Essays and notes
Source: Substack, cross-posted to Medium. Date: Feb to Jun 2026; one note from Oct 2024. Scope: Titles and dates as published. Links go to the originals. A proposal is filed as a proposal.
The essays are about what AI is doing to work, to markets and to responsibility. I publish them on Substack, with a copy on Medium. The dates here are the real ones, and the labels are plain: essays are essays, notes are notes. Published
| Title | About | Date |
|---|---|---|
| 2026 | ||
| The First RecallAlso on Medium | When the U.S. Government forced an AI model offline worldwide. | |
| Apple’s $1 Billion AdmissionAlso on Medium | When the world’s most valuable company gives up on building AI. | |
| The Trillion-Dollar Safety PremiumAlso on Medium | How the Anthropic–Pentagon standoff created $585 billion in market value. | |
| The Agent as Economic ActorAlso on Medium | When AI spends money, who owns the consequences? | |
| The Liability Vacuum in Agentic AIAlso on Medium | Why tort law breaks when the agent has no principal. | |
| The credible commitment problem in AI safety: lessons from the Anthropic–Pentagon standoff | A game-theory reading of a refusal that accepted blacklisting rather than capitulate. | |
| The $53 Billion Blind Spot: Why AI Agents Will Eat Travel Distribution From the Inside Out | And why the companies that control the pipes, not the interfaces, will win the next decade. | |
| They’re Not Coming For Your Job. They’re Coming For Your Purpose. | OpenClaw launched a mirror, and what stared back was uncomfortable. | |
| The Encoding Gap: How AI Is Reshaping the Architecture of Human Expertise | What happens to human knowledge when the machines that augment it begin to outperform the experts who built it? | |
| 2024 | ||
| Research proposal: advancing Random Forests for high-dimensional, imbalanced data in biomedical applicationsA note, on Medium | A proposal, not a paper: an enhanced Random Forest approach for imbalanced biomedical data. | |