Essays on AEC technology, construction operations,
and the long work of building at every level of the built world.
The biggest engineering firms are slow on AI, stuck with compliant but sub-tier tools. They hold every card — clients, experts, time. But AI deflates the billable hour the whole business is priced on. The dilemma, and the opening.
AI has changed the cost of judgment in the built world. Where Hesiod sees the opportunity — the regulatory layer, civil and geotechnical engineering, the contractor economy — and why domain expertise is the unlock.
Frontier models are good enough for real civil engineering and permitting work. If everyone has the model, the model isn't the moat — domain expertise is. Why this is our moment.
Permitting, compliance, and regulation are the slowest, most expensive, least-loved layer of the built environment — and the highest-leverage to rebuild with AI. How we build it from both sides, with EntitleHQ and Munivias.
The Hesiod thesis in long form. Why AEC software built from the outside fails, and what it means to be embedded at every level — permit desk, job site, back office.
How running real contracting businesses alongside software development creates compounding advantages — and why that feedback loop is structurally unavailable to outside-in builders.