Research
Research & Discovery Reports
Discovery Reports trace a single business term across documents, systems, and AI agents. Reference articles explain the model behind the work.
Semantic Debt Discovery Reports
One term. Traced end to end.
Each report takes one high-value business term and shows where its meaning diverges across documents, systems, and AI agents — and what canonical governance does about it. New reports added on a rolling basis.
- Report #001Available2026-06-15
How Many Meanings Does Coverage Have?
A single insurance term, audited across documents, systems and AI agents.
The word 'coverage' carries at least five legitimate, context-bound meanings inside a single insurer — claims handler, underwriter, reinsurer, compliance officer, AI assistant. This report walks through each, shows where they collide, and explains how Semantic Governance preserves the ambiguity while eliminating the drift.
Reference articles
Short. Citable. One idea each.
Why AI Amplifies Semantic Drift
Large language models infer meaning from training distributions, not from the organizations that deploy them. Without semantic governance, every AI deployment compounds existing drift.
Published 2026-06-15
Why Glossaries Fail
Business glossaries capture a snapshot of meaning at one moment. They have no concept of context, no history, and no mechanism to distinguish ambiguity from contradiction.
Published 2026-06-15
Ambiguity vs Semantic Drift
Ambiguity is a feature of natural language. Drift is a failure of governance. Conflating the two leads to brittle systems and unnecessary review work.
Published 2026-06-15
Governance vs Documentation
Documentation records what someone wrote down. Governance records who decided what, when, and why. Only governance produces an audit trail.
Published 2026-06-15
Meaning as Enterprise Infrastructure
Identity, payments, and data have all moved from local concerns to shared infrastructure. Meaning is following the same path.
Published 2026-06-15