The Canvas of 1973: How Harold Cohen and AARON Sparked the Era of AI Art

Today, typing a prompt into an AI generator to conjure breathtaking, hyper-realistic images in seconds feels routine. Yet, the story of artificial intelligence in art didn’t begin with modern neural networks or massive GPU clusters. It began more than half a century ago, inside a university lab, driven by a pioneering painter who decided to teach a machine how to draw.

That journey culminated in 1973, marking the birth of AARON—the world’s first true artificial intelligence art system.

From the Venice Biennale to the Computer Lab

To understand the magnitude of AARON’s creation, you have to look at its creator. Harold Cohen was a celebrated British painter who had already achieved mainstream success, representing Great Britain at the prestigious Venice Biennale in 1966. At the height of his traditional career, however, Cohen became deeply fascinated by the cognitive mechanics of human drawing—specifically, how the human brain translates abstract rules of perception into physical marks on a canvas.

In the late 1960s, Cohen moved to the United States to teach at the University of California, San Diego. Immersed in the burgeoning tech culture of the era, he taught himself computer programming. Rather than using the computer as a mere digital paintbrush, Cohen set out to encode his own understanding of artistic rules directly into software, birthing a program he named AARON.

How AARON Worked: Rule-Based Ingenuity

Unlike contemporary AI models that rely on machine learning, vast datasets, and probabilistic neural networks, AARON operated on symbolic artificial intelligence.

• Hand-Coded Logic: Cohen did not feed AARON thousands of historical paintings to mimic. Instead, he manually programmed human rules regarding shape, composition, perspective, and proportion.
• Autonomous Decision-Making: Within the rigid boundaries of those coded rules, AARON was given the autonomy to make its own artistic choices, deciding where to place lines, how to balance compositions, and how to structure a scene.
• Physical Execution: In its early days, AARON didn’t live on a high-resolution computer screen. Connected to a custom-built mechanical drawing machine—often called a “turtle plotter”—the computer physically moved pens across paper to bring its digital calculations into the physical world.

The 1973 Milestone and Decades of Evolution

By 1973, AARON successfully generated its first autonomous drawings. These early pieces were primitive, black-and-white line compositions featuring abstract shapes and rudimentary human figures.

To the art and tech worlds, however, they were nothing short of revolutionary. For the first time, an autonomous machine was creating original imagery based on an internal logic system rather than direct, stroke-by-stroke human teleoperation.

Over the following decades, Cohen continuously upgraded AARON. What started as simple geometric line work eventually evolved into sophisticated software capable of rendering complex botanical structures, rich jungle scenes, and vibrant color paintings. AARON’s works were eventually exhibited at prestigious venues worldwide, including the Tate Modern and the Whitney Museum of American Art.

A Timeless Artistic Legacy

Harold Cohen passed away in 2016, leaving behind a legacy that fundamentally reframed the relationship between humanity and technology.

Long before the ethical debates, commercial applications, and viral trends of modern generative AI, AARON proved that a machine could participate meaningfully in the creative process. Looking back at the humble black-and-white lines drawn by a 1973 computer program offers a powerful reminder: today’s technological leaps stand firmly on the shoulders of decades-old curiosity and visionary artistry.