Norwegian pension capital extending uranium positions. Emirati sovereign capital deploying into hyperscale compute. Singaporean state capital building data centre partnerships. Brookfield committing one hundred billion dollars to AI infrastructure. KKR launching Helix. Blackstone bundling QTS into BXDC.
These are independent decisions, by independent institutions, from independent jurisdictions. None reference the others. None coordinate. They are converging because the substrate is real and the convergence is rational.
The Physical Intelligence Economy is the build of physical infrastructure for intelligence at civilisational scale.
Eight substrate layers, converging into a single economic phenomenon.
These are not unrelated sectors. They are layers of one economy.
This is why institutional capital has been late on the Physical Intelligence Economy. Institutional capital is organised by sector vertical. Nuclear analysts do not speak with photonics analysts. Defence analysts do not speak with data centre analysts. Sell-side research is published by sector. Buy-side teams are hired by sector. Portfolio managers are benchmarked by sector. Every incentive in the industry reinforces the vertical organisation.
The convergence is happening across the boundaries this organisation creates. The boundaries are precisely where the most consequential movements are. Institutional capital, by construction, cannot see what forms across them.
This blind spot is structural. It does not resolve when the next analyst is hired. It is embedded in how the industry organises itself, reinforced by hiring, by benchmarking, by research publication, by every routine of professional practice. It persists because it serves the way the industry is organised, regardless of whether it serves the capital the industry deploys.
This is one expression of a broader failure.
The intelligence apparatus that sophisticated capital has historically relied on — sell-side research, mainstream financial media, expert consensus, the consultant class — has been late or wrong on every structural call of the last twenty years. Iraq weapons of mass destruction. The 2008 financial crisis. The election outcomes that shaped the current decade. The origins of COVID. The AI buildout itself.
The failure is not incompetence. It is structural. Institutional intelligence is produced inside incentive systems that shape what gets seen and what gets said. Sell-side research serves client relationships. Mainstream media depends on advertising. Expert opinion travels through networks that produce consensus before they produce accuracy. Each pipeline produces systematically distorted outputs aligned with the incentives that produce it. No conspiracy is required to explain the failures. The structure is sufficient.
The audience that matters has already adapted to this. Sophisticated allocators no longer treat the institutional intelligence pipeline as authoritative. They triangulate informally — across alternative sources, direct relationships, primary documents, their own pattern recognition — because they have learned that the captured pipeline produces outputs aligned with the incentives that produce it rather than with the reality the capital has to navigate.
What we have built is the systematic version of the triangulation sophisticated capital is already performing. Architecture designed to be epistemically disciplined rather than incentive-captured. Sources that sit outside the filters that distort the institutional pipeline. Synthesis that integrates across independent capital channels rather than aggregating consensus from captured ones. The trust mechanism is structural — falsification as discipline, disagreement as data, conviction below threshold suppressing action — rather than claimed.
The institutions that lost authority lost it for specific reasons. We built against those reasons rather than reproducing them.
The most patient capital in the world has been positioning around the Physical Intelligence Economy for some time.
Sovereign wealth funds with multi-decade time horizons. Pension capital that thinks in generations. Infrastructure capital that operates on twenty-year deployment cycles. They have been moving ahead of consensus because they have the time horizons to see structural rather than cyclical movement, and because they have the analytical capacity to recognise convergence before it has been named.
When independent capital pools, operating from independent jurisdictions, with no coordination between them, converge on the same substrate layers — that convergence carries information that consensus research cannot. Each pool has different incentive distortions. Each operates under different regulatory regimes. Each answers to different stakeholders. Convergence across these distortions is closer to actual signal than any single source could produce.
This is what we observe and what we act on. The convergence is the signal. The substrate is the destination. The Physical Intelligence Economy is the name of what these independent pools are collectively building, whether they have the vocabulary for it yet or not.
FORMIRA detects what is forming in the Physical Intelligence Economy by integrating intelligence across the conviction supply chain.
Capital does not appear in public markets fully formed. It moves upstream first — in investment committees, in sovereign strategic reviews, in patient capital deployments that compound over years before they show in any disclosure. By the time conviction has traveled through this supply chain to reach public visibility, it has been forming for some time.
Consensus research watches the public visibility. We watch the supply chain. The difference is whether you see what capital has decided after the decision has been priced, or whether you see the decision forming while it is still being made.
This requires architecture that operates across multiple capital channels simultaneously, multiple disclosure regimes, multiple structural forms of commitment. Sovereign wealth deployment. Infrastructure capital intermediation. Corporate operator positioning. Each channel surfaces different parts of the same underlying conviction. Convergence across channels is what we measure. Conviction is what we produce.
We say what is built. We say what is being built. We say what is designed. We never collapse those distinctions to make the claim sound stronger than the reality supports.
We built FORMIRA because we needed it to exist.
We recognised through building it that the need was structural, not personal — that sophisticated capital everywhere has been adapting to the same failures we were adapting to, and that the systematic response to those failures was infrastructure that did not yet exist.
What we built is the work the recognition demanded. We are now opening this work to others who have seen what we have seen.
If you have been adapting to the same failures, if you have been triangulating because you had to, if what we have articulated resonates with what you have observed in your own work — the conversation would be welcome.