Today that architecture sits inside graphics processors and AI chips. In the early 1990s it still felt experimental.
My assignment at KSR focused on visualization. Once computers begin generating mountains of data, another question appears immediately: how do humans understand what the machine has discovered?
I saw one answer during a visit to Cornell University.
In a dim lab filled with the hum of cooling fans, a supercomputer was drawing storms. Not weather forecasts exactly—fluid-dynamics simulations. On the screen ribbons of motion twisted through space, folding into eddies and whirlpools while colors shifted from cool blues to violent reds.
Researchers leaned over terminals debating what they were seeing. What had begun as pages of equations was suddenly moving across the screen.
The computer had solved the math. What mattered was that we could see the structure.
That moment stuck with me because it revealed something simple: when numbers become pictures, patterns appear that nobody had noticed before.
A few months later I saw the same idea again at the University of Nevada. Researchers there were analyzing health data for the National Institutes of Health. Instead of maps they used vectors—long strings of numbers describing individual lives.
When the results appeared on the screen it looked like a small galaxy.
Hundreds of colored points floated in space, each representing a person. Similar individuals drifted toward one another while others pushed apart. After a few minutes clusters began to emerge—constellations inside the cloud of points. Some reflected geography, others demographics or environmental exposure.
The computer hadn’t created anything new. It had simply revealed structure hidden in the data.
Watching scenes like that gradually changed the way I thought about intelligence—not artificial intelligence, just intelligence.
Humans do something similar all the time. Long before computers existed, our ancestors survived by noticing patterns: movement in tall grass, tension in someone’s voice, clouds thickening on the horizon before a storm. Those weren’t equations; they were clues.
Over millions of years, evolution turned the brain into a powerful pattern detector. Most of the time we don’t notice it working. We simply get a feeling about something—a place, a person, a situation.
Later science explained what the brain was doing all along. Neurons detect edges and motion. Layers combine signals into larger patterns.