The Pixel Before Time: The Story of AARON and the World’s First AI-Generated Image

Long before deep-learning models could render photorealistic portraits from a simple text prompt, a lone British painter sat in a room at the Stanford Artificial Intelligence Laboratory watching a machine carefully drag a pen across a sheet of paper. It was 1973, and art was about to change forever.

While today’s generative artificial intelligence tools rely on massive neural networks trained on billions of scraped photographs, the birth of AI-generated imagery happened in an entirely different era. There were no GPUs, no cloud servers, and no text boxes. Instead, there was a revolutionary computer program named AARON and its creator, artist turned programmer Harold Cohen.

Meet AARON: The Pioneer of Algorithmic Creativity

In the late 1960s, Harold Cohen was a successful, internationally acclaimed painter who had represented Great Britain at the prestigious Venice Biennale. Yet, he grew deeply fascinated by a radical question: What is the bare minimum required to make an image recognizable?

To answer this, Cohen abandoned his thriving conventional art career, moved to the United States, and taught himself computer science. He began developing AARON at the University of California, San Diego, eventually refining the software during a pivotal residency at Stanford in the early 1970s.

Unlike modern AI, which learns statistically by analyzing existing visual data, AARON was built on Good Old-Fashioned AI (GOFAI)—symbolic logic and explicit rules. Cohen didn’t train AARON on a dataset of pictures; rather, he painstakingly hand-coded rules about drawing, perspective, occlusion, and human anatomy.

“AARON was not a tool used to execute a human’s pre-planned drawing. It was an autonomous entity programmed to make its own artistic choices regarding composition, form, and line.”

The First Images: Lines That Breathed Life

The very first images produced by AARON in 1973 were humble, abstract black-and-white line drawings. Rendered using a specialized mechanical drawing machine connected to a computer, these early pieces featured primitive geometric shapes, organic curves, and rudimentary figures.

To a modern observer, they might look like simple vector sketches. But in 1973, their significance was monumental. For the first time in human history, a machine was executing a physical drawing completely autonomously, making real-time decisions about where to place a line without direct human intervention guiding the pen.

Evolution Over Decades

As computing power grew, so did AARON. Over the next several decades, Cohen continuously updated his creation:

• The 1980s and 90s: AARON learned how to handle color, eventually filling in its own shapes and branching out into complex botanical and human figures.
• The 2000s: The software evolved to project massive, vibrant jungle-like scenes onto gallery walls, shifting from paper-bound pen drawings to immersive digital environments.

When Cohen passed away in 2016, AARON’s journey froze in time, leaving behind a legacy that bridges the gap between traditional fine art and computational automation. Major cultural institutions, including the Whitney Museum of American Art, continue to celebrate AARON’s work as a cornerstone of digital art history.

Why the First AI Image Still Matters

Looking back from today’s era of instant-generated media, AARON reminds us that the quest to make machines creative is not a recent Silicon Valley trend. It is a decades-long philosophical experiment.