Field Notes
Field Notes are where I publish work that is still close to the source.
They are less polished than essays, but still relevant. Some are observations, experiments, reactions, or partial conclusions. Most come directly out of real projects, research, and day-to-day use of AI.
How the Ideal 2026 Political Platform Got Built
Our Political Platforms essay shows the finished argument. This note keeps the scaffolding visible. It explains how the platform was built, why weak planks were discarded, where AI helped, and why the final test still had to be human.
Five Figures on Fertility
The 2025 U.S. fertility figures are easy to summarize and harder to interpret. This work looks at delayed childbearing, long-run fertility decline, sperm-count deterioration, and the narrowing gap between births and deaths, then asks what these patterns may signal when placed next to South Korea, rising child costs, and a changing labor market.
Before Efficiency Shows Up: A Note from Early Industrialization
While working on AI data center infrastructure, I looked for a historical parallel in Britain’s Industrial Revolution. The data suggest a long phase where energy use grew faster than output, with efficiency improvements showing up later than the popular story implies.
How I stopped starting at the beginning
While researching data centers in space, I used a fresh AI prompt with no project history to surface the obvious questions I had started to miss.