Long-form pieces on the decisions that shape capital-intensive businesses and operating assets. Published as we have something worth saying.
The Mythos argument was that cost of offence had collapsed. The volume of offence had already risen 13-fold before that collapse registered. Institutional capital is still pricing the old world.
A $20 billion cyber insurance market and a $1.5 trillion utilities sector were both calibrated against the scarcity of human offensive expertise. That scarcity no longer applies.
Most capital stacks at opening assume operations will work. Operations rarely work at opening. The bridge between what was built and what operates is where capital gets eroded or preserved.
Frontier AI crossed a judgement threshold. Institutional procurement runs eighteen to twenty-four months behind. A principal with liquid capital reprices assets inside that lag, before the comp set catches up.
The question a family office should ask about an AI-native platform is not whether it works today. The question is what it looks like under procurement, handover, and ten thousand residents.
A capital-intensive asset that opens with three to five years of contracted revenue is underwritten on different terms. That work starts eighteen months before opening day.
Fujian commissioned on 5 November; the fourth Matsu-Taiwan cable incident since 2023 landed nine months earlier. The European grey-zone frame runs in Asia with a shorter stock window and a different response posture.
Four airport closures across three NATO capitals in three weeks and a 13-fold rise in GPS interference over nine months. The operating-asset principal should move ahead of the consensus.
On 6 April 2025 the UK non-dom regime ended. Ninety days in, London prime has been trading against a buyer pool the pre-reform market was not underwritten to reach. Notes from the audit seat.