"Colors of Noise"
The experiment progressed slowly but steadily. Each morning began with a system startup, each late morning with coffee and a discussion of what hadn't worked the previous day. Gradually, step by step, the researchers made progress. The hardest part was establishing signal acquisition.
The electroencephalograph, assembled from discarded parts, looked like a hybrid of a medical device and a radio receiver. Bundles of wires, wrapped in electrical tape, protruded from a housing constructed from an old cassette recorder. The contacts were attached to the head using electrodes modified in the department, and the signal was amplified by a homemade circuit using operational modulators. It all looked wild, but it worked.
The real breakthrough came when they modified the code. Now the program didn't just generate random numbers. It reacted. Reacted to the strength of the signal coming from the device. The range of probabilities changed depending on the pulse intensity. The color palette - 16 shades in total - began to reflect this interaction.
Black and dark gray meant no signal at all.
Light gray and white meant a weak, barely perceptible response.
Blue and dark blue meant a medium-strength signal.
Red and dark red meant a strong pulse, a sharp spike.
Over the course of a week, they observed how the screen, previously flickering meaninglessly, began to "respond" to internal states. Sometimes with almost nothing, sometimes with flashes of red when the subject was tense, angry, or, conversely, concentrating.
On the eighth day, the system malfunctioned. The machine's memory was limited, and at some point it began displaying old data mixed in with new. The screen became filled with artifacts - glitches, as Tom called them. The same fragments repeated, appearing where they shouldn't have. At first, this was mistaken for an error.
"It's just a buffer overflow," Professor Miller said, typing on the terminal.
But then it became clear that these "errors" were repeating themselves too systematically. They didn't disappear, they kept coming back. As if the machine was trying to remember something.
"We should mark them," the professor suggested. "Dark green. And if they reappear, green."
Sarah, always attentive to detail, added:
"And if the machine reacts to them, let it color them yellow. That means it's not just noise. It's a response."
The code changes were made that same night. By morning, the screen was no longer simply chaotic: yellow spots surrounded by green ones began to appear in the gray fog. They flickered, disappeared, and returned. The group watched them silently, hesitant to draw any conclusions.
But at some point, Jonathan pursed his lips and whispered:
"That face. I've seen that face."
On the screen, among the glitches and spots, features were indeed discernible. Not distinctly, not photographically. But enough for the brain to fill in the blanks. A face. A room. A book on the table. All of this flashed through my memory, and now - on the screen.
The machine seemed to have begun to "remember."
Whether this was a coincidence, a play of imagination, or something more - no one knew. But no one turned off the system. The experiment continued.
previous
Table of contents
continue