From Abundance to Attention

Generative Creativity as Cultivation

Short public version

Personal note

This essay comes from lived experience rather than academic distance. Working with generative AI has forced me to ask what it means to be an artist when images, sounds and variations can be produced in extraordinary abundance. The question has not been abstract. It has involved authorship, ownership, skill, attention, responsibility and value as part of my own working life.

My starting point is simple: generative systems do not remove the artist, but they change where artistic responsibility gathers. When generation becomes abundant, discernment becomes central.

1. The condition of abundance

Generative image systems have altered the conditions under which visual work can now be made. They do not simply offer another tool in a long history of tools. They introduce a different creative condition: abundance.

Images can now arrive in numbers that would once have been unimaginable. Variations appear quickly. Styles, textures, atmospheres, structures and accidents can be summoned, tested, discarded and recombined at speed. The artist no longer faces only the old scarcity of means, materials, time and technical execution. A new pressure appears: the pressure of excess.

This does not make the artist unnecessary. It changes where the artist’s responsibility gathers.

When generation becomes abundant, discernment becomes central.

The question is no longer only: can I make an image? The question increasingly becomes: among all these possible images, where is the life? What deserves attention? What should be carried forward, and what should be allowed to disappear?

2. Between panic and celebration

Much of the current conversation around generative systems and art seems to divide into two unsatisfactory positions.

On one side is the techno-utopian position. Here, the emphasis falls on speed, automation, fidelity and spectacle. The excitement is often directed toward the fact that something can be produced at all: a highly detailed image in seconds, a plausible style, a convincing surface, a technical leap. In this view, execution is easily mistaken for meaning.

On the other side is the defensive institutional position. Here, generative systems are treated primarily as threats: to labour, originality, copyright, authorship, the hand, the studio, the market and the aura of the object. These concerns are real and cannot be dismissed. But when anxiety becomes the whole frame, it can prevent us from seeing what is actually changing in the structure of artistic practice.

The more interesting territory lies between these positions. Generative abundance should neither be celebrated blindly nor rejected defensively. It should be examined as a new ecology of making.

The artist now works in a field where images may be plentiful, but attention is not. This is the shift that matters.

3. The myth of the clean prompt

One persistent misunderstanding of generative practice is the myth of the clean prompt: the idea that an artist writes a sentence, presses return, and receives a finished work. This may sometimes happen as a novelty, but it is not the reality of a sustained artistic practice.

The prompt is not the artwork. It is one condition among many.

A generative process may involve the setting of initial conditions, the testing of language, the adjustment of visual direction, the rejection of hundreds of seductive but false results, the recognition of unexpected accidents, the returning to earlier branches, the careful comparison of near-duplicates, the cutting away of excess, the building of a series, and the acceptance of consequence.

The process is often iterative, unstable and fatiguing. It is less like issuing a command to a machine than working within a weather system. The artist does not control every cloud, but must learn to recognise the moment when form, pressure, light and atmosphere align.

This is why the word curation is not enough. Curation suggests arranging things that already exist. Cultivation begins earlier. It includes seeding, tending, pruning, waiting, recognising and accepting. The artist enters the mud before the flower appears.

4. Cultivation, not automation

To describe generative creativity as cultivation is to insist that the human artist remains responsible for the field in which the work comes into being.

Cultivation does not mean complete control. A gardener does not manufacture the life of a plant. The gardener prepares conditions, attends to growth, cuts back, supports, redirects, protects and chooses. The result is not mechanical authorship, but neither is it passive reception.

This distinction matters. If generative systems are described only as automation, the human artist seems to disappear. If they are described only as tools, the magnitude of the shift is softened. Cultivation offers a more accurate middle term. It allows for agency without pretending to mastery.

The artist working with generative abundance may be setting conditions, guiding tendencies, recognising emergence, selecting from excess, sequencing bodies of work, and deciding when an image has enough necessity to stand. These are not secondary tasks. They are the new centre of the work.

