Essays

Memory, retrieval, and the externalization of knowledge: From Socrates to vector databases

Christian Universalism and the Ethics of Open Source

Christian universalism asserts that grace extends to all. Open source asserts that code should be available to all. Both face the same challenge: sustainability requires shared obligation, not individual sacrifice.

The Architect’s Notebook: System Design Begins With Writing

Writing design documents before code produces 23% faster delivery and 31% fewer defects. Clear architecture emerges from clear prose, not the other way around.

SEC Filing Processing as Data Engineering Microcosm

Processing 47,000 SEC EDGAR filings revealed every core data engineering challenge in miniature: schema evolution, semi-structured extraction, and source-of-truth design.

The SOPs That Survive: What Makes Them Standard

Studying 34 SOPs across 5 organizations found only 26% were followed. Surviving SOPs were practitioner-authored, under 2 pages, included decision points, and updated within 90 days.

The AI Ethics Officer Role Is a Systems Design Problem

AI ethics officers fail when positioned as compliance gatekeepers. The role succeeds when restructured as a cross-functional architecture position embedded in engineering.

Data Privacy Is Infrastructure, Not Policy

Replacing policy documents with 6 engineering controls reduced privacy violations by 91% across AI systems processing 2.4 million records monthly. Privacy must be built, not written.

The Trolley Problem Is the Wrong Framework for AI Ethics

The trolley problem was designed for individual moral agents. AI systems are sociotechnical institutions. Applying the wrong framework prevents the right questions from being asked.

Popper’s Falsifiability and Your A/B Test

A hypothesis that cannot fail is not a hypothesis. Popper's falsifiability criterion reveals why most A/B tests produce confirmation, not knowledge.

Ethics of AI in Healthcare Demands Systems Thinking

Model-focused audits miss 58% of ethical risks in clinical AI. Healthcare AI ethics demands systems thinking across EHR integration, clinician workflow, consent infrastructure, and failure cascading.