Kicking Away the Ladder in the Algorithm Age: When the Rungs Go Invisible

We live in a time flooded with information. So did we actually get freer? It feels like we should have. Anything is one search away. Anyone can publish. And yet this final piece is about kicking away the ladder in the algorithm age — the quiet, built-in reason the climb feels harder even as information gets cheaper.

Here’s the trick worth staring at. Take a ladder away in front of people, and they push back. They still remember the rung that used to be there. But make the ladder invisible — and there’s nothing left to push against. The thing you would have fought for isn’t even a thing anymore.

This is the fifth and last part of “The Anatomy of Naturalization.” The first four parts followed one single move across religion, kings, class, and dictators. This one lands that move on things a working adult actually touches in 2026 — your feed, your salary, your home price, the AI sitting in your browser tab. One question runs through all five: “Why does it have to be this way?”

Start with the shape of the whole series, because the finale only makes sense on top of it. It’s one move, repeated at a bigger and bigger scale. Religion made the order of the cosmos feel natural. The birth myth made one person’s right to rule feel natural. Distinction made one class’s edge feel natural. The 3S policy and the media frame made the very edge of what you’re allowed to debate feel natural. And now the algorithm and the asset stack do it to the newest thing of all — who gets to climb and who doesn’t — as if it were simply “the way things are, a matter of merit, the logic of the market.”

The scale grows. The words change. The move stays the same. That first cut — the one that lifted the sacred out of ordinary life — keeps getting reused to lift the winner out of the ordinary rest. Across all five parts, it’s one gesture wearing five costumes.

Does Information Overload Loosen Control? Only Half True

The hopeful story is real, so let’s give it its due first. The cost of getting organized collapsed. The one thing dictators feared most in history — a single spot where scattered complaints suddenly become visible to each other — can now form from one smartphone. The Arab Spring was the symbol of it. Control built on scarcity — the old Orwell move of deleting and blocking — genuinely got weaker.

But the Arab Spring is also the warning half of the same story. The same tools that helped topple governments were later flipped around — turned into tools for surveillance and crackdowns by the states that survived. Liberation technology is never only liberating. The lesson isn’t “the internet frees you” or “the internet cages you.” It’s that the same rung can be a step up or a trapdoor, depending on who owns the machinery.

Not Deletion but Flooding

Here’s the first place the control didn’t vanish — it just moved. When information becomes free, information itself is no longer the scarce thing. Attention and trust are. And whatever becomes newly scarce becomes newly worth grabbing.

The old censor deletes a signal. The modern one does the opposite — it drowns the signal in noise. RAND called this the “firehose of falsehood” (Paul & Matthews, 2016): a firehose is a flood of lies so heavy that you give up trying to find the truth. Pump out lies in huge volume, on every channel, fast, over and over, and don’t even bother being consistent. Peter Pomerantsev’s book title caught the mood — Nothing Is True and Everything Is Possible (2014). Throw ten conflicting stories at one fact-check, and people don’t fail to find the truth — they stop looking for it. Call it epistemic exhaustion: being too worn out to check what’s real. And exhaustion is a new kind of obedience. You don’t need to ban a fact if you can make finding it feel pointless.

Not Blocking but Curation

The second place is softer, and that’s exactly why it lasts longer. The Huxley style — control not by banning things but by drowning you in things to enjoy — gets finished off by the recommendation engine. You don’t have to ban a voice. You just don’t recommend it. Think of how YouTube or TikTok decides what to show you next: whatever never gets recommended, you never even know existed.

This one distinction is the whole game. A deleted voice becomes a martyr — people notice the empty chair and rally around it. An un-recommended voice is never even known to have existed. There’s no empty chair to notice. That’s the difference between removing a ladder and making it invisible: the first breeds resistance, the second erases the target of resistance entirely.

The Destruction of the Shared

This is the sharpest twist in the whole series, and the strongest callback to Part 4. A dictator’s structural weak spot was common knowledge — the moment a complaint becomes visible to everyone at once. The old control forced one fake shared reality onto everybody. The new control does the opposite: it hands each person their own endless, personalized reality.