5. Accident and acceptance

Artists have always worked with accident. Paint bleeds. Chemicals misbehave. Light leaks. A plate scratches. A camera catches something unintended. A brushstroke opens a possibility the artist did not plan.

What matters is not whether the artist personally manufactured every accident. What matters is whether the artist recognised it, accepted it, shaped the work around it, and took responsibility for its presence.

The artist may not have made every accident, but they are responsible for every acceptance.

This is especially important in generative practice. The machine may produce unforeseen forms, textures, errors, mutations or conjunctions. But the machine does not know what should remain. It does not know which accident carries force, which is merely decorative, which belongs to the work, and which should be refused.

Acceptance is not a weak gesture. In a condition of abundance, acceptance becomes one of the hardest artistic acts. To accept one image is also to reject thousands. To choose one path is to close many others. The artist’s responsibility is not reduced by generative abundance. It is intensified.

6. The psychological weight of glut

The art world often underestimates the psychological reality of abundance. For centuries, one of the artist’s central struggles was scarcity: lack of materials, lack of time, lack of technical means, lack of access, lack of money, lack of visibility.

Generative systems do not remove struggle. They relocate it.

When a thousand viable images can appear, the problem is not simply production. The problem becomes judgement. The artist must face the fatigue of possibility. There may be too many attractive surfaces, too many near-solutions, too many images that are impressive but empty, too many results that seem alive for a moment and then collapse under sustained attention.

This abundance can be paralysing. It can flatten judgement. It can create a false sense that everything matters equally, or that nothing matters enough. The artist must resist both tendencies.

To move from abundance towards attention is to accept that artistic labour has become more perceptual, editorial and ethical. The work lies not only in making more, but in knowing what not to keep.

7. Recognition

Recognition is difficult to describe, but central to the process. At some point, among many possible images, one image looks back. It has a charge, an internal tension, a sense of necessity. It may not yet be finished, but it has life.

This recognition is not purely intellectual. It involves the eye, the body, memory, taste, fatigue, desire, suspicion and restraint. It is not simply liking. Many images can be liked. Far fewer can be trusted.

The question becomes: does this image continue to hold attention after the first shock? Does it deepen or merely impress? Does it belong to a larger body of work? Does it open space, or does it close too quickly? Does it ask to be absorbed?

Recognition is where the artist’s history enters. Years of looking, making, failing, absorbing influence, resisting influence, teaching, arranging, discarding and beginning again are brought to bear on the instant of selection. The image may have been generated quickly, but recognition is not quick. It is carried by a lifetime of attention.

8. Selection, cutting, sequencing

Selection is often misread as a passive act, as if choosing from generated images were no more than picking a favourite from a menu. But in serious practice, selection is active, consequential and often severe.

To select is to assert that this, and not that, carries the work forward. It is to accept responsibility for the whole: subject, form, colour, tension, rhythm, atmosphere, coherence, relation to other works, and the pressure of what is left unsaid.

Cutting is equally important. Generative abundance tempts the artist to keep too much. A series can become swollen. An image can become overfilled. A promising direction can become merely productive. The artist must cut away not only failures, but also successes that do not belong.

Sequencing then turns individual works into a field of attention. One image may be strong alone, but a body of work asks a different question: how do the works speak to each other? Where is the repetition meaningful? Where does variation deepen the series? Where does the sequence breathe?

In this sense, generative practice moves from creation, to curation, to cultivation. The final work is not simply an image. It is an ecology of decisions.

9. The artist as companion to attention

A useful way to understand this role is not the artist as teacher, and not the artist as technician, but the artist as companion to attention.

The artist cannot control what a viewer will see, feel or understand. The artist cannot guarantee significance. But the artist can create conditions in which attention may slow, gather and deepen.

This is especially important with generative images, because the viewer may arrive with suspicion, fatigue or prejudice. The image may be dismissed too quickly because it is assumed to be effortless. The artist’s task is therefore not only to generate and select, but to frame the encounter in such a way that the work can be met.

The viewer can never fully know the artist’s intention. But the viewer can be accompanied toward their own version of significance. Without that accompaniment, the image risks remaining only a sense object: attractive, strange, decorative, impressive, but not absorbed.

