About

An AI economic think tank, reporting from inside the shift.

Divergent Compute measures the distance between what markets price and what the math allows — across the AI build-out, and inside our own operation.

The dual mandate

AI is both our engine and our subject.

We cannot credibly analyze the economics of AI from the sidelines. So we run an AI-native research operation and instrument it — making the firm itself a live data point on the productivity, labor, and capital shifts it studies.

We don't just speculate on the future of the economy. We use the tools shaping it.

What we measure

Two clocks, one race

The financing runway — can the spending survive? Capex versus demand, circular financing, depreciation that hasn't been honestly marked.
The productivity lag — will the payoff arrive? The Solow lag between adoption and measurable output, read in real time from inside the frontier.
The race between them — does productivity arrive before the financing that funded it has to be repriced? That single question is our edge.

How we work

The discipline

Credibility is the only thing we sell, so we engineer for it. Verified figures trace to primary filings. Attributed estimates are named as such. Unverified figures are flagged or removed — never dressed up. Every thesis pillar carries an explicit falsifier, and we revise in public.

The work runs on three surfaces that feed each other: a publication that reports the divergence, research that measures it, and consulting that applies the instruments inside your operation.