You Asked the Stars. Now You Ask the App. Same Deal.

Somewhere in ancient Rome, a haruspex—a trained priest whose entire job was interpreting the liver of a freshly sacrificed animal—was telling a general whether to march his army into battle. The general probably already knew what he wanted to do. He just needed something external to confirm it. Something with authority. Something that, if things went sideways, wasn't him.

That general would have been very comfortable with a LinkedIn job recommendation feed.

The Oracle Was Never Really About the Future

Here's the thing about ancient divination systems: they weren't primarily predictive tools. They were decision-laundering machines. When a Roman magistrate consulted the augurs before a major vote, or when a Greek city-state sent a delegation to Delphi before going to war, the point wasn't to actually learn what the gods thought. The point was to get a legitimizing stamp on a decision that was already half-made—and to distribute the blame if it turned out badly.

The Delphic Oracle was famously ambiguous, which wasn't a bug. It was the whole product. 'A great empire will fall if you cross the river' tells you nothing about which empire. Croesus of Lydia found that out the hard way in 547 BCE when he interpreted that prophecy as permission to invade Persia and lost his own kingdom instead. The oracle wasn't wrong. The oracle was never accountable.

Sound familiar? When your credit score comes back with a number you can't fully explain, generated by an algorithm you'll never see, run by a company that doesn't have to justify its methodology to you—that's Delphi. When TikTok's recommendation engine serves you a particular kind of content so consistently that you start to feel like the algorithm knows you—that's augury. When a hiring platform's AI filters your resume before a human ever reads it, and you don't get the job, and no one can tell you exactly why—that's a haruspex reading entrails and shrugging.

The System Decided. Nobody's Fault.

The psychological function here is almost embarrassingly consistent across five thousand years of recorded human behavior. We don't want to be responsible for outcomes we can't control. And since most major outcomes—who gets credit, who gets hired, what we consume, who we date—feel genuinely uncontrollable, we prefer systems that take that decision out of our hands.

This isn't stupidity. It's a rational response to uncertainty. If the system decided, you can't be blamed for the outcome, and you can't be accused of favoritism or bias (even if the system is riddled with both). Ancient societies understood this intuitively. Astrology in Mesopotamia wasn't just personal guidance—it was a bureaucratic tool. State astrologers in the Assyrian Empire produced regular reports for the king that functioned like policy briefings. The stars said. The stars said to wait. The stars said the harvest would fail. Nobody had to own that.

Modern algorithmic systems are doing the exact same thing at scale, with better PR. The credit score diffuses blame so thoroughly that no individual lender has to tell a Black family in a redlined neighborhood that they're being denied a loan because of who they are. The system decided. Objective. Data-driven. Nothing personal.

Who Benefits From 'Objective'

Here's where the history gets uncomfortable. Ancient divination systems consistently—not occasionally, consistently—reinforced existing power structures. When Roman priests declared the omens unfavorable for a vote, it was usually a vote that powerful people didn't want to happen. When the Oracle at Delphi blessed a particular military campaign, the temple had often already received significant donations from the people proposing that campaign.

The augurs weren't independent arbiters of divine truth. They were institutionally embedded in the same elite networks as the people consulting them. Their 'objective' readings had a funny way of confirming what the powerful already wanted.

You can probably finish that thought yourself. The companies building recommendation algorithms, credit scoring systems, and automated hiring tools are not neutral parties. They're embedded in the same economic networks as the institutions using their products. When Amazon's hiring algorithm turned out to systematically downrank resumes that included the word 'women's' (as in 'women's chess club'), that wasn't a glitch in an otherwise objective system. That was the system accurately reflecting the preferences of the people who built and used it. The algorithm didn't have a bias problem. It had a honesty problem—it was telling the truth about existing preferences while claiming to be above them.

Why We Keep Asking Anyway

None of this means people are naive. The Roman generals consulting augurs knew, on some level, that the priests were not receiving transmissions from Jupiter. The Assyrian king reading his astrologers' reports knew the stars weren't literally dictating policy. And most adults today understand, in the abstract, that Netflix's recommendation engine is not a neutral window into their own taste—it's a tool designed to maximize their time on the platform.

We use these systems anyway because the alternative is harder. Making genuinely autonomous decisions, without external validation, under real uncertainty, while accepting full accountability for the outcome—that is cognitively and emotionally expensive. The oracle, the algorithm, the credit score: they all offer the same product. A sense that something larger than you has weighed the situation and rendered a verdict. That the decision has been made by a process, not a person.

And that feeling is worth a lot. It's worth sacrificed animals. It's worth data privacy. It's worth not knowing why you didn't get the apartment.

The One Thing That Did Change

Ancient divination systems required physical presence, trained specialists, and elaborate ritual. They were slow and expensive, which meant they were mostly used by states and wealthy individuals for high-stakes decisions. Your average Roman farmer wasn't consulting a haruspex about whether to plant barley.

The modern version runs on your phone, costs nothing to access, and operates continuously in the background of daily life. The scale is genuinely new. The number of micro-decisions now being outsourced to algorithmic systems—what to read, who to follow, what to buy, which route to drive, which person to swipe right on—represents something quantitatively different even if it's psychologically identical.

What that means is that the comfort of 'the system decided' is now ambient. It's not a special occasion anymore. It's just Tuesday.

The haruspex would have loved the subscription model.