Large language models, briefly
What they are, what they are not, and why "it's just autocomplete" is both true and unhelpful.
Read explainer →Section 04 — Story & Explainers
Things worth understanding before the next news cycle.
Concepts, terms, and frequently-asked questions translated into language a working journalist or a curious student can use.
Retrieval-augmented generation
Why "the model looked it up" is doing a lot of work in that sentence.
Read explainer →Attribution, consent, and synthetic media
Three questions every newsroom should answer before publishing AI-assisted imagery.
Read explainer →Prompting for journalists
A short guide to writing instructions the model can follow — and a longer warning about over-trusting the result.
Read explainer →When to disclose AI use
A working framework for transparency: what counts as 'material' use.
Read explainer →Training data, in one diagram
Where models learn from, why it matters, and the licensing questions still unresolved.
Read explainer →Frequently asked
Questions we keep getting.
- Is AI replacing journalists?
- No. The microsite's editorial position is that AI is a tool for journalism. Humans remain the primary editorial decision-makers, and AI use is disclosed when material.
- Can I reuse content from this microsite?
- Most explainers and OER materials are released under open licences. Each item carries its own licence notice — check before redistributing.
- How do you decide what to publish?
- Content must be based on real practice, learning, or relevant references. We stay critical of both the opportunities and the impacts of AI on media and the public.