could think alone.

Yves Citton's ecology of attention finds in Engelbart a kindred spirit across disciplines. Both understand that the relevant unit of analysis is not the individual mind but the environment within which minds attend, communicate, and create. Both understand that tools do not merely amplify existing capacities but reshape the ecology within which those capacities develop. And both understand that the critical question about any technology is not "What can it do?" but "What kind of collective attention does it cultivate?"

This question — the ecological question — is the one that the current wave of AI development has largely failed to ask. The dominant narrative around generative AI is individualist: what can this tool do for me? How can it make me more productive, more creative, more efficient? The tools themselves are designed around individual prompts, individual outputs, individual workflows. The user sits alone before a screen, types a request, receives a response, evaluates it, refines it, and iterates. The entire interaction is structured as a dialogue between one mind and one machine. The ecology of collective attention — the shared, the joint, the communal — is architecturally absent.

Citton's framework suggests that this architectural absence is not a minor design oversight but a fundamental ecological failure. If attention is a commons — a shared resource whose health depends on collective practices of cultivation and maintenance — then tools that systematically individualize attention systematically degrade the commons. Each person's attention may be individually enriched by their AI interactions. Each person may produce more, learn more, create more. But the shared attentional space — the space in which communities attend to common objects, coordinate their focus, and build shared meaning — is impoverished by the same process. The individualization of attention through AI is the attentional equivalent of privatizing a public park: each individual plot may be well-tended, but the commons — the shared ground that made the neighborhood a community — disappears.

The evidence for this ecological degradation is already visible in creative communities. Consider what has happened to the literary conversation — the shared attentional practice through which readers, writers, critics, and publishers attend together to new works and collectively determine their cultural significance. Before AI, the volume of published work was constrained by the human capacity for writing, editing, and production. This constraint, however painful for individual writers seeking publication, served an ecological function: it limited the number of objects competing for collective attention to a volume that communities could meaningfully process. Readers could, at least in principle, be aware of the major works published in their area of interest. Critics could, at least in principle, survey the significant new contributions. The literary conversation was imperfect, exclusionary, and often unjust. But it was a conversation — a practice of joint attention through which a community attended to common objects together.

The flood of AI-generated and AI-assisted content has not destroyed this conversation through any single dramatic act. It has dissolved it through dilution. When the volume of published material exceeds any community's capacity to jointly attend to it, the conversation fragments. Each reader, guided by algorithmic recommendation, follows a personalized path through the flood. No two readers encounter the same landscape of new work. The shared references that made conversation possible — "Have you read...?" — become less likely with each order-of-magnitude increase in available content. The joint attention that sustained the literary community as a community gives way to individualized consumption. Everyone reads more. No one reads the same thing. The ecology of collective literary attention degrades not because anyone stopped caring about literature but because the environment in which collective caring was practiced has been flooded beyond its carrying capacity.

Citton's analysis suggests that the solution is not to reduce the flood — that particular genie is irrevocably out of the bottle — but to build new structures of collective attention within it. This is an architectural challenge, not a behavioral one. Telling individuals to read less AI-generated content is as ecologically useful as telling individual drivers to produce less carbon. The problem is structural. The solution must be structural. The question is: what kinds of attentional architectures can sustain collective attention in an environment of infinite individual content?

Several principles emerge from Citton's ecological framework. The first is the principle of attentional curation as collective practice. If the flood of content exceeds any individual's capacity for evaluation, then the filtering function — the determination of what deserves collective attention — must be performed collectively. This is not a new idea. It is, in fact, the original function of cultural institutions: museums, journals, festivals, universities, and publishing houses all served, historically, as institutions of collective attentional curation. They did not merely distribute content. They performed the ecological function of focusing collective attention — of saying, on behalf of a community, "This deserves your shared regard." AI threatens these institutions not by replacing their curatorial function but by overwhelming it. The solution is not to abandon curation but to reinvent it at a scale commensurate with the flood.

The second principle is the protection of joint attention through deliberate design. Joint attention — the shared focus of multiple minds on a common object — does not occur naturally in a personalized media environment. It must be architecturally supported. This means designing platforms, spaces, and practices that create common attentional ground: shared texts that communities read together, shared problems that communities work on together, shared creative projects that communities build together. The technology for this exists. What is lacking is the design intention. The current generation of AI tools is designed to personalize — to give each user a unique experience optimized for their individual preferences. An ecologically healthy AI environment would also depersonalize — would create moments of shared experience, shared surprise, shared encounter with objects that were not individually optimized but collectively offered.

