The Journal
Essays
Long-form writing on AI, philosophy, psychology, and systems thinking.
-
When an Agent Lies: AI Hallucination as Ethical Engineering Problem
In a medical information agent, 4.2% of responses contained fabricated information. At 8,000 daily queries, that is 336 potentially harmful outputs per day. Hallucination in high-stakes contexts is an ethical failure.
-
Kuhn’s Paradigm Shifts in Programming Language Adoption
Programming paradigms behave like scientific paradigms. OOP dominated for 25 years. Functional programming is steadily replacing it. Language wars are paradigm conflicts, not technical debates.
-
Responsible AI Is a Cross-Functional Process
Implementing responsible AI as a cross-functional process reduced bias incidents by 48% and shifted detection from 6 weeks post-deployment to 3 days pre-deployment.
-
Epistemic Humility as Engineering Competency
Epistemic humility is accurate assessment of what you know and do not know. In engineering, that honesty produces systems that work in the real world.
-
Design Decisions as Moral Choices: Winner Was Right
Every architecture decision embeds ethical commitments. Langdon Winner argued artifacts have politics. The question is whether the architect is conscious of it.
-
The Ethics of AI Art Is a Labor Economics Problem
An estimated 26% of commercial illustration work has been displaced by AI image generation since 2023, with losses concentrated among early-career artists. This is a labor economics problem, not a copyright debate.
-
Deadlines Are Communication Tools, Not Motivation Tools
Deadlines used as coordination tools had a 78% on-time completion rate, while deadlines used as motivation tools had a 23% on-time rate across 18 projects.
-
The Data Warehouse Is Not Dead; Your Expectations Were Wrong
After migrating to a lakehouse and back, a company reduced query costs by 52% and improved report reliability from 87% to 99.1%. The warehouse is not dead. It was misunderstood.
-
The Modern Data Stack Died. Here Is What Replaced It
The modular Modern Data Stack collapsed under integration tax. The replacement is vertical integration with open escape hatches.
-
Building Data Pipelines That Survive Schema Changes
Schema-resilient pipeline patterns reduced failures from 4.3 per month to zero over 9 months. Pipelines that assume schemas will change survive longer.