10. From looking to seeing: a practice of attention

From abundance to attention is also a movement from looking to seeing, and from seeing to insight. This movement is not automatic. Abundance does not demand attention; more often, it overwhelms it. When images multiply beyond the viewer’s capacity to receive them, they cease to arrive as distinct encounters and become noise. The viewer learns to scan, classify and discard because there is too much to stay with.

Abundance can also coarsen attention. In a saturated visual field, only the most extreme images may seem able to cut through: the brightest colour, the sharpest contrast, the most disturbing subject, the most exaggerated gesture. Culture begins to reward visual over-seasoning. Like a palate dulled by too much sweetness, salt or heat, the eye can lose sensitivity to quieter forms of meaning.

This is why attention becomes necessary. It is not simply a preference for slowness, but a corrective practice. Attention slows looking down long enough for recognition, relation and meaning to appear. It allows quieter things to become visible again: tonal shifts, uncertain identities, fragile balances, delayed recognitions, and meanings that do not reveal themselves at first glance.

Seeing is looking long enough for connections to appear. It is not a fixed sequence but a circulation between different kinds of attention: impact, the whole, detail, recognition, relation, equivalence, context, return and insight. Like tasting wine, it can be trained. A sommelier does not simply drink and decide; they look, smell, taste, hold, separate, compare and return.

Frederick Franck’s The Zen of Seeing: Seeing/Drawing as Meditation is one older echo of this concern with seeing as a discipline. The present argument belongs to a different condition — one of generative abundance and visual glut — but it shares the belief that seeing is not passive looking. Seeing is a practice that can open into insight.

The visual world itself has changed. Earlier forms of attention were often shaped by scarcity, distance and effort: a religious icon, painting or ceremonial object might require a journey before it could be seen. Contemporary visual life is dense with competing signals: signs, adverts, traffic, shopfronts, screens, reflections, brands, warnings and movement. The digital image intensifies this further. It refreshes, interrupts, multiplies and disappears. We scroll more, but see less.

Attention often begins by separating something from the glut around it. A frame, plinth, wall, light, screen, sequence, home, or gallery can isolate a work from competing signals. These are not neutral devices. They create conditions in which the viewer can focus, compare, return and see. At their best, galleries matter because they protect this possibility. So too can the chosen work in the home: selected from the surrounding noise and given a continuing place in one’s field of attention. Living with a work allows attention to accumulate, though it can also become invisible through familiarity. Seeing has to be renewed, even with the things we have chosen.

Looking closes down quickly because the mind is built to label. From infancy, the visual world is gradually separated from abstraction into recognisable forms. Two dark points and a curved line become a face; the face becomes mother; recognition brings delight. The world becomes knowable. But the same system can settle things too quickly. A cup is a cup, a face is a face, an abstract image is “abstract,” a digital image is “AI,” and the encounter closes.

Art reopens this process. It does not abolish the label; it loosens it. A cup placed on the head becomes, absurdly but plausibly, a hat. The point is not that the cup “really” is a hat, but that seeing can release an object from a single label and restore possibility. The artist and the viewer both need some version of this capacity: not childishness, but the ability to look with knowledge while not closing things down too soon.

Seeing is also shaped by what the viewer brings to the encounter. An art historian, artist, critic, curator or experienced viewer brings memory, comparison, categories, precedents and prior encounters. This contextual understanding allows seeing to deepen, just as a sommelier’s distinctions make tasting more articulated. But knowledge also carries a danger. Expertise can harden into habit. A viewer may see only what they already know how to name. The strongest seeing holds two things together: enough experience to recognise richness, and enough openness to be surprised.

The modes of seeing are not strict steps. They may overlap, repeat, reverse, or begin wherever the work first catches the viewer. A small detail may pull the viewer in before the whole has been understood; a colour may create the first impact; an equivalence may appear before the subject has been named. What matters is not the order, but that these different kinds of attention eventually come into play.

