What problem does it solve? Texts about AI often collapse distinct concepts like visibility, relevance, authority, legitimacy, truth, and power into a single vague claim, producing sloppy analysis of how AI actually participates in information processes. This Skill provides a rigorous analytical framework that keeps these categories separate and reconstructs the concrete relations through which AI-mediated information becomes consequential. ## Core Features & Use Cases - Conceptual Distinction Enforcement: Preserves the analytical boundaries among mediation, selection, visibility, relevance, authority, legitimacy, truth, auditability, and power, flagging common collapses such as "visibility = truth" or "automation = authority". - Structured Analysis Protocol: Applies a 42-section framework covering institutional embedding, delegated and perceived authority, feedback loops, material infrastructure, dependency, and counterfactual tests (e.g., remove AI, remove institutional uptake). - Standardized Output Format: Produces findings with claim, evidence, mediation chain, classification (A-E), confidence level, and recommendation for each analyzed passage. - Use Case: When reviewing a policy paper claiming that an AI ranking system "controls public opinion," use this Skill to decompose the claim into mediation, selection, visibility, and institutional uptake relations, then assess where actual power asymmetries and dependencies lie. ## Quick Start Analyze this text about AI-driven content recommendation using the power-mediation-authority framework and report findings with the structured output protocol.
