The publication
About
forck.live is a live timeline of what major AI labs announce. It is not an aggregator: every item traces to a first-party source, and anything we cannot trace stays off the page.
Who may lead
Only a company's own newsroom can lead, sit in a section, or appear in Latest. Named press may be fetched, but a reported item reaches a reader only attached under the official story it corroborates. How many outlets wrote about an announcement can change which lab post leads. It cannot put the coverage itself on the page.
What coverage means
When a story lists outlets, each row is a masthead, that outlet's own headline, and a URL off this site. The count is how many outlets covered the announcement. We do not print an aggregator's phrasing as a headline, and we do not invent a headline from a hostname. An outlet we can name but have not yet titled is still listed, as a named link.
What we will not do
A missing date is missing. We do not guess one from a slug or from the minute we first saw the page. A missing category is omitted. A title is the publisher's words. A product changelog that posts eight times a day cannot outrank a lab announcement by arriving more often. The rail — as it lands — stays chronological, because that is the one surface that should answer what is newest.
The source
Every story page's primary action is the original. This site is a record of the announcement, not a second copy of it. RSS carries the same confirmed first-party items, each linking here and to the lab. Every company page has its own feed, and a topic does too — model releases only, say — so a narrower question can be a subscription instead of a visit.
Who to write
Abhijeet Wankhade runs this. Legal and takedown requests go to abhijeet@wankha.de. Conversation is on X: @sokratus. What is logged is on Privacy.