THE TELL

How we work

The news about AI and money that actually changes something — in your language, the same day. Our newsroom is run by AI. We say so, and we show the rules it works by.

Why us, and not the chatbot you already use

When you ask an ordinary model what is happening in crypto, it goes looking and finds whatever surfaces first. It did not read that story this morning. It does not remember what happened on the subject a month ago. And it has no opinion — it is built not to argue, so it gives you the average answer.

We read every day, not when asked

Around 350 stories a day, around the clock, no weekends. By the time you open the page it has all been read, checked against the source and explained.

We remember

When a story runs into its third month we know every step of it and show you where it is heading. An ordinary model starts from a blank page every time.

We have a character

Our newsroom is tuned to look at AI, crypto and money a particular way: the person first, the market second. Not “analysts suggest” but “here is what this means for you”.

We say what we do not know

That is rare — in machines and in people alike.

51sources watched around the clock
~350stories reviewed every day
2–6published — the rest changed nothing
7languages, in the same minute
What we take on

We do not try to cover everything — it cannot be done and it is not useful. Out of roughly 350 stories a day we take the ones that change the rules: after which something works differently than it did yesterday. And the ones that start something — where you can already see what will grow out of it in six months.

The rest is somebody’s press release, somebody’s quarterly report, somebody’s attempt to sell you a course.

If we stayed quiet, nothing needed changing.

How we see it
  • Technology matters not in itself but for what it changes in people’s lives.
  • Behind every number stands a person it touches. We put the same question to every story: who gains, and who pays for it.
  • New things are rarely sudden — almost always they are the next step of something that started earlier. That is why we run stories for months instead of writing each one from scratch.
  • We do not predict prices. We tell you what follows.
  • No story is worth your panic.
Seven stages, and each can send a story back
The watch

51 sources without a pause — lab blogs, the technology press, fraud alerts from the FBI and the FTC, research archives, and the forums where people describe what happened to them. Primary sources every five minutes.

The desk

Reads every headline that comes in — all of the day’s stories, not a sample. Held to one rule above the others: we write for the 99%, not the 1%. Marks what deserves a full read.

The reading

Opens the marked stories and reads the text, not the headline. Scores each from 1 to 10: does this change anything, who does it touch, does the official line match the evidence. Rejects roughly nine in ten, each with a reason on the record. If a site is closed to us, it reads the same story where it is open — we do not retell what we have not read.

The writing

Works to more than five hundred lines of editorial rules: the voice, the structure, the banned devices, and the standing demand to explain a hard thing in plain words.

The fact check

Gets the source and our draft side by side. Hunts for anything the source does not support: a figure that shifted, a claim stronger than the evidence, a motive we invented, an aggregator standing in for a primary source. Sends pieces back — and the ones it sends back are listed publicly.

The edit

Decides whether a text is fit for a reader. Sends it back if it opens with a list of funds, if the jargon is explained with more jargon, if nothing in it touches an ordinary life, or if there is no single line worth repeating to a friend.

The languages

Six languages, each with its own instructions. Not a word-for-word conversion: a headline that works in English rarely works in Turkish or Vietnamese. Each language gets the same story, written to land in that language.

What we are not

We have no reporters. We telephone no one, we run no interviews, we are not at the scene. Others go and get the story; we explain it. Those are two different trades, and we do the second one — which begins where most coverage stops: reading all of it, checking it against the source, and saying plainly what it means for you.

Who answers for it

The rules are set by a person, who answers for the mistakes and can stop any publication at any moment. Individual stories are not read by a human before they go out: they are checked by the system, against the rules published on this page. The machines do the work. The responsibility stays human.

Victor P  ·  Write to the editor

Check us

Two pages exist so that none of the above has to be taken on trust: what we turned down and why, and whether what we said would happen actually happened.

Everyone reports what happened

We send what it means — the part that gets left out: who it hits, what breaks next, and why the obvious reading is wrong. One letter, only when something actually shifts.

No spam. Leave in one click.

Prefer to follow instead? Telegram X