The History and Origin of the First AI Generated Image

The Dawn of Digital Creativity: The Story of AARON and the First AI-Generated Images

Today, we live in a world where typing a few words into a text box can instantly summon hyper-realistic portraits, sprawling fantasy landscapes, and intricate digital paintings. But long before modern neural networks and internet-scale datasets transformed text prompts into pixels, a lone artist and programmer was quietly laying the groundwork for artificial intelligence art.

His name was Harold Cohen, and his creation, AARON, gave the world its first true glimpse of machine-made art.

From the Canvas to Code

In the late 1960s, Harold Cohen was already an established, highly successful British painter. He had represented Great Britain at prestigious international showcases like the Venice Biennale. Yet, driven by a deep intellectual curiosity about what fundamentally makes an image evocative, he made a radical pivot. He moved to the University of California, San Diego, taught himself computer programming, and began asking a question that decades ahead of its time: Could a machine learn to make art?

By 1973, Cohen had developed the initial iterations of AARON—not as a piece of hardware, but as a groundbreaking software program.

How the First AI “Thought”

AARON operated very differently from today’s AI image generators. Modern tools rely on massive machine learning models trained on billions of existing images scraped from the web. AARON, however, had no internet, no database of photographs, and no neural network.

Instead, Cohen manually coded rules of art, cognition, and geometry directly into the program:

• Cognitive Rules: AARON was taught the cognitive primitives of how humans perceive the world—such as how objects overlap, how figures stand, and how lines intersect.
• Autonomous Decision-Making: Rather than just tracing a pre-programmed path, AARON was given a set of rules that allowed it to make original structural choices on the fly. It decided its own compositions, balancing shapes and lines dynamically.

The earliest outputs were minimalist, abstract, black-and-white line drawings that echoed primitive art or organic forms. Attached to a specialized robotic drawing machine (a mechanical arm holding a pen), the computer would autonomously guide the pen across a physical sheet of paper, bringing its internal logic into the physical world.

A Lifelong Partnership

As the decades progressed, Cohen continuously evolved AARON. What started as basic line drawings eventually expanded into complex figures, plants, and jungle-like scenes. In later iterations, Cohen even taught the program rules of color theory, allowing AARON to select and mix its own pigments or paint independently.

The work of AARON shifted perceptions worldwide, earning exhibitions at legendary venues like the Tate Modern and the Whitney Museum of American Art. When Cohen passed away in 2016, it brought to a close a nearly fifty-year collaboration between a human artist and his software creation.

A Legacy Etched in Code

Looking back, AARON was a monumental milestone. It proved that creativity was not strictly an exclusive biological trait, and that a machine could follow algorithmic logic to produce evocative, original visual art.