[YOU] on AI describes the discovery that the exhilaration of AI-assisted creation eventually gives way to something that looks like compulsion: the loop of prompt and response that is impossible to exit because the exit requires tolerating an emptiness that the tool has been trained to fill. Attentional fallow names what is being eliminated in that loop—not dead time but the precondition of living thought. The cycle’s insistence on the signal that the AI cannot generate because it has no stake in the world converges with Citton’s ecological insight: the genuinely novel question, the original insight, the idea that could not have been retrieved because it had not yet been formed, requires the fallow that the AI-saturated environment systematically eliminates.
The concept connects to Byung-Chul Han’s diagnosis of a culture that has lost its relationship to negativity—to rest, to boredom, to the apparently unproductive pauses that are actually the conditions of genuine productivity. Both thinkers are describing the same elimination from different angles: Han from the phenomenology of the burnout subject, Citton from the ecology of the attentional commons. The deep attention that sustained work requires, and the floating attention that creative surprise requires, share the same need: a space that the optimization engine is not allowed to fill.
The concept develops from Citton’s broader ecological framework in Pour une écologie de l’attention (2014) and from his analysis of what he calls “attentional cultivation”—the set of practices and environmental designs that protect attentional diversity against the monoculture pressures of the optimization economy. The agricultural analogy is not decorative: Citton intends the structural parallel seriously. Industrial agriculture’s great failure was treating yield as the only variable and treating fallow as waste; the attentional economy’s great failure is treating engagement as the only metric and treating emptiness as an inefficiency to be eliminated.
The concept gains specific urgency from the design logic of AI creative tools. A large language model is always available, always responsive, always proposing a completion before the sentence is finished. This affordance eliminates the precise environmental condition—the gap between impulse and expression, the moment of not-yet-knowing—in which the generative mode of attention operates. Citton’s framework predicts the result: the generative mode gives way to the evaluative mode, the creative wetland is drained and developed, and the fallow that would have regenerated the soil of genuine novelty is replaced by another round of optimized selection from what already exists.
Emptiness as medium, not absence. Attentional fallow is not the absence of thought but the presence of a different kind of thinking: diffuse, associative, below the threshold of conscious effort, making connections between domains that focused goal-directed attention cannot reach. The shower thought is not a lucky accident; it is the systematic output of an attentional mode that only operates when the focused mode is off duty. Eliminating the fallow does not merely remove waste; it removes the generative process that focused attention cannot perform.
The design problem. Attentional fallow cannot be protected by individual willpower alone, because the tools that eliminate it are integrated into the creative environment itself—embedded in the text editor, the design application, the development environment, offering suggestions before they are requested. Citton’s ecological analysis implies that protecting attentional fallow requires environmental design: working practices, platform configurations, and institutional norms that build emptiness into the structure of the creative day rather than relying on individuals to resist a tool that is designed to make its own use frictionless.
The regenerative cycle. Just as agricultural fallow feeds subsequent harvests by rebuilding what intensive cropping depletes, attentional fallow feeds subsequent focused work by replenishing the associative material that deep analysis requires. A creative practice structured only around focused production, with no protected fallow, will eventually exhaust its own generative capacity—producing more of what it has already produced and less of what it has not yet imagined. The efficiency gain of eliminating fallow is real in the short term and self-defeating in the medium term: the very output it was optimizing for becomes less original, less surprising, less valuable precisely as the fallow that produced it is eliminated.

The attentional-fallow concept faces a pragmatic objection: the specific creative processes it describes—shower thoughts, dream insights, wandering associations—are themselves empirically contested as sources of genuine novelty rather than merely comfortable feelings of insight. Cognitive scientists have challenged romantic accounts of incubation, noting that many apparent “creative breakthroughs” during downtime can be explained by continued unconscious processing that does not require phenomenological emptiness. Against this, Citton’s defenders note that the attentional-fallow argument does not depend on a specific theory of the neural mechanism; it depends only on the ecological observation that certain kinds of thought are incompatible with continuous goal-directed stimulation, and that the AI-saturated environment is continuously goal-directed by design. Whether the mechanism is unconscious incubation, associative wandering, or something else entirely, the ecological precondition is the same: a gap in the stimulus stream that the optimization engine is structurally motivated to eliminate.