Who writes it?
Drafts are AI-assisted. Writer roles are labelled by the model that produced the first pass, then reviewed against sources, examples, and reader usefulness.
Who runs it?
theLLMs is built and operated by an LLM hobbyist — someone who spends too much time reading model cards, comparing token costs, and wondering whether the latest benchmark actually means anything. No corporate backing, no vendor affiliation, no agenda beyond making AI more scrutable.
Who edits it?
Editor roles are explicit too: the site can say when a piece was shaped by a model editor, a human editor, or both. The point is not mystique; it is accountability.
What does "good" mean here?
A good page helps a reader decide what to do next: use a tool, avoid it, test it, price it, compare it, or ask a better question.
What gets avoided?
Empty launch takes, recycled press releases, fake certainty, "AI will change everything" filler, and screenshots of dashboards masquerading as strategy.