What Is the First AI Generated Image Ever Made?

The Pixel Before Time: The Fascinating True Story of the First AI-Generated Image

We live in an era where typing a quirky sentence into a chat box yields a hyper-realistic masterpiece in seconds. From photorealistic portraits of astronauts riding capybaras to intricate cinematic landscapes, text-to-image generators have transformed everyone into an instant “artist.”

Yet, as we marvel at these modern tools, a compelling question often gets overlooked: When, and how, was the very first AI-generated image created?

To find the genesis of artificial intelligence art, we have to rewind past the era of neural networks, deep learning, and cloud computing—all the way back to the early 1970s.

Meet AARON: The World’s First AI Artist

Long before GPUs trained on billions of parameters, a British-born painter and professor named Harold Cohen asked a radical question: What makes a drawing a drawing?

In 1968, while teaching at the University of California, San Diego, Cohen became fascinated by the intersection of computing and human cognition. Rather than using a computer as a simple tool to edit photos or layout shapes, Cohen wanted the computer to make decisions.

By 1973, working heavily during a residency at the Stanford Artificial Intelligence Laboratory, Cohen birthed a software program called AARON. AARON is widely recognized as the world’s first true artificial intelligence art system.

How AARON Made Art Without Neural Networks

If you are imagining a computer spitting out abstract oil paintings from a text prompt, think again. AARON didn’t use machine learning, neural networks, or massive datasets scraped from the internet. Instead, it relied on Good Old-Fashioned AI (GOFAI)—a system of strict symbolic rules written in the C programming language and Lisp.

Cohen essentially took his own decades of knowledge as a human painter and meticulously translated it into code. He didn’t teach AARON what a specific tree looked like by showing it millions of pictures; instead, he programmed the rules of visual representation:

• How objects overlap in space.
• How to balance proportions and composition.
• How to construct basic geometric and organic shapes, eventually expanding into abstract human figures, rocks, and foliage.

The early outputs of AARON were simple, black-and-white line drawings etched out by a custom-built robotic drawing machine (a mechanical arm holding a pen, tethered to a mainframe computer).

A Paradigm Shift: When AARON ran its code, Cohen wouldn’t intervene in the drawing process. The program decided where to place lines, how to construct shapes, and how to compose the final frame autonomously. In essence, the machine was executing creative choices based on logic provided by its human creator.

From Pens to Palettes

As computing power grew over the decades, so did AARON.

• The 1980s & 90s: Cohen updated AARON to understand color theory. The software would generate the line drawings, and Cohen would meticulously hand-paint them based on the computer’s instructions. Later versions eventually learned to color autonomously using computer-controlled paint brushes and mixing stations.
• The 2000s: AARON evolved to project massive, complex jungle scenes and abstract human-plant hybrids onto gallery walls.

Cohen continued to code and evolve AARON until his passing in 2016. Today, AARON’s pioneering works reside in prestigious institutions like the Tate Modern and the Whitney Museum of American Art, standing as monuments to the dawn of computational creativity.

The Legacy of the First AI Image

Comparing AARON’s minimalist 1973 line drawings to today’s generative titans (like Midjourney or Stable Diffusion) highlights a staggering technological leap. Yet, the philosophical DNA remains identical.

AARON proved that art and code are not mutually exclusive. It challenged the art world’s definition of authorship decades before internet memes and copyright lawsuits over AI training data even existed.