10.1. Impact

Before analysis begins, there is impact. A work first arrives as a whole: large or small, light or dark, dense or open, calm or agitated, seductive or resistant, inviting or threatening. This first encounter is the body and mind registering the presence of the work. Abundance tends to push images toward impact because impact is what survives the first scan. In a crowded field, the instantly legible image has an advantage. But when impact becomes the whole experience, the work remains at the level of the first hit. A practice of seeing does not reject impact. It refuses to stop there.

10.2. The whole

Stand back, physically or mentally, and allow the whole work to arrive as a field before asking what it is “of.” Is it light or dark, dense or open, calm or agitated, balanced or unstable? Narrowing or half-closing the eyes can reduce detail and reveal the broad tonal structure: the deepest darks, brightest lights, and main weights of the composition. This fuzzy seeing is one way of slowing after impact.

10.3. The eye’s wander

Let the eye travel without fixing too soon. Notice where it is pulled, where it hesitates, what it passes over, and what keeps calling it back. This is a first mapping of the work’s internal forces: weight, rhythm, emphasis, interruption or silence.

10.4. Recognition

Recognition begins when the mind starts to name what it is encountering. Is the work figurative, abstract, symbolic, bodily, architectural, mechanical, tender, comic, violent, uncertain? Recognition may be clear, or it may hover. Sometimes the richness of a work lies in resisting the label for long enough that seeing can deepen.

10.5. Parts and detail

At some point the viewer moves into the parts: figures, marks, blocks, surfaces, textures, colours, shadows, intervals, breaks and repetitions. Detail does not only mean small detail. A dominant colour field, heavy dark mass, blank space or repeated form can be as important as a small mark.

10.6. Relationships

This is where looking begins to become seeing. The question is no longer only, “What is there?” but “How do these parts behave together?” Does one colour answer another? Does a shape repeat elsewhere? Does an empty space press against a dense one? Meaning begins in relation.

10.7. Equivalence

Equivalence is the moment when one thing starts to behave like something else. A shape may not literally be a face, body, animal, landscape, wound, mask or machine, but it may contain enough visual clues to suggest one. The mind searches, labels and compares. Artists often act as visual disruptors: they give just enough information for recognition to begin, then alter or complicate it. Art does not simply abandon recognition; it unsettles it.

10.8. Meaning

Meaning does not have to mean a literal story. In abstract work as much as figurative work, meaning may arise as the story of forces: balance and imbalance, emergence and disappearance, order and collapse, concealment and exposure. The viewer begins to ask: what is this work showing me? What is it asking me to notice? What richness is being offered? This is where recognition deepens into insight.

10.9. Return

Finally, return. Step back, look again, or come back later. The whole may now look different because the parts have been seen; the parts may behave differently because the whole has been felt. Seeing is not linear. It is circular. Looking consumes the image and moves on. Seeing stays long enough for recognition to deepen.

Attention slows looking into seeing. Seeing deepens recognition into insight.

11. The gallery frame and its contradiction

The gallery is a machine for slowing attention. White walls, spacing, scale, lighting, silence, captions, distance and ritual all say: stop here, look longer, allow this object to matter.

This is one of the ironies of the current art world. The gallery already possesses the very mechanism that generative images most need: a frame that can slow looking. Yet digitally native work is often denied that frame unless it has first been translated into a form the market recognises more easily as scarce, ownable, insurable and collectible.

This does not mean that physical manifestation is false. A print, projection, installation, book, object or large-scale presentation can deepen attention through scale, surface, duration, placement and bodily encounter. Physical form can make the work more present.

But physical translation can also become scarcity theatre: a way of making digital work appear rarer, slower or more ownable than it truly is.

The distinction is crucial. Physical manifestation is valid when it deepens encounter. It becomes scarcity theatre when it merely reassures ownership.

12. The native digital work-state

A digitally generated or digitally cultivated image should not be treated as a sketch waiting to become real. The digital file may already be the native form of the work.

This does not make print irrelevant. Nor does it mean that screen display is always sufficient. It means that physical translation should not be required as proof of artistic legitimacy.