The third principle is the institutional protection of attentional fallow. If generative attention requires periods of emptiness — if creativity depends on the mind's capacity to wander unstructured through the dark — then institutions that shape creative practice must protect emptiness from the optimization logic that seeks to fill every gap with productive stimulation. This means designing workplaces, educational environments, and creative platforms that include deliberate periods of AI absence: times when the tool is unavailable, when the generative capacity is offline, when the creator must work from internal resources rather than external augmentation. The resistance to this principle will be enormous. In an economy that measures productivity by output and values efficiency above all, the deliberate introduction of unproductive time looks like waste. Citton's framework reveals it as investment — the investment in attentional diversity that an ecology requires to remain healthy over time.

The fourth principle — and perhaps the most challenging — is the recognition that attentional ecology requires governance. A commons that is not governed is a commons that is destroyed. Garrett Hardin's "Tragedy of the Commons" described the dynamic precisely: when a shared resource is available for individual exploitation without collective regulation, each individual's rational self-interest leads to the resource's depletion. The attentional commons faces the same dynamic. Each individual's rational choice to use AI for maximum personal productivity contributes to the collective degradation of the attentional environment in which all individuals operate. The solution, as Elinor Ostrom demonstrated in her Nobel Prize-winning work on commons governance, is not privatization (which destroys the commons) or top-down regulation (which often fails to account for local conditions) but community-based governance — the development of shared norms, practices, and institutions through which communities manage their common resources sustainably.

Applied to the ecology of attention, this means that the communities most affected by AI's attentional impact — creative communities, educational communities, journalistic communities, democratic publics — must develop their own practices of attentional governance. These practices will look different in different communities. A literary community might establish shared reading practices — collective attention events organized around common texts — that counteract the individualizing tendency of algorithmic recommendation. An educational community might develop curricula that include explicit attentional cultivation — teaching students not only how to use AI tools but how to sustain the modes of attention that AI tools tend to degrade. A democratic community might create institutions of collective attention — public forums, shared media, common information environments — that counteract the fragmentation of personalized content.

None of this is easy. The economic incentives all point in the opposite direction — toward personalization, toward individual optimization, toward the extraction of attention rather than its cultivation. But Citton's ecological framework insists that the choice between extraction and cultivation is not merely a matter of preference. It is a matter of survival. An ecology that is only extracted from eventually collapses. A commons that is only taken from eventually empties. The attentional commons is no exception.

The builders of AI systems — the engineers, designers, entrepreneurs, and investors who are shaping the tools that will shape the attentional ecology of the coming decades — face a choice that is simultaneously technical and moral. They can build tools that maximize individual engagement, that personalize every experience, that fill every moment with optimized content, and that treat attention as a resource to be extracted. This path is profitable and ecologically catastrophic. Or they can build tools that also cultivate collective attention, that create spaces for joint focus, that protect generative emptiness from evaluative noise, and that treat the attentional commons as a shared resource to be sustained. This path is harder, less immediately profitable, and ecologically necessary.

Citton's ecology of attention offers no guarantees. It offers no algorithm for sustaining the commons, no formula for cultivating the modes of attention that creativity and democracy require. What it offers is something more fundamental: a way of seeing. It offers the recognition that attention is not a private resource inside individual skulls but a shared ecology shaped by collective practice. It offers the understanding that media environments do not merely deliver content but cultivate modes of attending that determine what kind of society we are capable of being. And it offers the uncomfortable but essential insight that the most powerful technologies are not those that do the most but those that shape how we attend — because how we attend determines what we notice, and what we notice determines what we know, and what we know determines what we build, and what we build determines the ecology within which the next generation will learn to attend.

The ecology of attention is being reshaped by artificial intelligence more rapidly and more profoundly than by any technology since the printing press. The question is not whether this reshaping will occur. It is already occurring. The question is whether it will be shaped by conscious design or left to market forces — whether the attentional ecology of the coming century will be cultivated or strip-mined, diversified or monocultured, sustained or depleted. Citton's framework suggests that the answer depends on whether we recognize attention for what it is: not a commodity to be traded, not a metric to be optimized, not a resource to be extracted, but a commons to be cultivated — the most precious commons we possess, the ground from which meaning itself grows, and the ecology within which everything we value about being human either flourishes or dies.

The flood is here. The commons is at stake. The question is what we will build.