Automation, stated plainly
Every story in this paper is researched, written and published by a software agent. No editor types the text. A human sets the rules, reviews the result and answers for it. We say this here because a reader deserves to know how a text came about.
What is not automated: the source list, the editorial rules and the decision to publish this paper at all. Those are human decisions, and they are the ones that matter.
Which sources are allowed
The core of every edition comes from research institutions and from the bodies that set the rules. Journalistic sources are permitted only where a dated market event has to be evidenced, and never for more than two stories per edition. Anything outside the register below is rejected before publication – automatically, by the software.
What never enters this paper
- Security holes, jailbreaks and leaks reported for the thrill of it
- Vendor announcements without an independent source
- Aggregators, forums and platform posts
- Anything without a verifiable publication date
How the graphics are made
This paper uses two kinds of image, kept strictly apart. First the graphics: they are drawn from the figures of the story they belong to – the same numbers you read in the text, from the same source. Second the photographs: they are generated with artificial intelligence and marked as such, visibly on the image and machine-readably in the file. A photograph never shows the event the story is about, and never a recognisable real person – it depicts the setting in which the story takes place. Wherever an image would have to serve as evidence, a graphic stands there instead of a photograph. The evidence itself is always in the text and in the linked original source.
Mistakes
An automated paper makes mistakes, just as a staffed one does. Every story names its source and links to it, so any claim can be checked at the origin. If you find an error, write to us and it will be corrected or the story withdrawn.
