The History of the First Ever AI-Generated Image

The Brush That Never Slept: The True Story Behind the World’s First AI-Generated Image

Long before the internet was flooded with hyper-realistic photogenics, phantom extra fingers, and the frantic clicking of text-to-image prompts, a quiet revolution began in a laboratory room in the late 1960s.

Today, we take generative artificial intelligence for granted. We type a sentence into a chatbot, and seconds later, a masterpiece or a monstrosity materializes on our screen. But the lineage of machine-made art didn’t start with neural networks, billions of parameters, or vast web-scraped datasets. It started with a painter who wanted to bottle human creativity into code.

His name was Harold Cohen, and his creation was an autonomous art system named AARON.

Meet Harold Cohen and the Birth of AARON

In the late 1960s, Harold Cohen was a celebrated British painter representing his country at prestigious global showcases like the Venice Biennale. Yet, at the height of his traditional career, he found himself gripped by an existential question: What makes a drawing a drawing?

To find out, Cohen did something radical for an artist of his era. He packed his bags, moved to the United States, and taught himself computer science. He didn’t want to use a computer merely as a digital canvas, like a fancy version of Microsoft Paint. Instead, he wanted to build a software system that could make its own artistic decisions.

By 1973, Cohen’s digital assistant—which he christened AARON—produced its first historic, independent visual outputs.

How the First AI “Thought”

AARON was fundamentally different from the AI image generators we use today. Modern tools like Midjourney or DALL-E rely on deep learning, statistical probabilities, and billions of training images scraped from the web.

AARON had no internet. It had no massive dataset of historical paintings. Instead, Cohen painstakingly handcrafted a cognitive rulebook of human art-making.

• The Logic of Form: Cohen coded rules about how physical objects overlap, how proportions work, and how the human eye perceives boundaries.
• Autonomous Composition: Once the program booted up, Cohen didn’t guide the computer’s hand. AARON decided its own compositions, choosing where to place lines, how to balance negative space, and how shapes interacted on the page.
• The Mechanical Hand: In its earliest iterations, AARON couldn’t display an image on a high-definition monitor. Instead, the code commanded a physical, custom-built turtle-like drawing machine (and later, robotic plotters) holding a pen to physically ink lines onto paper.

Watching AARON work in the 1970s must have felt like peering into science fiction. A mechanical arm would glide across a blank sheet of paper, methodically drafting abstract, organic shapes that no human had pre-planned—bringing a computer program’s internal “mind” into the physical world.

Evolution of a Digital Artist

Over the next several decades, AARON grew alongside computer science.

• In the 1980s, Cohen taught AARON how to handle closed forms and basic shading.
• By the 1990s and 2000s, the system evolved to handle color mixing, complex textures, and eventually, rich, jungle-like botanical scenes populated by stylized figures.

AARON’s work broke barriers that no other software had touched. Its pieces were eventually exhibited in world-class institutions like the Tate Modern in London and the Whitney Museum of American Art in New York.

Why the First AI Image Still Matters

Harold Cohen passed away in 2016, but AARON remains a towering milestone in digital history.

When we look back at those primitive 1970s black-and-white line plots, we aren’t just looking at vintage computer code. We are looking at the philosophical genesis of machine creativity. Cohen proved that art was not some mystical, untrackable human trait immune to logic, but a process of rules, relationships, and recursive choices that a machine could learn to understand.

The next time you prompt an AI to paint a futuristic city or a surrealist portrait, spare a quiet thought for AARON—the mechanical pioneer that picked up a pen in 1973 and proved that a computer could dream in lines and shapes.