The AI build-out is missing its third metric.
We measure how much infrastructure can do, and what it costs to run. We have no measure of whether a facility and the systems that host it can sustain one another.
Grid, watershed, airshed, land, and community — that missing property has a name. This site is its home: the framework, the specification, and the argument, published openly for anyone to use, adapt, and improve.
The AI build-out is missing a third metric
Every dashboard in the AI build-out tracks two families of numbers. Capacity asks: can we build it? Efficiency asks: what does it cost to run? Both are indispensable. Both are measured entirely inside the fence line.
Neither can see a slipped interconnection, a water moratorium, a denied permit, or a community that organizes against the hum — the failures now delaying billions of dollars of infrastructure. Those failures do not happen inside the machine. They happen at the boundary between the machine and the systems that host it.
Infrastructure Health is the degree to which a facility and its host systems — grid, watershed, airshed, land, and community — sustain one another over the asset's operating life. It is a boundary property, it is bidirectional, and it is an engineering objective rather than a virtue: something to be specified, measured, designed for, and traded off like any other requirement.
- Capacity
- Efficiency
- Capacity
- Efficiency
- Infrastructure Health new
The paper that proposes the objective
This paper argues that AI infrastructure now fails at its boundary as often as inside its fence line — a slipped interconnection agreement, a water moratorium, a denied permit, a community that organizes against the hum. None of these are captured by capacity or efficiency, the two metrics every facility already tracks. The paper defines Infrastructure Health as a boundary property: the degree to which a facility and its host systems sustain one another over the asset's operating life. It proposes six measurable dimensions — metabolism, hydration, thermal regulation, immunity, symbiosis, and regeneration — each reported as an evidence-graded indicator rather than averaged into a single score. The paper also states plainly where it is weakest — consent indicators without social-science rigor, normalization classes that still need reference data, thermodynamic claims published with their bounds rather than their best case — and its refusal to collapse six dimensions into one composite number is a choice that deserves to be argued with.
Six dimensions. Measured at the boundary.
A healthy infrastructure behaves more like a living system than a static machine. That is not a metaphor for its own sake — each dimension resolves to something an engineer can specify and measure, at the interface between the facility and its hosts.
The six are reported side by side as a profile — never averaged into a single score. No physician averages hemoglobin with cholesterol, because the average would conceal exactly what matters. Every indicator additionally carries an evidence grade, from measured operational data down to unverified assertion, so that a promise and a meter reading can never wear the same clothes.
Who wrote this, and on what terms
Founding Paper No. 001 is written by Dr. Perry Ho, co-founder of AeroSphereTech, a company that develops closed-loop energy and AI-factory infrastructure. That interest is disclosed here for the same reason it is disclosed throughout the paper: a measurement framework proposed by someone with a stake in what it measures should be read with that in mind, and should be built so that it can be checked.
Three commitments follow from it. The framework text — the definition, the six dimensions, the indicators, the measurement rules, and the evidence grades — is published under Creative Commons Attribution 4.0: anyone may use, adapt, and build on it, for any purpose, including in competition with its author. Nothing in the vocabulary is trademarked or licensed. And should the framework ever mature into a certification, that certification will not be owned by the author's venture — under certification-mark law, the owner of a certification may not supply what it certifies, which makes the independence the paper promises a legal requirement rather than a courtesy.
The paper also states plainly where it is weakest: the consent indicators need social-science rigor the author does not claim, the normalization classes need reference data, and the reference architecture's thermodynamic claims are published with their bounds rather than their best case. A reviewer should be able to reject the author's implementation entirely and still accept the framework.
Review before public release
Review of the founding paper is by invitation. If we have been in touch — or your expertise bears directly on one of the areas below — request a review copy here. Most reviewers read a single section closest to their field; an hour of critical attention is worth more than a general endorsement.
- Symbiosis / consent indicators — proposed without the social-science rigor a specialist would bring.
- Normalization classes — still awaiting real reference-population data.
- Reference architecture — thermodynamic claims published with their bounds, not their best case, and a deliberate refusal to collapse six dimensions into one composite score.
Reviewing does not imply endorsement. Reviewers are free to reject any part of the framework — including all of it — and their feedback will be credited accordingly in the first public edition.
Review copies are provided solely for invited feedback before publication. Please do not redistribute the manuscript without permission.
From connected systems to systems that understand themselves
Infrastructure has evolved from passive assets to connected systems.
The next evolution is infrastructure that can understand its own condition, adapt to changing environments, and sustain healthy operation throughout its lifetime. This initiative explores that future.