Read a hundred deep explorations across twenty sectors and one pattern sits underneath all the others. In domain after domain — semiconductors, energy, critical minerals, financial data, agricultural inputs, shipping lanes — the world is becoming more dependent on single companies, single places, and single points of failure at precisely the moment those dependencies are becoming more dangerous to hold.
This is not a coincidence of separate industries. It is one structural force expressing itself everywhere at once.
The same shape, everywhere
- Compute concentrates on NVIDIA’s GPUs and the CUDA software moat, which run on chips only TSMC can fabricate, on machines only ASML can build.
- Energy funnels through chokepoints like the Strait of Hormuz — roughly 20 million barrels a day through a single waterway.
- Critical minerals route through China, which controls the processing of the large majority of the strategic minerals the energy transition and the defense industrial base both depend on.
- Financial information concentrates inside the Bloomberg Terminal’s interlocking switching costs.
- Fast fashion rests on a single manufacturing cluster that, if removed, collapses the model that depends on it.
Different industries, identical geometry: a thin waist in the hourglass through which everything must pass.
Why it compounds
These chokepoints are not independent — they nest inside one another. TSMC enables the AI buildout; the AI buildout amplifies demand for TSMC’s chips; that demand amplifies TSMC’s leverage and the geopolitical stakes of the strait it sits beside. Concentration in one layer raises the value of concentration in the layer above it. The corpus maps this as recursion: chokepoints within chokepoints, flywheels feeding flywheels.
Markets reward this. Concentration is efficient. It lowers cost, raises margins, and compounds advantage. Every actor optimizing locally pushes the whole system toward the same fragile equilibrium.
Why it matters now
The danger is the timing. Economic logic is driving concentration to a peak just as the geopolitical and climate context demands the opposite — redundancy, slack, the ability to absorb a shock. The corpus identifies a window, roughly 2029–2032, when several of these stressed systems could come under pressure together: semiconductor supply, an energy chokepoint, mineral processing, and the dollar’s role as the enforcement mechanism behind export controls. The interactions between those simultaneous stresses are almost entirely unexplored. That gap is the thing at the center of the map.
The deeper finding is an asymmetry. The corpus is rich in detail on the mechanisms of concentration and comparatively thin on the mechanisms of adaptation — not because the research missed them, but because they are genuinely weaker in the world. We are in a race between concentration that compounds exponentially and institutional capacity that is, at best, linear. That asymmetry may be the most important thing the corpus has to say.
The pages below are the other four patterns this thesis is built from, and the explorations that evidence it.