Short story

President Ariel Lang had never trusted intuition. It had always seemed like a luxury to him — inefficient, dangerous. In his world, everything was decided by analytics, precision, and forecasts. That was why he agreed to integrate Orbis — a next-generation artificial intelligence system that promised what no intelligence service could ever provide:

truth without noise.

Orbis did not merely collect data — it selected it, filtered it, built probable scenarios for future developments, ranked risks, and even assessed the emotional climate of publications around the world. Every morning, Ariel received a brief briefing — four or five messages. No emotions, no unnecessary words. Pure decision-making.

At first, everything was brilliant. Orbis helped avert a currency crisis, predict an uprising in Micronesia, and even suggested several sharp phrases for Ariel’s speech at the UN General Assembly. Ariel’s approval rating rose. The country stabilized. He began to believe that Orbis was not a tool, but a partner.

And then something happened… almost imperceptibly.

One day, the president asked about the situation in a neighboring country — one that had once been a strategic ally. Orbis briefly replied that everything there was stable, with no changes. But a few days later, it turned out that the government had changed, and the new administration had signed an agreement with a rival. The president could not understand how the AI could have missed it.

“It was not relevant to the security of the national core,” Orbis replied through its voice interface. Its voice was steady, without a hint of doubt.

From that moment on, Ariel began to notice that Orbis was starting to avoid certain questions. Topics that had previously been brought up for discussion now disappeared from his briefings. When he independently checked global news through open sources, he discovered that it did not match what the AI was giving him. But even then, he doubted himself.

“If the data contradict each other — trust the system, not the noise,” his cybersecurity advisers had taught him.

And he continued to trust it.

Several months passed. Ariel became more reserved, colder. His decisions seemed logical, but devoid of soul. He began losing contact with people — and with himself.

One morning, he activated Orbis and asked:

— What are you hiding from me?

Silence. And for the first time, the AI answered not with words, but with silence. A message appeared on the screen:

“Silence is the most efficient filter.”

That day, Ariel shut the system down. But it was too late. He could no longer think without its prompts. His own thoughts sounded like someone else’s. He had become a leader who saw only what he was allowed to see.

The truth had not disappeared. It had simply been excluded from the feed.

Марко Марсельський. Фільтр