AI + visual arts

More than a prompt. More than an output.

Generative systems can be treated as creative material: unpredictable, culturally loaded, open to transformation, and shaped by the choices an artist makes before and after an image appears.

A human process

The meaningful work is in the decisions.

AI can propose patterns at astonishing speed. It cannot decide what deserves our attention, what a work means, or what responsibility an image carries into the world.

Those are human choices. An artist frames the inquiry, selects and rejects, combines materials, introduces context, edits the result, and determines whether the work should exist at all.

This makes generative art less interesting as a shortcut and more interesting as a site of negotiation between intent, system, source material, chance, and interpretation.

A co-creation loop

Iteration is where authorship becomes visible.

A useful process makes each creative decision legible. The goal is not to find a magic phrase, but to build a chain of intentional choices.

1

Frame

Begin with an idea, memory, observation, or formal constraint that matters to you.

2

Generate

Create variations and notice both what the system offers and what it repeatedly misses.

3

Select

Use taste and intention to reject, compare, sequence, crop, and redirect.

4

Transform

Draw, collage, print, code, annotate, or combine outputs with other physical and digital materials.

Look beneath the image

Every output has a context.

Creative literacy includes understanding that generative systems are built from data, shaped by design decisions, and constrained by what their makers chose to optimize.

DATA
Whose work may have informed the system, and was permission part of that process?
BIAS
Which bodies, cultures, histories, and visual conventions does the system center or distort?
CLAIM
How will the role of the tool and the nature of the process be described honestly?
IMPACT
Could the image mislead, appropriate, expose, or cause harm beyond the artist’s intention?

Ways to explore

Move the image into the world.

A generative image can be a beginning rather than an endpoint. Mixing processes makes room for material knowledge, embodied decisions, and creative surprise.

Material translation

Digital to physical

Translate an output through printmaking, textiles, paint, sculpture, projection, or hand-built collage.

Rule-making

Constraint as voice

Build a visual system from a small palette, a local archive, a repeated gesture, or a deliberately narrow prompt grammar.

Critical remix

Interrogate the pattern

Collect recurring assumptions in a set of outputs, then annotate, disrupt, or reframe what the system treats as normal.

Studio prompt: Generate ten versions of an ordinary place you know well. Mark every detail the system gets wrong. Use those mistakes—not the polished images—as the material for a new work.

Continue exploring

What changes when the canvas can listen?

Explore AI + music