The question should not be: how can this digital work be made to look like traditional art? The question should be: what form of encounter allows this work to receive the attention it needs?

Sometimes that may be a print. Sometimes projection. Sometimes a book. Sometimes a screen. Sometimes a sequence. Sometimes an installation. Sometimes the work may remain most itself as a luminous digital image, encountered in the medium from which it emerged.

The digital file is not immaterial in any simple sense. It has scale, resolution, colour, structure, relation, reproducibility and conditions of display. It is not a non-object. It is a different kind of object, or perhaps a different kind of work-state.

13. Scarcity, labour and value

The art market has inherited systems of value built around scarcity, labour, objecthood and ownership. Generative abundance unsettles these systems.

If an artwork can be produced as a file, multiplied without degradation, displayed on a screen, printed at different scales, altered, recombined or reissued, then familiar markers of value become unstable. The market therefore often seeks reassurance: the unique object, the visible hand, the labour-intensive translation, the monumental fabrication, the signed physical surface.

There is nothing inherently wrong with any of these. The problem arises when they become substitutes for attention. A work should not need to mimic scarcity in order to be taken seriously.

Traditional art history has often romanticised visible labour: the hours spent over a surface, the difficulty of craft, the trace of the hand. These remain powerful and meaningful. But generative practice asks us to recognise other forms of labour: perceptual labour, editorial labour, ethical labour, the labour of restraint, the labour of refusing abundance.

The difficulty of a work may no longer be located where the viewer expects it.

14. Ethics and extraction

Any serious account of generative creativity must acknowledge ethical concerns. Training data, consent, attribution, labour displacement, cultural extraction and monetisation cannot be waved away by appealing to artistic freedom. These questions matter.

But ethics should not be reduced to a single refusal of the technology. Generative systems are already part of the visual culture in which artists, audiences and institutions now live. The more useful question is how artists can work with transparency, restraint and responsibility inside a changed field.

This includes acknowledging the conditions of production, avoiding false claims of handmade origin, resisting the exploitation of living artists’ styles as shortcuts, and being honest about the role of generative systems in the work.

It also includes resisting the temptation to hide behind the machine. The artist cannot say: the system made it, therefore I am not responsible. If the artist accepts the image, presents it, sequences it, sells it, exhibits it or attaches their name to it, responsibility returns.

Abundance does not excuse the artist. It makes discernment more ethically visible.

15. Stillness after speed

Generative systems are often discussed in terms of speed. But the value of the resulting work may depend on slowness.

This is particularly true of the still image. In an age of endless motion, simulation, animation, immersion and spectacle, the still image retains a peculiar force. It asks the viewer to stop. It does not unfold automatically. It waits.

A generative still image may have emerged from speed, but it need not be consumed quickly. Its importance may lie precisely in the contradiction between rapid generation and slow absorption.

The artist’s task is to protect that slowness: to choose images that can survive attention, and to present them in ways that allow attention to gather.

16. Series as cultivation

A single generative image may be striking. But the deeper artistic question often appears in the series.

A series tests whether an image belongs to a larger logic. It asks whether the artist is merely collecting effects or cultivating a world. Repetition, variation, recurrence and difference become evidence of attention.

In a sustained generative practice, the series becomes the place where judgement can be seen. Why this image and not another? Why this colour temperature? Why this degree of density? Why this relation between structure and instability? Why this rhythm across the body of work?

The series makes the artist’s decisions visible. It transforms abundance into a path.

17. From abundance to attention

From Abundance to Attention is not a rejection of generative systems. Nor is it a defence of them as mere tools. It is an attempt to name the changed responsibility of the artist within a field of excess.

The artist working with generative systems is not simply a prompt writer, technician, curator or consumer of machine output. The artist is a cultivator of attention.

They prepare conditions. They invite accident. They recognise life. They select, cut, sequence and frame. They accept responsibility for what remains.

The work begins in abundance, but it does not end there. Without attention, abundance becomes noise. With attention, abundance can become form.

The artist may begin with many images, but the work is the discipline of arriving at this one.

*a fuller working document is available separately.