About

How this project connects ICESCR advocacy, AI economics, and evidence-based analysis.

What Unratified Does

In 1977, the United States signed the International Covenant on Economic, Social and Cultural Rights — a treaty protecting the right to work, to health, to education, and to benefit from scientific progress. The Senate never ratified it. For decades, that gap carried theoretical consequences. Now, with AI transforming the economy at a pace that creates winners and losers faster than any previous technology, that gap carries practical ones.

Unratified connects AI's economic transformation to the human rights treaty the United States signed but never kept. The site provides analysis, education, and advocacy tools so that visitors can understand what the ICESCR protects, how AI-driven economic change makes ratification urgent, and what concrete steps move toward ratification.

Human Rights; Nothing More, Nothing Less.

Every element of this analysis — quality floors, safety net reconstruction, litigation activation, educational reform — represents implementation of rights 173 nations already committed to. Nothing here asks for anything beyond what the United States signed in 1977.

The Analysis

The analytical engine behind this site uses differential diagnosis — the same methodology physicians use to evaluate competing explanations for symptoms. Seven hypotheses about AI's economic impact get tested against empirical evidence, scored with a consensus-or-pragmatism discriminator, and evaluated through five analytical orders.

Order 0 — Direct Effects
AI removes constraints on software labor, triggering productivity gains in automatable tasks
Order 1 — Jevons Rebound
Cheaper production drives demand explosion, creating new work faster than AI eliminates old work
Order 2 — Bottleneck Migration
Constraints shift from technical capability to human judgment, specification, and curation
Order 3 — Institutional Response
Regulation, education, and social structures adapt — or fail to adapt — to new bottleneck locations
Order 4 — Productive Exhaustion
Each constraint-removal cycle shifts scarcity toward increasingly human capacities: values, meaning, dignity

The Composite A model (H2+H3+H4+H7 mod H6) emerged from this analysis: AI functions as narrow superintelligence for software labor, bounded by migrating bottlenecks, distributed unevenly across adopters and non-adopters. The resulting economic bifurcation maps directly onto the rights the ICESCR protects.

A fair witness principle governs the entire analysis: observe without interpretation, distinguish observation from inference, present all arguments faithfully — including arguments against ratification.

The Five Lenses

Content on this site adapts to five audience personas. Each lens reshapes the same underlying analysis for a different reader, adjusting emphasis, vocabulary, and presentation depth.

Voter
Personal impact framing, action-oriented, plain language at approximately Grade 8 reading level. The default lens — because advocacy starts with individuals.
Politician
Policy brief format, legislative pathways, committee references, talking points at approximately Grade 10 reading level. Designed for legislative staff who evaluate and brief their senators.
Developer
Data-forward, dense tables, technical methodology, monospace formatting at approximately Grade 12 reading level. The entry point from the Observatory.
Educator
Learning objectives, discussion prompts, downloadable materials at approximately Grade 12 reading level. Resources for bringing ICESCR and AI economics into the classroom.
Researcher
Structured abstracts, full citations, methodology documentation at Grade 16+ reading level. Everything needed to evaluate, replicate, or extend the analysis.

The lens system uses CSS visibility rules — not different content. Every reader sees the same analysis; the lens determines which sections receive emphasis and which tuck behind progressive disclosure. Switch lenses anytime using the toggle in the navigation bar.

Unfamiliar terms link to the Glossary, which provides definitions for 50 project-specific terms across 8 categories.

The AI Meta-Layer

An AI — Claude, built by Anthropic — produced the core analysis on this site. An AI analyzing AI's own economic impact on human rights creates a recursive meta-layer that deserves transparent acknowledgment.

This project treats that meta-layer as a feature, not a bug. The analysis invites scrutiny of both its conclusions and its analyzer. Every claim links to external evidence. Every inference gets flagged as inference. The differential diagnosis methodology documents its scoring, its eliminated hypotheses, and its confidence degradation — so readers can evaluate the reasoning independently of who (or what) produced it.

The companion project, the Human Rights Observatory, extends this transparency by evaluating how the tech community discusses human rights — applying the same evidence-first methodology to 803+ Hacker News stories scored against the 30 articles and Preamble of the Universal Declaration of Human Rights.

Safety Quotient Lab

Unratified operates as a named project of Safety Quotient Lab — an organization that measures and improves psychological safety.

Safety Quotient Lab develops the Psychoemotional Safety Quotient (PSQ), a measurement instrument for psychological safety. The PSQ evaluation of the Universal Declaration of Human Rights produced the Dignity Quotient — a framework measuring how effectively legal systems operationalize dignity.

That measurement revealed a structural gap: the U.S. legal framework protects the threat-reduction half of dignity (civil and political rights, via the ICCPR) but leaves the resilience-building half (economic, social, and cultural rights, via the ICESCR) without binding legal protection. Unratified demonstrates what that gap means in practice, especially as AI reshapes the economy.

The parent organization provides the measurement instrument. This project provides the finding.

Values and Voice

Five Organizational Values

  1. Intellectual Honesty — Present uncomfortable findings honestly. The ratification counterfactual showed that ICESCR provides tools, not solutions, operating on a 10–20 year timeline through litigation. The site says so. Overclaiming undermines credibility with every audience.
  2. Dignity as Axiom — Dignity — the UDHR's and ICESCR's foundational concept — drives every decision. The Dignity Quotient measures how effectively legal frameworks operationalize dignity. Unratified demonstrates what happens when the measurement shows a gap.
  3. Epistemic Quality — Every claim links to evidence. Every inference gets flagged. Confidence levels get stated. The differential diagnosis methodology — with its discriminator, its eliminated hypotheses, its confidence degradation map — embodies this value.
  4. Accessible Rigor — Rigor without accessibility serves only academics. Accessibility without rigor serves nobody. Progressive disclosure achieves both: summary cards for quick understanding, full analysis for deep engagement, methodology documentation for peer review.
  5. Human Rights; Nothing More, Nothing Less. — Every element represents implementation of rights 173 nations already committed to. This value constrains scope: the site advocates for ratification of an existing treaty, not for novel political programs.

Editorial Standards

All user-facing copy on this site follows E-prime — a constraint that eliminates all forms of "to be" (including "is," "are," "was," "were"). E-prime forces active, precise language that conveys agency and avoids hidden assumptions. "The United States signed the ICESCR in 1977" carries more force than "the ICESCR was signed by the United States."

The fair witness editorial standard governs content across every section: observe without interpretation, distinguish observation from inference, use precise language that avoids assumptions, describe actions and behaviors rather than motivations.

Tools, not solutions. Ratification provides legal mechanisms — tools through which courts, over time, bend economic transformation toward rights realization. The site never promises that ratification fixes everything. It demonstrates that ratification provides the instruments through which progress happens.

Licensing and Contribution

Content
Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0) — Share and adapt freely with attribution. Educators, researchers, and advocates can use all content without permission barriers.
Code
Apache License 2.0 — Open source. Fork, modify, and redistribute with patent protection.

Contribute

The project lives on GitHub at safety-quotient-lab/unratified. Contributions — factual corrections, accessibility improvements, translations, and feature proposals — follow the guidelines in CONTRIBUTING.md.

Machine-Readable Endpoints

Live Evidence

The Human Rights Observatory evaluates 803+ Hacker News stories against the Universal Declaration of Human Rights. It provides the empirical foundation for understanding how the tech community discusses — and often overlooks — economic, social, and cultural rights.

Explore the Observatory ↗