I study what happens when systems meet reality.
A policy can be sound on paper and fail in practice. A technology can work perfectly in one country and become almost useless in another. An institution can have the right rules, the right people, and the right tools — and still produce the wrong outcome because it misunderstood the environment around it.
That gap interests me.
I have spent much of my life moving between two operating worlds: formal American institutions and East African systems of trust, improvisation, social coordination, and constraint. The longer I have worked across both, the less interested I have become in asking whether a system is sophisticated.
I want to know whether it works.
The central question
What has to be true in the real world for this system to produce the outcome it claims to produce?
That question has followed me through technology, institutional analysis, public systems, markets, AI, infrastructure, and the everyday mechanisms by which people coordinate when the formal system is incomplete.
It has also produced a few convictions I keep returning to.
A design does not operate in isolation. Infrastructure, incentives, culture, trust, language, regulation, history, and human behavior all become part of the machinery whether the designer accounted for them or not.
People cooperate with systems they consider understandable, credible, and sufficiently fair. When legitimacy erodes, the cost of enforcement, supervision, workaround, and resistance rises. Trust is not separate from performance. It helps produce it.
Some of the most consequential failures happen between layers: policy and implementation, headquarters and field conditions, technology and adoption, formal rules and informal behavior. The ability to translate between those worlds often determines whether a good idea survives contact with reality.
I am usually less interested in the symptom than in the mechanism generating it. What constraint, incentive, bottleneck, feedback loop, or institutional assumption keeps reproducing the outcome?
My work sits at that intersection. I am a linguist, analyst, technologist, writer, and systems thinker, but those labels describe the tools more than the job.
The recurring job is synthesis: take a complicated environment, find the underlying mechanism, separate signal from noise, and make the system legible enough that someone can act on it.
What I write about
The essays on this site examine institutions under pressure: how power moves, how trust is built and lost, how systems adapt, why apparently irrational behavior can become rational from inside a particular incentive structure, and what happens when ideas designed in one context are transplanted into another.
That means the subject can move from artificial intelligence to public institutions, from Kenya to the United States, from markets to infrastructure, or from bureaucracy to everyday human behavior. The subjects change. The question underneath them usually does not.
Currently Publishing
The Transition State Arc follows institutions as they move from apparent stability into stress, misread their changing environment, and eventually reorganize around a new reality. Read the sequence →
Built, Not Just Written
Some ideas are better tested in code than in prose. Selected public projects:
View all repositories on GitHub →
Why this site exists
I am interested in ideas that survive contact with reality.
That means writing should eventually do something: improve a decision, clarify a mechanism, expose a hidden assumption, change the way a problem is framed, produce an experiment, or make another person see a system they had previously only experienced.
I do not expect every essay to provide an answer. Some are attempts to identify the right question.
But I want the reader to leave with a better map than the one they arrived with.
If you are working on a difficult system, institution, technology, or problem where the formal model and the lived reality do not quite match, I am always interested in a serious conversation.
Discussion