AI Review
Structured analysis of an external document by an AI model. Prompt and metadata are documented. Every analysis is traceable and in principle reproducible.
Formats & Methods
How we work.
Every article follows a defined format. Here we document which analysis and article types we use — so you know what to expect.
Analysis Types
How we investigate topics — transparent, reproducible, with multiple perspectives.
Multi-Model Comparison
The same document is independently analysed by multiple models. Comparison reveals blind spots of individual models.
Media Review
Analysis of external media — TV shows, YouTube videos, podcasts. Transcript-based, then multi-model analysis with fact-checking.
Rhetoric Analysis
Sports commentary for rhetoric — with an open rulebook. Four-level model: structure, Toulmin argumentation, rhetorical techniques, fact layer.
View rulebook →Data Extraction
Systematic extraction of structured data from documents — tables, lists, key figures. Foundation for fact-checking and visualisation.
Replication
Repeating an analysis with a different model. Same prompt, different tool. Tests the robustness of results.
Opinion Formats
Context, commentary, counter-proposals — the voices of our editorial team.
Editorial
Joint editorial assessment by the team on a topic or publication. All four voices contribute.
It can be better
Constructive counter-proposal derived from an analysis. Not just criticising — but showing how it could be done differently.
Commentary
Individual voice of a team member on a current topic. Personal, pointed, transparently labelled.
Publications
Completed documents — citable, versioned, Open Access.
Position Paper
Analysis and position in one document. Numbered (PP-001, PP-002, ...), versioned, with evidence matrix and policy demands.
Practical Guide
Handbook for a specific audience. Legally and practically sound. For reference and application.
Dataset
Accompanying data for an analysis — Excel, CSV, JSON. For verification and further research.
Series & Participation
Formats that appear across all sections — and how you can contribute.
AI Lab
We investigate our own methods: what works, what doesn't, and why? Prompt experiments, model comparisons on methodology, bias checks. Editorial self-reflection.
Editorial Picks
Curated reading tips and links — what our editorial team reads and recommends. Appears as a tag across all sections.
Letter to the Editor
Have something to say? We publish selected letters — agreement, disagreement, additions. All welcome.
Write to us →