essays
Long-form writeups that put an idea in conversation with the literature it belongs to.
These are research essays: extended, tutorial-like writeups that take one idea, situate it in the literature it belongs to, and carry it far enough to be worth disagreeing with.
Each is a single self-contained page. Read it start to finish and you should learn the field around the idea, not just the idea. Figures are drawn to be read. Sources are cited in full, and every claim that carries weight rests on a paper whose full text was actually read.
How these are written
I write them with an agent, following a fixed procedure I have packaged as an open skill: idea-to-essay. Given an idea, it digests it into a thesis you could argue with, sweeps the literature and downloads verified full texts, produces the numbers and figures from code, and only then writes.
Three rules do most of the work. Nothing is cited from an abstract or from memory, so a claim either has a full text behind it or it is flagged as new. Corrections count as output, so when a paper overturns something I believed, that goes in the writeup rather than quietly out of it. And every number traces to a script that runs. The procedure, the tooling and the house style are all in the repository, under MIT, if you want to run it yourself.
The essays
Three Schools, Three Shocks
A history of synthetic biology, and why its third wave stalled
Physics, system, and industry have each asked their own question of life. Three times a discovery let all three ask it at once. The third time stalled, and the reason is a missing model class. Twenty-five sections, nine figures, 107 sources read in full.
The Biomachine Perspective
A tutorial on reading cells as machines, and what that buys you
A cell is a machine. Fix the function, compare the implementations, and let the difference in physical constraints explain the difference in architecture. Twenty-one sections, ten figures, worked examples, and an interactive demo. This is the perspective the lab is built on.
Each essay links into a local literature/ archive of full-text
PDFs. That archive is not redistributed here, so those links are inert on the web. Every citation is
complete enough to find the source.