Mission

The Distraction Index publishes a weekly, frozen record of U.S. political events scored on two dimensions: constitutional damage (A-score) and media hype / manufactured distraction (B-score). The goal is to help citizens distinguish between events that cause real democratic harm and events that are engineered to dominate attention.

Every week runs Sunday through Saturday. Once a week closes, its scores freeze permanently. Post-freeze corrections are append-only and publicly documented on the corrections page.

How It Works

Articles are ingested from multiple news sources every four hours. AI clustering groups articles into distinct events, which are then scored by AI using a transparent, versioned algorithm. Every scoring decision can be audited on the methodology page.

Creator

The Distraction Index is an independent project by Steve Harlow, an AI and civic tech advocate. The source code is publicly available on GitHub.

Principles

• Transparency: Every algorithm weight, modifier, and formula is documented publicly. No black boxes.
• Immutability: Frozen scores are never silently changed. Corrections are append-only and timestamped.
• Independence: No political affiliation, no advertising, no editorial influence. The algorithm scores events; humans review edge cases.
• Open Source: The full codebase is available for inspection, critique, and contribution.