Emergence and Authorship

There has always been debate about where creative work lives.

Some locate it in the execution: the brushstroke, the final render. This stance was challenged when digital art entered the picture, as people decried the cheapening of the craft when creative decisions became non-destructive and reversible. But people realized that the deeper, more meaningful work happens in the creative process itself. Creativity is in the choices and constraints that make certain outcomes possible and others unlikely.

Consider improvisation. What makes a jazz solo work is not freedom but structure: the chord changes, the key, the tempo, the other players' contributions. These constrain the viable choices available to the improviser, and the music emerges from the interaction between intention and limitation.

This is how all good creative work functions. Painters choose their palette and composition. Writers establish voice and structure. None of them specify outcomes directly at the point of conception. They design or inherit conditions, and what emerges from those conditions is the work.

Like photography before it, AI forces us to confront the question of authorship. I believe, as most creatives do, that what machines create without human intervention is slop. If that is true, then we cannot pretend that creative work was ever in the execution alone. AI strips away the illusion that creativity lives in the hand rather than the mind directing it.

Control vs. Constraint

We then have to ask what exactly the mind contributes to the creative process.

Control attempts to realize predetermined outcomes. Constraint shapes the space of possible outcomes. These represent different relationships between creator and work.

Control says: I know what I want, and I will specify it until the system produces exactly that. This works when you can articulate precisely what you want beforehand, and when the task is execution rather than discovery. Control is the realm of automation, because it is measured in how efficiently the creator's pre-existing vision can be realized.

Constraint says: I know the properties I want the outcome to exhibit, and I will structure conditions that make such outcomes likely. This works when the desired output cannot be fully specified in advance. Crucially, it involves iteration: what emerges may prompt new choices, and works when the task involves creative evolution. Constraint is an essential element for creation.

Control is a tool-oriented construct. It is a measure of the ability for creative intent to be articulated and expressed. However, much of what is valuable about creative work operates through constraint rather than control.

The constraints a creator applies to their work shape what can appear, gives them the direction to make creative choices, and very often change as the work itself is realized.

Emergence

When constraints are imposed on a dynamic system, we allow for emergence. Put simply, emergence is when the whole is becomes more than its parts alone.

I built the simulation below through constraint.

Click Start. Watch what happens. No two runs produce identical dynamics, yet all exhibit recognizable patterns: boom-bust cycles, niche formation, near-extinction.

The simulation took dozens of iterations to create. Early versions collapsed into extinction. By adjusting constraints and observing what emerged, I found parameter combinations that produced rich, stable dynamics.

This is where the control/constraint distinction reveals something non-obvious: the best outcomes are often ones you could not have specified in advance.

It is not the brush on the canvas that is art. It is not the words printed on paper that is art. It is what emerges from that paint, from those stories -- It is what is invoked within us that we perceive as art.

Constraint creates conditions from which art can emerge.

Authorship

Traditional authorship assumes direct causal chains: artist conceives, hand executes, work manifests. AI seems to threaten authorship because it breaks this chain.

But the constraint model locates authorship differently. You author a work by shaping conditions, by recognizing value in what emerges, by iterating toward alignment with intent.

This framing does not eschew control, but it sees control as a vehicle for realizing creative decisions made at a higher level in the creative process. Control allows for the execution of creative decisions made within the constraints decided upon by the artist.

I am the author of the above simulation not because I wrote every line of code (thanks, Claude!), but because I designed the constraints and iterated toward a result that satisfied my intent.

This is a form of authorship that has always existed. The jazz musician does not author a solo by controlling every vibration of air. The photographer does not author an image by placing each photon. They shape conditions and select from what emerges.

What makes this kind of creativity demanding is that it requires holding two things at once: clear enough intent to evaluate emergence, and enough openness to let emergence change what you're trying to make. This is what distinguishes productive use of AI from both automation and slop.

The creative's work is in finding constraints that produce emergence worth encountering.