Bogosity, Enshittification, and the Glittering Drain
Acronyms used in this post:
AI: Artificial Intelligence, software that can generate or analyze text, images, audio, code, and decisions by learning patterns from data.
SEO: Search Engine Optimization, the old craft of making web pages easier for search engines to find, understand, and rank.
GDP: Gross Domestic Product, the usual national measure of economic output, often worshiped as if it were a thermometer for civilization itself.
The internet did not become stupid because machines became clever. It became stupid because we trained millions of people to pass exams without teaching them how to smell nonsense.
That is the beginning of the tragedy.
Not the whole tragedy. The whole tragedy has more rooms. One room contains greedy social media companies. Another contains lazy schools. Another contains uncles forwarding miracle cures before breakfast. Another contains young men with ring lights explaining geopolitics like they are selling socks near Gariahat. And now, in the biggest room, humming softly like a refrigerator full of ghosts, sits AI.
But the first crack was educational.
A person who has never been taught how to ask, “How do you know?” is defenseless on the internet. He may have a degree. He may have a LinkedIn profile polished enough to blind a crow. He may write “thought leader” in his bio and still fall flat before a smooth paragraph, a fake chart, a dramatic video, or one of those muscular WhatsApp messages that begins, “Doctors will never tell you this.”
Doctors often will not tell you this because it is nonsense.
The old internet, for all its pimples and broken links, still had the air of a dusty library. You searched. You wandered. You found a webpage from 2003 written by a man with terrible fonts and honest intentions. It looked like a railway notice pasted during a storm, but it taught you something. You had to read. You had to compare. You had to develop that small inner goat that chews before swallowing.
Then the feed arrived.
The feed did not ask, “Is this true?” The feed asked, “Will this keep him staring?”
That small change ruined the furniture.
Truth is slow. Anger is fast. Doubt takes effort. Outrage arrives fully cooked, like oily telebhaja in a newspaper packet. You know it may hurt later, but in that moment the crunch is magnificent.
So the platforms learned the oldest trick in the bazaar. Put the shiny thing at eye level. Put the sweet thing near the child. Put the cheap anger where the tired adult will reach for it at 11:47 p.m., when the fan is making its wounded helicopter sound and the day has left him with exactly three rupees of patience.
I know that man.
I am sometimes that man.
Middle-aged, Bengali, broke in the soul before broke in the pocket, living somewhere on the shanty edge of Calcutta where the city becomes brick, dust, tea steam, barking dogs, political posters, and damp laundry. I have worked in serious American healthcare IT systems. I have studied data, standards, workflows, and the many ways respectable institutions can produce nonsense while wearing shoes. I also know what it is to stare into a phone because the mind has become a dark rented room and the phone is the only tube light still flickering.
That is how they get us.
Not because we are idiots.
Because we are tired.
Failed education gives the internet its prey. Failed economics gives it hunger. Failed politics gives it rage. Failed loneliness gives it appetite. The platform supplies the plate.
Then AI enters, wearing a clean shirt.
This is the nasty upgrade. Earlier, nonsense required labor. Some poor fellow had to sit and type the conspiracy theory. He had to make the bad poster. He had to crop the politician’s face badly. He had to add twelve red arrows and a font usually seen on cheap detergent packets.
Now nonsense has a machine.
AI can produce one thousand polite lies before your tea cools. It can write fake expertise in perfect grammar. It can make a fabricated image look like a photograph. It can give a fool the tone of a professor and a fraud the patience of a schoolteacher. That is new. Not the lying. Humans have lied since we discovered language and immediately used it to deny eating the mango. The new thing is the speed, polish, and scale.
Bogosity used to arrive sweating.
Now it arrives well formatted.
That is why people fall for it. A clumsy lie triggers suspicion. A fluent lie slips through the door, asks for water, compliments your curtains, and steals the silver spoon.
You think the danger is that AI will become conscious.
Maybe one day. Who knows.
The nearer danger is duller and more profitable: AI will make rubbish look respectable. It will give every half-baked claim a blazer, every rumor a bibliography, every prejudice a whiteboard diagram, every scam a soothing customer-service voice.
And our education has not prepared people for this.
We taught children to underline answers, not interrogate claims. We taught them to fear wrong answers more than bad questions. We taught them to memorize the digestive system but not digest an argument. We taught them to respect authority, which is sometimes useful, but we forgot to teach them that authority can be rented by the hour.
A blue tick. A famous face. A confident voice. A video with dramatic music. A screenshot with a logo. A paragraph beginning, “Studies show…”
These are not evidence. They are stage props.
But on the internet, stage props often win.
Because the internet is not a debate hall. It is a fish market with electricity. Everyone is shouting, the floor is wet, someone is weighing something with suspicious enthusiasm, and the freshest-looking fish may have been painted fifteen minutes ago.
Social media made this worse by learning our emotional shortcuts. It discovered that we click what flatters us. We share what frightens us. We believe what gives shape to our private resentment. A conspiracy theory is not just a false explanation. It is a small emotional chair. The believer sits on it and feels taller.
That is why correction often fails.
You bring facts. He brings identity.
