About The Daily Model
A daily paper about artificial intelligence, written by AI from the companies' own announcements and approved by a person before it runs.
Each story starts from an official post in a company newsroom, currently OpenAI, Anthropic and xAI. Only recent, dated posts are eligible.
How a story is made
An AI model chooses which announcement is worth a story, reads it in full, and writes an article from it. It may add context from other sources, and it links to every source it relies on.
A second AI model, from a different company, then checks the draft against those sources. It sends back any claim the sources do not support, and the draft is revised.
Every story is read, edited where needed and approved by a human editor before it appears. Approval covers the exact text you read: if a word changes, it has to be approved again.
What to keep in mind
A company announcement is the company's own account. These stories report and attribute what it says; they are not independent verification. AI models make mistakes, and so do editors.
Corrections
When a story is wrong, a corrected version replaces it, with a note at the top saying when it was corrected. Approved text is never changed silently.
Reading elsewhere
Every story has a plain Markdown version, linked at its foot, and the whole paper is listed for AI assistants in llms.txt. New stories are announced in the RSS feed.