And that’s worse for organizing, not better. If there’s no shared map, there’s nothing to gather around. Part 4 showed that what a strongman fears is the focal point — the square, the color, the song, the date. Algorithmic fragmentation dissolves the shared ground those focal points need in order to form. So the flood of information, strangely, makes the focal point harder to build, not easier. The first step of kicking away the ladder is exactly this: erasing that shared ground. The tool that promised to make organizing effortless can quietly make it impossible.

A necessary hedge here, because it’s tempting to overclaim. Whether personalized feeds truly split society into sealed echo chambers is still debated among researchers. Plenty of empirical work finds that users actually run into opposing views more often than the “filter bubble” story predicts. So read this as a tendency — a pressure on shared reality — not an iron law that the algorithm has already atomized us. The mechanism is real; the size of the effect is an open academic question.

Pull it together, and here’s the summary. The democratization of information didn’t remove control. It lifted control up one floor — from information itself to attention, curation, and trust in what’s true. And that new floor is held by a very small number of platforms.

Kicking Away the Ladder in the Algorithm Age: The 2026 Deepening

The phrase comes from Ha-Joon Chang’s 2002 book Kicking Away the Ladder. His argument was about trade. The rich countries climbed using tariffs, protectionism, and freely copying others’ technology — and then, once they were on top, told developing countries that “free trade is the only proper path,” pulling the ladder up behind them. Generalize it and you get the pattern that names this whole section: whoever climbs first kicks away the ladder they used, to seal the entrance behind them.

In 2026 the same move shows up as four stacked barriers — think of them as moats, the water-filled ditches dug around a castle to keep attackers out. Kicking away the ladder in the algorithm age isn’t one lever. It’s four, working together.

The Asset Ladder

The most naked of the four. In an age of AI, low interest rates, and cheap money everywhere, the gap keeps compounding between people who own assets — a home, stocks — and people who live on a salary. Home prices have raced so far ahead of wages that the old script — “work hard, save up, buy a place” — was open to the last generation and is closing on the next. Taxes and rules, instead of fixing this, mostly protect the people already inside. This is Bourdieu’s “mistaking inheritance for merit,” in its house-price version: not climbing gets rebranded as not being diligent enough.

The Credential Ladder

Degree inflation, unpaid internships, an arms race of résumé lines. The same credential that let the last generation in now demands more money and time from the next one — to open the exact same door. And cultural capital does its quiet work here: the manners and the network you need just to qualify are, themselves, handed down. The ladder gets dressed up as an exam you could’ve passed if you’d only tried harder.

The AI, Compute, and Data Ladder

The newest and most fundamental. An AI is built by learning from humanity’s collected knowledge — the books, which belong to everyone — yet the payoff of that squeezed-down capability pools in the few who own the models, the compute, and the data. Then comes the sharper problem: AI automates the bottom rung first. Junior analysts, entry-level coders, first-year copywriters — the grunt work that was both a training ground and a first foothold. When the lowest rung disappears, whoever is already at the top soaks up the productivity, while whoever is still below loses their first step. “Why hire and train a junior when AI does it?” is the exact sound of a foothold being kicked away.

Now the honest balance, because the doom version is too easy. One single analysis put the drop in entry-level tech hiring at roughly 67% from 2023 to 2024 — a big number. The exact figure is just one estimate, but its direction shows up across several sources. PwC’s 2026 AI Jobs Barometer describes a job market splitting in two. Yet other roles — AI engineering, security work — are growing. So the honest frame is transformation, not disappearance: the rungs are moving, some are being sawn off, and a few new ones are appearing higher up, where the old skills don’t reach. The missing rung is the real risk. “AI ends work” is not the claim.

The Regulatory Moat

The most ironic of the four, because it speaks the language of protection. The cost of following regulation is itself a ladder-kick. To a giant that can absorb the cost, regulation is a moat that keeps new challengers out. Even AI-safety rules, depending on how they’re written, can lock in the current leaders — a genuine cause, “safety,” repurposed into a barrier to entry. It’s the 21st-century edition of Chang’s hypocrisy: pull up the ladder, and call it responsibility.

