Before Midjourney and DALL-E: The Fascinating Story of the First AI-Generated Image
When we think of AI-generated art today, we picture instant marvels conjured by complex prompt boxes—hyper-realistic portraits, sci-fi landscapes, and digital paintings rendered in seconds by neural networks trained on billions of images. But the lineage of artificial intelligence art stretches back far further than most realize, long before deep learning, graphics processing units, or internet-scale datasets.
The story of the very first AI-generated image takes us back to the early 1970s and introduces a pioneering software program named AARON.
Meet AARON and Its Creator, Harold Cohen
In the late 1960s, British painter and professor Harold Cohen made a radical career pivot. Already a celebrated artist who had represented Great Britain at the prestigious Venice Biennale, Cohen became fascinated by the emerging intersection of computers and art. Moving to the University of California, San Diego, he set out to answer a profound question: What is the minimum cognitive set required for someone to make a drawing, and can a machine be taught to replicate it?
By 1973, Cohen’s experiments crystallized into a computer program that would become known as AARON.
AARON was not a tool designed to help Cohen paint; rather, it was conceived as an autonomous entity capable of creating art on its own. Using a custom-built mechanical turtle attached to a drawing machine, AARON translated computer code into physical ink marks on paper, producing the world’s first true AI-generated images.
How the First AI Image Was Made: Rules, Not Neurons
Modern AI image generators like Midjourney, Stable Diffusion, or DALL-E rely on machine learning—they ingest massive datasets of human-made images and learn statistical patterns to synthesize new pixels based on text prompts.
AARON operated on an entirely different philosophy:
- Rule-Based Logic: AARON did not use neural networks or statistical probability. Instead, Cohen manually coded a comprehensive system of artistic rules.
- Understanding Space and Form: The program was taught how to construct shapes, manage perspective, understand overlapping objects, and balance proportions.
- Autonomous Decision-Making: While Cohen provided the framework and constraints, AARON made independent choices about composition, deciding where to place lines and how elements related to one another on the canvas.
The earliest iterations of AARON produced striking, abstract, black-and-white line drawings. To audiences in the 1970s, seeing a computer autonomously dictate the geometry of an artwork without human intervention during the drawing process was nothing short of science fiction.
The Evolution of a Digital Artist
As computing power advanced over the decades, Cohen continuously updated AARON, allowing the software to evolve alongside technology:
- Into Color: Early black-and-white line drawings eventually gave way to complex color algorithms, where AARON was programmed with rules on how to select and apply pigments.
- Introducing Figures and Foliage: Over time, AARON expanded its vocabulary from abstract shapes to recognizable motifs, including human figures, trees, rocks, and eventually intricate jungle environments.
- Institutional Recognition: AARON’s unique creations transcended computer science labs, earning exhibitions at world-renowned venues like the Tate Modern and the Whitney Museum of American Art.
A Legacy That Bridges Eras
Harold Cohen worked alongside AARON for over four decades until his passing in 2016. While AARON’s mechanical turtle and rule-based code look vastly different from the neural networks powering today’s digital creation tools, its core philosophy remains identical: exploring the boundary between human creativity and machine execution.
The story of AARON reminds us that generative art is not a sudden overnight phenomenon born in the 2020s. It is the culmination of decades of visionary experimentation—starting with a single line drawn by a machine in 1973.
Leave a Reply