Protobyte essays

Essay 01

Protobyte Phylogeny: a practice coming together

For decades, I’ve been working toward a closer relationship between painting, drawing, and computation.

The connections have been there for a long time. What is changing is my ability to bring them into a more expansive system: one in which form, variation, inheritance, and an organism’s history can develop together.

AI has become an integrating force in that work.

I came to programming as a painter. Code offered another material to explore, another way to discover something through the act of making it. Trigonometry became an expressive paintbrush. A few lines of code, a change of parameter, a little randomness: something unexpected would appear.

I began the Protobyte inquiry in 2003. The early creatures grew from code, mathematics, and variation. They gave me a way to approach computational form with an artist’s curiosity about what a process might reveal. The surviving studies from 2004 already suggest anatomy and unfamiliar kinds of life. Protobyte chronology, Bryn Mawr's early-work record.

That way of working also entered my teaching and books, beginning with Processing: Creative Coding and Computational Art in 2007. Learning to program could open a direct relationship with the material. An artist could change the rules that produced an image, experiment with those rules, and discover forms through them. Publisher's record.

My work also extended into a custom C++ Protobyte graphics library. C++ library account.

Later, Glomularus extended the Protobyte work through browser-based three-dimensional organisms. Generative editions created another way for the work to circulate beyond my studio. My experiments with generative AI opened further connections to drawing and painting. BEASTS, presented in 2023, brought code, AI, and hand drawing into the same exhibition. Artist chronology, BEASTS exhibition record.

There were other paths into questions of life and participation. In The Dorises, my collaboration with Marina Zurkow, computational oysters gathered into a collective reef. Made for Chain Reaction on Feral File, the project connected generative software and AI imagery with an invitation to support oyster conservation. These projects have their own identities, but they belong to the larger context from which the current work is developing. Project record.

Throughout these experiments, I continued imagining a more integrated artificial-life system. I wanted to work with populations and histories as well as individual forms: creatures whose variations could be inherited, whose relationships could accumulate, and whose lives could continue beyond a particular view of them.

The engineering required to hold those ideas together was a persistent constraint.

As AI moved from code completion toward more sustained technical collaboration, my working process changed. I could explore an architecture, interrogate a problem, revise a component, and move between parts of the system with greater continuity. Ideas I had kept in separate experiments became easier to bring into relation.

That is the context for Protobyte Phylogeny.

The work extends the Protobyte inquiry into an evolving population of computational organisms. Generative code develops their bodies and patterns. Breeding and mutation produce inherited variation. An authored ecology governs energy, survival, and reproduction, while changing conditions affect which traits are favored.

AI participates within that ecology, too. A vision model assesses rendered organisms against particular criteria; those assessments contribute to selection alongside rules expressed in code. A language-model naturalist writes records from information about the organisms and their histories. The geometry remains a product of the generative system.

This makes the relationship between making and judging part of the work. I choose the developmental rules and the conditions under which variation matters. The models participate in assessment and description. The exhibition is being developed around visitors contributing organisms whose subsequent lives and descendants unfold in the shared world.

The word phylogeny places ancestry at the center of the project. An organism has a form, but it can also have parents, descendants, a sequence of events, and a recorded death. Those relationships give the work a duration beyond any single image.

Related studio projects are exploring how those forms might be encountered differently: as detailed specimen portraits, offline renders, paper interpretations, voxel creatures in Protokraft, and installations through ProtoTanks. Each direction introduces its own material decisions. Together, they give me further ways to investigate the relationship between an organism, its history, and its representation.

What excites me is the possibility of working across these relationships with greater continuity. Drawing, coding, observing, selecting, and revising can become parts of an ongoing practice whose results continue to change.

I’m grateful to be developing Protobyte Phylogeny for Modal × Gray Area in San Francisco. The group exhibition is scheduled for October 28–November 1, 2026. Modal's computational infrastructure also supports the development of the world, giving the partnership a direct relationship to the material conditions of the artwork. Organizer announcement.

The work feels new and familiar at once. I recognize the pleasure that first drew me into code: changing something small and discovering a form I could not quite have anticipated.

Now I’m trying to give that form a longer life, and a history in relation to others.

Explore Protobyte Phylogeny · Modal × Gray Area exhibition · Related installation work: ProtoTanks

Essay 02

The conditions of a creature

In Protobyte Phylogeny, making a creature also means making the conditions under which it will be assessed.

The generative system gives the organism a body. A set of parameters shapes its proportions, surface, appendages, and movement. Breeding and mutation allow differences to pass into subsequent generations. But variation alone does not determine what happens to a population. The work also needs a way for some differences to matter.

That is where selection enters the piece.

The current system combines criteria expressed in code with assessments from a vision model. The model receives a rendered view of an organism and a rubric associated with an ecological niche. Its assessment contributes to the fitness value used by the ecology. Energy, aging, population capacity, and breeding rules also influence what happens next.

This arrangement gives selection an explicitly authored character. The criteria come from decisions about what the world should value. A model's reading becomes one input into those conditions. Changing the conditions can change which forms persist and reproduce.

As an artist, I’m interested in what happens when that act of judgment becomes part of a generative process. My decisions continue to act through the system even when I am no longer choosing each individual result. The resulting population can carry traces of those decisions in ways that are difficult to anticipate creature by creature.

The naturalist introduces another relationship to the same population. A language model receives information about an organism and writes a field record. That description can help us attend to a creature, but it remains a generated interpretation. The organism's stored ancestry and events provide the record against which the language can be read.

These are different ways of encountering computational life: developing form, assessing it, describing it, and preserving evidence of what happened. Bringing them together is part of the artistic experiment.

Protobyte Phylogeny is in development for the Modal × Gray Area group exhibition in San Francisco. Project, organizer.

Essay 03

A portrait of a particular life

A computational organism can produce many images. What interests me in the developing Protobyte work is the possibility of keeping those images connected to a particular life.

The distinction begins inside the system. A genome describes a generative foundation. A life record identifies an individual in the world. A capture preserves a state of that individual and the information needed to interpret it. These records have different jobs. Two lives can share a genome without becoming the same life.

That gives a specimen portrait a different relationship to its subject. It can be connected to an organism that had parents, produced descendants, or died. Its history extends beyond the moment represented by the image.

ProtoPortraits and ProtoMax are being developed around that continuity. A retained capture can supply the source for a more detailed representation. Lighting, materials, rendering, and additional artistic detail shape how the organism appears. The source record keeps those choices connected to what came from the world.

Other studio experiments make the change of material more conspicuous. ProtoGami interprets a genome through animated paper facets. Protokraft samples the generative form into a block lattice. A crease or a voxel introduces its own visual logic. Their current outputs interpret the genome without preserving the complete life record.

The larger artistic question is what carries across these changes. It might be a proportion, a pattern, a recognizable arrangement of parts, or a documented ancestry. Different representations can make different aspects of the organism available to attention.

For Protobyte Phylogeny, the proposed exhibition archive begins at the opening. The developmental populations have their own records, but they do not acquire an exhibition history retrospectively. The artistic inquiry reaches back to 2003; the particular lives in this new population will have their own beginning. Artist chronology.

That relationship between a continuing practice and a particular life is becoming one of the most compelling parts of the work for me. A form can persist through many representations while the record allows us to ask what, exactly, we are looking at.