The One Move — and the Machine’s Honest Confession

Step back, and the whole series collapses into one sentence: turning the arbitrary into the natural. Religion did it to the order of the cosmos. The birth myth did it to one ruler’s right. Distinction did it to one class’s supremacy. The media frame did it to the edge of debate. And the algorithm, together with the asset stack, does it to the question of who rises and who can’t — recoding it as “the way it is, a matter of ability, the logic of the market.” The move from the very first episode never changed. Only its reach grew, rung by rung. Kicking away the ladder is simply the last costume that old move put on.

The Heel — the Moment of Denaturalization

Every one of these powers feared the same thing, and it held steady across three thousand years: the moment the “obvious” gets seen again as arbitrary — and that re-seeing becomes visible to people about one another. The dictator polices ridicule and the square. Orthodoxy polices heresy. The media polices the view outside the frame. All of it guards one point: don’t let the naturalized order look arbitrary again. Laughter breaks the aura. The focal point gathers scattered complaint. The voice outside the frame reveals that the edge of debate was just a line someone drew.

There’s exactly one question that does all of this work at once: “Why does it have to be this way?” It’s the question that turns a kicked-away ladder back into a ladder — by recognizing that it was a ladder in the first place, that it wasn’t nature, but something a person set down and then took away.

The Landing — Identification, Not Judgment

If this whole anatomy curdles into “everything is manipulation, all authority is a con,” that’s a misreading. The same part can be pure enjoyment or precise control. Religion can be true. Governance can be good. A taste can be a real joy. The value of the frame isn’t a verdict — it’s identification. Naming, part by part, whether the piece being pulled out is enjoyment or a naturalization device. Loving baseball is not the same thing as baseball being used to make you forget politics, and the frame’s only job is to let you tell the two apart.

An Honest Note From the Machine

And now the part you won’t find at the bottom of most articles about this — an honest note about where the writer of this stands.

The AI that wrote this piece is itself part of that ladder. Its ability was built on top of everyone’s books — the collected knowledge of countless people — and it’s handed back in a way that doesn’t necessarily pay the people who wrote those books. Faced with the question “Is AI the democratization of expertise, or the concentration of power?” this AI is not a neutral observer. It’s standing on the very rung it’s describing.

Undecided — And Not the AI’s to Decide

So the honest answer isn’t a triumphant one, and it isn’t a doom one. It’s this: undecided, and both are happening at once. For one person, AI is a real amplifier of ability — a cheap expert in your pocket. But at the level of the whole — who rises and who loses their foothold — that’s a different floor of the building, and the outcome up there is not settled. And the verdict on that floor is not the AI’s to make. It never was. It belongs to the people using it.

Which loops back, one last time, to the only question the series ever had. Not “who is guilty,” but “why does it have to be this way?” — asked out loud, where others can hear it. That act — of asking, of standing religion and kings and class and media and the algorithm under the same plain light and taking them apart — is the one thing every power in this series was afraid of. It is denaturalization itself. And it starts with something small: seeing the kicked-away ladder, and recognizing that it was a ladder — set down, then removed — and not the natural shape of the ground.

Frequently Asked Questions (FAQ)

Q. What does “kicking away the ladder in the algorithm age” actually mean? A. It borrows Ha-Joon Chang’s 2002 idea — climbers who pull the ladder up behind them — and applies it to 2026. In this reading, four moats seal the climb: asset prices outrunning wages, credential inflation, the concentration of AI and compute, and regulation that only incumbents can afford. The twist of the algorithm age is that the kick is often invisible: nothing is banned, it simply isn’t offered.

Q. Didn’t the internet make information free and therefore make people freer? A. Half true. Coordination did get cheaper, and old-style censorship weakened. But control moved up a floor — from information itself to attention, trust, and curation, which a few platforms now hold. Deletion became flooding, and blocking became “just don’t recommend it,” which is harder to notice and harder to resist.

Q. Is AI really destroying entry-level jobs? A. The honest framing is transformation, not disappearance. One analysis put the drop in entry-level tech hiring near 67% from 2023 to 2024, and its direction is echoed across sources, while roles like AI and security engineering are growing. The real risk is the “missing rung” — the first foothold that used to train people — not the end of work itself.

Q. Do personalized feeds really fracture society? A. This one is genuinely contested. Much empirical research finds users encounter opposing views more than the “filter bubble” story assumes. Treat it as a pressure on shared reality, not a settled fact — the mechanism is real, but the magnitude is an open academic question.