/** * Add Human Vetting Notice to Architectural Boundaries Banner * Demonstrates AI-curated content requires extra scrutiny */ const { MongoClient } = require('mongodb'); const DEV_URI = process.env.MONGODB_URI || 'mongodb://localhost:27017'; const DEV_DB = process.env.MONGODB_DB || 'tractatus_dev'; const SLUG = 'why-ai-safety-requires-architectural-boundaries-not-just-training'; const ENHANCED_BANNER = `
Human Vetting Process: This AI-curated blog was initially withheld from production deployment pending human review. After revision to meet inst_017 compliance, it was approved for publication.
The original content contained absolute assurance language ("guarantees") that violated inst_017. The content has been revised to use evidence-based language:
This demonstrates governance working end-to-end. AI-curated content triggers additional scrutiny. When violations are detected, content is withheld, revised, and approved by human reviewers before publication. Rather than silent editing, we document the entire process transparently.
Note: The revised language is more precise and accurate. "Deterministic constraints" better describes what architecture provides than "guarantees," which implies perfection.