🫴Artists You Need to Know: Harold Cohen Harold Cohen (1928–2016) is widely recognized as a pioneer of AI art. Starting in the late 1960s, he built AARON, a system that draws and paints autonomously, using rules that mimic how an artist thinks: from lines to shapes to figures. For Cohen, the computer wasn’t just a tool; it was a collaborator with its own kind of agency. With his solo show, "AARON at Tsukuba," at our gallery in London, here's a look at his journey ↓

Before “AI art” had a name, Cohen pivoted from a successful painting career to code. He began building AARON, a rule-based program that could compose images from scratch without using source images.

At Stanford’s AI Lab (1973–75) he formalized AARON’s logic and introduced it publicly through plotter drawings. Viewers watched the machine “think,” pausing mid-stroke as compositions emerged organically rather than as geometric patterns.

One breakthrough was turning AARON’s drawings into color paintings. Cohen engineered custom plotters and painting robots so the system could lay down dyes and paints, translating software decisions into physical brushwork, no hand “touch-ups.”

Tsukuba Expo ’85, Japan became a landmark moment: inside the U.S. Pavilion, AARON drew live for months, producing thousands of unique works that visitors took home, cementing the project’s public impact and scale.

Another key moment was AARON KCAT (2001), a figurative phase distributed with Kurzweil CyberArt. Plants, people, and interiors unfolded algorithmically at public scale, marking AARON’s shift from abstract structures to a richer symbolic world.

Cohen also wrote influential texts like “Parallel to Perception” (1973), arguing that computers can parallel mental functions, so a program’s “autonomy” isn’t illusory. The idea underpinned decades of exhibitions where AARON performed live drawing.

A favorite example of his dry clarity: he resisted calling AARON “creative” unless it later did something it wasn’t explicitly programmed to do, setting a high bar that still challenges today’s model-driven art.

Across his work, the thread is consistent: perception, representation, and the line between human intention and machine procedure. He steered rules into that in-between space where we project meaning onto emerging forms.

Taken together, Cohen’s practice says creativity can be specified: encoded, executed, performed, AND still feel human. His legacy is the blueprint many contemporary AI artists are iterating on now.

"AARON at Tsukuba" in our private space in Notting Hill, London. 🔗 For more information, visit