zartsnarf

WHO: The wealthy are the only market

As automation and AI displace most labour, the majority transitions onto Universal Basic Income. That income exists, but it is set at subsistence, stays flat, and does not grow. Discretionary spending thins out. Purchasing power does not disappear, it concentrates on the people who own scarce and hard assets, and the wealth divide widens as a mechanical result rather than a side effect.

So the customer of the next economy is not the mass market. The mass market keeps an income but loses its pricing power, because UBI funds survival and nothing above it. The customer is the high-net-worth and ultra-high-net-worth individual, the only segment whose spending grows. Every product and service worth building should be tailored to that segment. Luxury, privacy, and exclusivity are the durable value drivers. Mass-market plays lose their economic engine.

WHAT: Scarcity that survives abundance

Price tracks scarcity. When something becomes abundant, its value falls to marginal cost and stops rewarding whoever supplies it. The question is not what becomes cheap, but what cannot. The instinct is to reach for hard commodities, gold and silver and copper and oil, on the logic that finite raw assets must hold value. The cost of extracting and refining almost any commodity is mostly energy and labour, and both go to zero, so driving them down does not make commodities scarcer, it makes them cheaper to pull out of the ground. The same convergence that carries this thesis pushes most raw-material prices down, not up.

What survives abundance is scarcity that no amount of energy, labour, or intelligence can manufacture. It splits three ways.

Structural scarcity is supply fixed by nature or geography, independent of the cost curve. Waterfront, prime location, capital-city perimeters, land, and the rights bolted to specific land: water, mineral, development. You cannot print a coastline or automate the supply of a city centre. This is the core, and the wealth concentration above only bids it higher.

Positional scarcity is value that depends on others not having it: trophy assets, blue-chip art, the authenticated and the original. As synthetic everything floods the world, the verifiably real and human-made gains a premium precisely because it cannot be reproduced.

Monetary scarcity is hard money: gold, silver, and bitcoin. Their value now primarily comes from currency debasement, not from production limits. Funding UBI for billions implies a sustained expansion of the money supply, and hard money hedges that expansion.

Commodities carry one transitional trade. The buildout creates bottlenecks before it delivers abundance, and copper, uranium, and grid metals see demand spikes during the transition even as the end state drives energy cheaply. That is a timing position on transition inputs, not necessarily a permanent store of value.

The structural bias is toward what abundance cannot touch, over anything intelligence, energy, or labour can now manufacture for essentially free.

WHERE: The Global South

The inputs to the J-curve are exogenous. Cheap energy, labour, and intelligence arrive as global utilities, regardless of any local government's competence. Historically those were the binding constraints on developing nations. Remove them and what remains, governance and infrastructure, becomes far faster and cheaper to build. A century of Western development compresses into years.

But a region being lifted by the curve and an owner capturing that lift are two different things. The lift is exogenous; capturing it runs on governance: property rights, rule of law, capital mobility, and protection from expropriation. The same compression that rewards an owner in a well-run jurisdiction strands capital in a badly run one, where the upside accrues to whoever seizes it rather than whoever holds title. So governance is the first thing to screen for, ahead of anything about the land itself.

That narrows the field. The compression lifts hardest where the base is lowest and the room to grow is largest, which rules out the West: saturated, expensive, overcompeted, and barely moved by the curve. The slope is steepest in the Global South, but only in the parts where ownership survives the holding period. Southeast Asia, stable South America, and the well-governed economies of Southern Africa such as Botswana and Namibia. Jurisdictions with active insurgency, capital controls, or debt distress come off the list regardless of how cheap the land looks. Within the surviving set, the targets are waterfront, lagoons, lakefronts, capital-city peripherals, and any corridor likely to absorb future infrastructure.

WHY: The three cost inputs go to zero

Within the decade, the three cost inputs of civilization collapse toward zero at the margin: energy becomes cheap and abundant, automation absorbs most physical labour, and AI absorbs most knowledge work. They do not arrive together. Intelligence collapses first, because it is software and scales at the speed of compute. Physical labour follows on a longer lag, because automating atoms means building, deploying, and maintaining hardware, and that runs at the speed of factories rather than models. The destination is the same, only the timing differs: knowledge work first, embodied work behind it.

Energy carries both, since cheap power runs the compute behind AI and the machines behind automation, and energy bends toward abundance from both sides at once. The AI build-out is the largest new load placed on the grid in generations, and that load pulls in the capital that expands supply. Supply answers on three curves: solar and storage, already the cheapest new electricity across much of the world and still falling; nuclear, with small modular reactors entering deployment and retired plants restarted to feed data centres; and fusion, drawing serious private capital with first commercial designs targeted for the 2030s.

That convergence forces all three positions. Strip out energy, labour, and intelligence as constraints, and only physical things stay scarce (the what), only asset owners keep growing purchasing power (the who), and the fastest growth lands where the base was lowest (the where). The three vectors intersect at one position.

Scarce physical assets, in the Global South, held for and sold to the wealthy. Concretely, land and real estate in strategically positioned developing geographies. Buy at pre-development prices and hold through the transformation.

Everything else compounds on that core and serves the same buyer on the same curve.

Luxury hospitality and experiential travel monetise the land directly, as the wealthy seek privacy, natural beauty, and untouched destinations.
Private wealth and financial services own the toll booth, as incoming capital needs structuring, custody, lending, and local vehicles.
Infrastructure concessions own the rails the curve runs on: ports, telecoms, power, logistics, water, and digital.
Compute and data-centre siting turn cheap local energy into the one input the AI build-out cannot get enough of. Land plus power in a stable jurisdiction is a site for the load that funds the whole curve.
Resource processing captures the refining margin locally instead of exporting raw material. This is a bet on owning the processing step, not on the price of the commodity, which the convergence pushes down.

WHEN: Now, before it's priced in

The logic is durable, but the entry price is not. The window is likely under ten years and possibly closer to five. After that, prices begin to discount the boom that is coming, and the asymmetry disappears. The move is to accumulate scarce physical assets in stable Global South geographies now, while they are still cheap, and hold for the J-curve.

The harder half is the exit. These assets are illiquid by nature, so it has to be planned at the point of purchase rather than assumed at the end. Buy where a future buyer will exist: titled land, clean ownership, and a jurisdiction whose rules will still hold when it is time to sell.