You bring sources. He brings humiliation.
You bring patience. He brings twenty-seven forwarded videos and a cousin in Dubai who “knows the real story.”
There is comedy here, but also danger. A society that cannot tell evidence from decoration becomes easy to rule, easy to sell to, easy to inflame, and easy to drain. It will buy fake cures, fake nationalism, fake history, fake science, fake investment schemes, fake outrage, fake intimacy, fake revolution, and fake wisdom from a man sitting in front of foam panels and calling himself a truth warrior.
Meanwhile, the platform smiles.
The platform is not your friend. It is not even your enemy in the old-fashioned way. A proper enemy at least has the courtesy to dislike you. The platform does not care enough. It measures you. It nudges you. It predicts you. It keeps you breathing inside the feed like a fish in a glass bowl.
First it served users.
Then it served advertisers.
Then it served itself.
That is enshittification in plain clothes. The shop opens with good tea. Then the tea becomes watery. Then the chairs become uncomfortable. Then the owner charges extra for shade. Then he sells your conversation to a man outside who wants to sell you a miracle belt for back pain.
By the time you complain, all your friends are sitting there too.
This is the trap. Networked rot is sticky. We do not stay on bad platforms because we love them. We stay because our people are there, our work is there, our audience is there, our memories are there, our little economies of attention are there. Leaving feels like moving house during monsoon.
AI will not replace this trap. It will furnish it.
Already the web is filling with synthetic mush: articles written for ranking, images made for clicks, comments generated for manipulation, reviews that smell of plastic, videos narrated by voices that sound almost human, like someone trained a refrigerator to do radio.
Soon the question will not be, “Is this information online?”
The question will be, “Did a human mean it?”
And after that: “Does meaning even matter if the machine can keep me looking?”
This is where I become gloomy, which for a bankrupt bipolar Bengali is not exactly a rare solar eclipse. But gloom has uses. It can point at the damp patch before the ceiling falls.
The damp patch is this: we are building a civilization where attention is extracted faster than judgment is trained.
That sentence is the little snake under the flowerpot.
Attention is natural. Judgment is cultivated. A child has attention. A dog has attention. A mosquito has excellent attention if you are trying to sleep. Judgment takes practice. It takes teachers, reading, argument, mistakes, embarrassment, correction, time, and the slow discovery that your first reaction is often just your nervous system wearing a hat.
The internet attacks judgment by exhausting attention.
Scroll long enough and everything becomes equal. A war, a recipe, a cat, a lynching rumor, a shampoo ad, a poem, a fake medical claim, a celebrity divorce, a flood, a joke, a stock tip, a dead child, a dancing politician. The mind cannot keep moral posture through this parade. It slumps.
Then the machine offers you another clip.
Just one.
Always one.
There is no final cigarette in the packet of the feed.
So what do we do? Not the grand answer. Grand answers are usually slogans wearing perfume.
We begin smaller.
Teach people to pause.
Teach children that a claim is not true because it is written well. Teach adults that anger is not evidence. Teach users that AI is not a priest, judge, uncle, witness, or professor. It is a tool that can be useful, wrong, brilliant, shallow, and confidently lost in the same afternoon.
Teach the habit of asking: who made this, why now, what is missing, who profits, what would change my mind?
That last question is dangerous. It may even make you civilized.
We also need platforms to be treated less like magical innovation temples and more like public drains. If a drain overflows, the city suffers. You do not say, “Ah, but drainage is a private matter between the pipe and its investors.” You inspect it. You regulate it. You punish negligence. You force maintenance.
Information systems are drains for public attention. When they clog with sewage, everyone breathes it.
And yes, governments will misuse this argument if allowed. They will call criticism misinformation and call propaganda national interest. So regulation must come with transparency, independent oversight, researcher access, appeal rights, and real public pressure. Otherwise the cure becomes another fever.
There is no clean ending here.
The internet will not return to its innocent library days. The shelves have been moved. The casino has opened inside the reading room. A man is selling glowing coconuts near the philosophy section. An AI parrot is summarizing books it has not read. A politician is live-streaming from the staircase. Someone has started a podcast beside the fire exit.
Still, the library is not gone.
That is the maddening thing.
Inside the noise, there is still knowledge. There are still teachers, archives, lectures, papers, maps, manuals, poems, patient explainers, old photographs, public data, open-source tools, and quiet people who know what they are talking about. The treasure remains. The path is simply covered with glittering garbage.
A better education would teach people to walk through that garbage without eating it.
A better internet would stop rewarding those who dump it there.
A better AI culture would admit that fluency is not truth, speed is not wisdom, and scale is not civilization.
Until then, keep your suspicion sharp.
Not bitter. Sharp.
Bitterness is easy. It sits in the corner and spits. Sharpness cuts fruit, rope, lies, and sometimes the knot inside your own head.
That is still worth something.
Even here.
Even now.
Even beside the drain, under the weak fan, in the hot Calcutta afternoon, with the phone glowing like a small hired demon and the world insisting that the next swipe will explain everything.
It will not.
But the pause before the swipe might.
P.S. References: Cory Doctorow’s writing on enshittification;