From Anecdote to Accounting: measuring the AI boom

This essay grew out of an independent research project on the economics of artificial intelligence.
As AI capital spending became the central macro-financial question of 2025–26, Allan was struck that the public “bubble” debate was conducted almost entirely by anecdote — useful technology cited against alarming spending, with neither side measuring the same thing.

The piece offers a way to move the AI-investment debate “from anecdote to accounting”.
It argues that “is it a bubble?” is the wrong question to put to the data and replaces it with three quantities, each buildable from public sources: private coverage (is the capital earning enough to pay for itself?), social coverage (is the economy gaining value even where investors are not?), and financing fragility (whose money is at risk, and where in the capital structure it sits).