Made for You, or Made to Keep You?
"Personalised" was supposed to be the best thing that ever happened to the news. For most of history the same paper landed on every doorstep; the promise of the screen was a paper that finally knew you — your interests, your level, your time. Then the word went sour. Say "personalised news" now and what comes back isn't tailored — it's filter bubble, echo chamber, clickbait, rabbit hole. A promise about being known turned into a warning about being trapped.
Here's the question hiding inside the word, the one nobody asked out loud: made for you — or made to keep you? Those are not the same product. And for fifteen years we've been sold the second one wearing the name of the first.
The Bait-and-Switch
Strip the marketing away and most "personalisation" was a single trick: watch what you click, then show you more of whatever you click. Put like that it sounds servile — the system bending to your wishes. It isn't. It's optimisation, and the thing being optimised was never your understanding. It was a number on the platform's dashboard: time on app, sessions per day, the odds you tap one more time before you put the phone down.
Those goals only look like yours. A system tuned to keep you watching will happily learn that you click outrage faster than nuance, that you linger on what confirms what you already believe, that a cliffhanger headline beats an honest one. It is not misreading you. It is reading you perfectly, and then serving the platform. Eli Pariser saw the shape of this early and, back in 2011, gave it a name — the filter bubble, the private universe of information each of us ends up living in, assembled by algorithms we never see and can't argue with. The trouble was never that the machine got you wrong. It's what it did once it got you right.
Whose Side Is It On?
There's a simple test for whose side a system is on: does it ever tell you to stop?
Something built for you would, sometimes, show you less. It would say: that's the important news today, you're caught up, go live your life. Nothing optimised for engagement has ever said that, because "you're caught up" is a failure state — a closed app, an ended session, a number that goes down. So the scroll has no bottom. There is always one more thing, tuned to be a little more arousing than the last. Zeynep Tufekci described watching this happen on YouTube: start with a campaign rally and the recommendations march you toward extremes; start with jogging and you arrive at ultramarathons. You are never hard core enough, as she put it. That isn't a glitch. Escalation is the product.
The echo chamber is the same machinery pointed at your beliefs. Show people only what they already nod along to and engagement climbs while understanding sinks; Cass Sunstein called the result an information cocoon — a comfortable room with no doors. And the cruel twist is that none of it even makes people happy to be there. When researchers ask why nearly forty percent of people now actively avoid the news, the answers aren't "I don't care." They're: it's repetitive, it's exhausting, it's hard to tell what's true. Those are the symptoms of a system optimised for reaction, not for you. People can feel the hand on their sleeve, even when they can't name it.
Tailored to Digest, Not to Scroll
None of this means personalisation was a bad idea. It means we let it serve the wrong master. Tailoring the news to a person is a good instinct — it was just aimed at holding your attention when it should have been aimed at respecting it.
Think about what "made for you" should mean for something you actually want to digest. The same story, pitched at the depth you bring to it — a headline's worth when it's a passing interest, the full picture when it touches your life. In the language you think in. At a length you can finish and then close, instead of one engineered to spill into the next thing. That's personalisation in the service of understanding, and the giveaway is that it ends. The measure of it working isn't how long you stayed — it's that you got what you came for and got on with your day.
We've made the case for depth elsewhere: one story, told at Beginner, Intermediate, or Expert, because you contain more than one reader. And we've written about why active reading beats the passive scroll for the one brain you have to keep. Put those together and you get the opposite of a bubble — not a system deciding what's good for you, but a tool shaping the news around you so the reading is yours to finish.
The Thing They Never Gave You: Control
Underneath all of it is the one thing the old model never handed over: control.
For all the talk of "personalisation," you were rarely allowed to set anything that mattered. You didn't choose how deep a story went, or when the news arrived, or how much of it there'd be. The single dial within reach only ever turned one way — more. Notifications you didn't ask for. A homepage reshuffled by someone else's priorities. "For You," decided entirely without you.
Brief puts the controls back in your hands, on purpose:
- Depth is yours. Read any story as a Beginner, Intermediate, or Expert — a thirty-second primer or the full five minutes — and choose differently from one story to the next.
- Language is yours. Read the world in the language you actually think in, not whatever the original happened to be filed in.
- Timing is yours. One digest, arriving when you want it. Breaking news is optional, and off by default — your day is interrupted only if you say so.
- The whole picture is yours. Every story is drawn from multiple trusted sources, so you can see where they agree and where they don't — the opposite of an echo chamber built to show you a single side.
You decide what you see and how you see it. That sentence sounds obvious until you notice how little of the news you read this decade was ever actually up to you.
So Where Does the AI Go?
It would be easy, and dishonest, to stop there — as if the cure for bad algorithms were no algorithms. Brief runs on AI, a lot of it. It reads across thousands of sources, clusters the day's events, writes each story at three depths, and translates all of it, in about the time it takes you to make coffee. Pretending otherwise would make us exactly the kind of news source this series keeps arguing against.
So here's the honest part. The problem was never that the old products used AI. It's the question they pointed it at. You can aim the most powerful model in the world at "what will keep this person here the longest?" or at "how do I help this person understand this faster?" — and you get two opposite products out of the identical technology. One studies your weaknesses. The other does your reading.
Brief points it at the second question, every time. We use AI to compress the world so you can take in more of it in less time — never to decide what you should read. It doesn't pick your interests, rank your importance, or curate a version of reality engineered to hold you. You choose the story, the depth, the language, the moment; the AI just clears the road so the trip is shorter. The reader stays in the driving seat — where personalisation should have put you all along.
Made for You
Made for you, or made to keep you. The technology can't tell the two apart — only the goal behind it can, and for fifteen years that goal belonged to someone else. Personalisation was never the villain. It was just never once on your side.
So we built the other one. Made for you — and made to let you go.
The news, on your terms.
Sources
- Pariser E. (2011). The Filter Bubble: What the Internet Is Hiding from You. Penguin Press. TED talk: Beware online "filter bubbles".
- Sunstein CR. (2017). #Republic: Divided Democracy in the Age of Social Media. Princeton University Press. press.princeton.edu
- Tufekci Z. (2018). YouTube, the Great Radicalizer. The New York Times. nytimes.com