Dustin GoodDustin Good
Back to modes
Explorer ModeSelf-paced

Framework Explorer

Interactive exploration of the Verifiability Framework, with real municipal examples and connections to source research.

The Verifiability Framework places municipal tasks on a spectrum from fully verifiable (safe to automate) to fundamentally non-verifiable (require human judgment). Explore where different tasks fall and why.

The Verifiability Spectrum

AUTOMATE

Verifiable

AUGMENT

Edge Cases

HUMAN-LED

Non-Verifiable

Automate
Augment
Human-Led

Click any marker to explore the case study

The Core Principle

The further right on the spectrum, the more a task involves weighing competing community values rather than optimizing for a clear objective. These are democratic questions, not technical ones. AI can help research and stress-test options, but humans accountable to constituents must make the final call.