The History of the First AI-Generated Image

Before the Prompt: The Untold Story of the First AI-Generated Image

When we think of artificial intelligence and art today, our minds instantly race to high-tech digital canvases summoned by a few typed words. We type a prompt into an interface, and seconds later, a hyper-realistic, brilliantly lit masterpiece appears.

Yet, long before neural networks, diffusion models, and billions of parameters dominated the conversation, someone had to answer a radical question: Can a machine be an artist?

To find the true “first AI-generated image,” we have to rewind not to Silicon Valley’s recent boom, but to the late 1960s and early 1970s—long before the internet as we know it even existed.

The Birth of AARON: A “Surrogate Self”

The story of the first true AI art system begins with Harold Cohen, a distinguished British painter who represented his country at the prestigious Venice Biennale. At the peak of his traditional career in 1968, Cohen made a startling pivot. He packed up his brushes, moved to the University of California, San Diego, and immersed himself in computer science.

Cohen wasn’t merely interested in using a computer as a fancy paintbrush. He wanted to understand the cognitive mechanics of human creativity itself. He wanted to build an autonomous system that could make its own artistic choices.

By the early 1970s, Cohen had birthed a software program named AARON.

AARON was unlike anything that came before it. While engineers like A. Michael Noll and Georg Nees were already using mainframe computers at places like Bell Labs to plot geometric, algorithmic patterns, AARON was fundamentally different. It was designed to possess rules of representation. It didn’t just calculate repeating grids; it understood boundaries, occlusion, perspective, and composition.

What Did the First AI Images Look Like?

The earliest images produced by AARON in the 1970s were sparse, striking black-and-white line drawings.

Running on massive, room-sized mainframe computers, AARON would output code that drove a mechanical drawing machine—a flatbed plotter with a robotic arm holding a pen. As the pen glided across the paper, it drew abstract human figures, rocks, plants, and organic shapes.

Crucially, AARON made autonomous decisions. Cohen did not draw a sketch and ask the computer to trace it, nor did he hardcode where every single line should go. Instead, he coded a deep understanding of cognitive rules: how a body fills a space, how objects overlap, and how line weight conveys depth.

When the machine drew, it improvised within those rules. Every single output was unique. In essence, AARON was making original artistic choices, rendering it the world’s first true artificial intelligence artist.

From Mechanical Arms to Neural Networks

As decades passed, Cohen continuously evolved AARON.

• In the 1980s and 1990s, AARON learned to handle color, with Cohen building complex color-mixing systems for the robotic arm.
• Later iterations began populating vibrant, jungle-like canvases that were eventually exhibited in world-class institutions like the Tate Modern and the Whitney Museum of American Art.

Cohen spent over forty years working alongside his digital creation, viewing AARON not as a tool, but as a collaborative partner—a digital “surrogate self”. When Harold Cohen passed away in 2016, AARON went quiet, but its legacy was permanently cemented into art history.

Why the First AI Image Matters Today

Looking at today’s landscape of instant text-to-image generators, it’s easy to feel like AI art sprang out of nowhere. But looking back at AARON’s primitive black-and-white lines reminds us of a vital truth: the desire to merge machine logic with human imagination is more than half a century old.

AARON proved that creativity isn’t exclusively a biological miracle; it can be unpacked, understood, and mirrored through code. The next time you watch an algorithm render a breathtaking digital painting in seconds, spare a thought for Harold Cohen and his mechanical arm, tracing out the very first steps of a revolution on a blank sheet of paper.