What Was the First AI-Generated Image in History?

The Genesis of Pixel Art: What Was the Very First AI-Generated Image?

When we think of AI-generated images today, our minds instantly jump to hyper-realistic photos, epic fantasy landscapes, or viral internet memes conjured up in seconds via tools like Midjourney, DALL-E, or Stable Diffusion.

However, the story of artificial intelligence making art didn’t start with neural networks or text prompts in the 2020s. It actually began more than half a century ago, rooted in the clever mind of an artist who wanted to see if a machine could learn how to draw.

Meet AARON: The World’s First AI Artist

The crown for creating the first AI-generated image belongs to Harold Cohen, a British painter and professor. In the late 1960s and early 1970s, while teaching at the University of California, San Diego, Cohen became fascinated by computing. Rather than just using a computer as a tool, he wondered if he could write a program that actually made creative decisions on its own.

In 1973, Cohen unveiled AARON, widely recognized as the world’s first automated art-generation program—and the grandfather of all modern AI art.

Did you know? AARON wasn’t a “deep learning” model trained on billions of photos scraped from the internet. In fact, deep learning didn’t exist yet! Instead, AARON was built entirely on hand-coded rules of art and logic.

How AARON Made Its Historic Images

The very first images generated by AARON looked vastly different from today’s colorful digital creations. They were simple, abstract, black-and-white line drawings printed on paper using a mechanical drawing machine (a plotter) hooked up to a mainframe computer.

Instead of copying existing art, Cohen manually taught AARON the “rules” of the visual world. He programmed concepts like:

• Proportion and scale: How objects relate to one another in space.
• Occlusion: How one object overlaps and hides another (