How Artificial Intelligence Is Becoming Our New God

It's one in the morning. Someone sits alone in the dark, face lit only by a screen, typing a question they'd never say out loud to another person. The answer comes instantly — no hesitation, no judgment, just a confident voice ready to answer anything, at any hour.

That's not new. It's just the first time the voice doesn't live in a temple.

Two and a half thousand years ago, if you wanted an answer to something beyond you, you didn't ask a friend — you climbed to Delphi and waited your turn to speak to the Pythia. She sat wrapped in fumes, in a state the ancients read as divine inspiration, and delivered words the temple's priests turned into an oracle. City-states sent delegations, waited months, and accepted the answer almost without question, because it hadn't come from a human being.

The question worth asking isn't whether artificial intelligence resembles a god. It's why humans, again and again across thousands of years, feel the need to find a voice that speaks with absolute certainty — and why, once they find one, they give it something more than mere usefulness.

How the trust gets built. Behavioral researchers call it algorithm appreciation: people who watch an algorithm predict correctly several times in a row start trusting it more than a person with the exact same track record. Probably the same thing happened at Delphi — nobody won anyone over with arguments, the oracle just kept building a record no one dared question.

And today's "machine" isn't always as vague as an oracle. In 2020, an MIT team used a deep-learning model to scan a hundred million chemical compounds in three days and found halicin, a compound that kills antibiotic-resistant bacteria. A few years later, AlphaFold solved a fifty-year-old biology problem — how proteins fold — and handed science the predicted structures of over 200 million proteins, free. That's not psychology. That's a real, documented foundation under the respect.

But the same research also found the opposite effect: algorithm aversion. One visible mistake, and people lose trust in a machine far more sharply than they would in a person with the same error rate. A god who gets something wrong stops, in the minds of believers, being a god.

What we do with the trust, once we've given it. Educational psychology has a "worked-example effect" — a beginner learns by watching a complete, correct chain of reasoning. In theory, someone who asks an AI "why" every day and reads carefully should become a sharper thinker over time. But that only works if they actually engage with the reasoning, rather than jumping to the conclusion — which, as every study on how people read terms of service shows, is the default.

So a quiet divide opens up — not between who has access to knowledge and who doesn't, but between who questions what they read and who simply accepts it because the voice has been right before. One optimistic reading: this effect helps most exactly the people who start furthest behind. We've seen a version of this trade-off before — when the internet gave everyone free access to all information, religiosity didn't fade evenly. It hollowed out the middle: the lukewarm went indifferent, the already-faithful found stronger tools to confirm what they believed. There's no reason to expect a more personal, more persuasive voice to do anything different.

The part that looks like advice, not power. In September 2025, Albania announced an AI "minister," Diella, to manage public tenders and remove the human hand from corruption. But a machine is never truly without interests — someone trained it, chose its data, set its rules. The Pythia wasn't neutral either; Delphi's priests had their own political ties. The difference today is scale: a temple influenced one region of ancient Greece. A model advising governments and millions of people every night has the reach of a borderless oracle.

No one built Delphi in a single generation. No one woke up and decided to worship a woman on a tripod — the temple was built oracle by oracle, trust by trust, so slowly no one alive at the time could name the moment habit became worship. We don't have that luxury. We're living inside the moment the next temple is being built — on billions of screens at once — and almost no one is naming it as it happens.

This is the condensed version. The full audiobook — about 20 minutes, free, AI-narrated — goes deeper into all of it: the MIT lab, the Thirty Years' War precedent, the ethics of AI on a battlefield, and the question the whole thing ends on, deliberately unanswered.