The Subterranean Layer: Why Critical Minerals Became the New Geopolitics
PriyankaThe critical minerals market hit $328 billion in 2024. It's projected to nearly double to $586 billion by 2032. Lithium demand alone surged 30% last year. Uranium demand is set to jump 28% by 2030. There are 75 nuclear reactors under construction worldwide and another 120 planned.
This is not a commodities cycle. This is a structural realignment of what the global economy runs on.
For decades, the bottleneck in the world economy was energy -- oil, gas, the infrastructure to move it. That era hasn't ended, but a new layer has emerged underneath it. The atoms that power electrification, AI infrastructure, defense systems, and the energy grid itself have become the most geopolitically contested resources on the planet. Not oil. Lithium. Uranium. Copper. Rare earths. Cobalt. Gallium. Germanium.
The thesis: critical minerals are no longer a niche commodity play. They are the infrastructure layer beneath the infrastructure -- and whoever controls the processing stack controls the next industrial era.
The Trigger: Electrification, AI, and Defense Collide
Three megatrends are converging on the same set of minerals simultaneously.
Electrification. A single battery-electric vehicle requires roughly six times the mineral input of a conventional car. That means lithium, nickel, cobalt, graphite, manganese, copper, and rare earth magnets -- per vehicle. Global EV adoption is still in early innings relative to the total vehicle fleet. The demand curve doesn't flatten; it steepens.
AI infrastructure. Every data center expansion requires massive copper for wiring, rare earths for high-performance magnets, and an exponentially growing energy supply. The hyperscalers -- Microsoft, Google, Amazon -- are signing nuclear power purchase agreements because renewables alone can't meet the baseload demands of training and inference. Nuclear means uranium. Grid expansion means copper. Compute at scale is a mining problem disguised as a software one.
Defense autonomy. Modern weapons systems -- from the F-35 to precision-guided munitions to satellite constellations -- are built on rare earth permanent magnets. The Pentagon's supply chain runs, in uncomfortable proportion, through Chinese processing facilities. Every hypersonic missile, every military drone, every EW system depends on materials the U.S. does not currently refine at scale domestically.
The Inflation Reduction Act and the CHIPS and Science Act both contain critical mineral provisions that are often overlooked. Tax credits for domestic mineral processing. Incentives for battery recycling. Funding for the Defense Production Act to stockpile strategic materials. The policy infrastructure is being built. The question is whether the industrial base follows.
The Real Bottleneck: Processing, Not Mining
Here is the part most people get wrong: the problem is not finding the minerals. It's refining them.
China controls approximately 90% of global rare earth processing. For heavy rare earths -- the ones most critical for defense applications and high-performance magnets -- that number is 99%. China also dominates midstream processing for lithium, cobalt, and graphite.
In October 2025, China expanded export controls on rare earths and related processing technology779220_EN.pdf). Earlier that year, it restricted exports of gallium, germanium, and antimony -- three materials critical to semiconductors, defense optics, and ammunition. These weren't idle threats. Shipments dropped measurably.
The U.S. and its allies can mine. Australia has lithium. Canada has nickel. The U.S. has rare earth deposits. But mining without refining capacity is like having crude oil and no refineries. The value -- and the leverage -- sits in the processing layer.
This is why startups like Durin Mining (which raised $3.4M in seed funding in April 2025, backed by a16z crypto and Bedrock Capital) are interesting not just because they automate drilling, but because they accelerate the upstream that feeds the midstream. Durin builds semi-autonomous drill rigs designed to make mineral exploration faster and cheaper. Today, roughly 3 in 1,000 exploration attempts successfully find a deposit. Automating that bottleneck changes the economics of domestic mineral development.
But Durin is upstream. The bigger gap is midstream: separation, purification, refining. That's where the geopolitical leverage concentrates.
The Control Question
Critical minerals are embedded in everything that matters to the next economy:
- Batteries -- lithium, nickel, cobalt, manganese, graphite
- Grid infrastructure -- copper, aluminum, rare earths
- AI hardware -- gallium, germanium, copper, rare earths
- Defense systems -- neodymium, dysprosium, terbium, samarium
- Nuclear energy -- uranium, zirconium
But the contrarian view is worth sitting with: the next generation of "American dynamism" may not be software companies or defense primes. It may be regulated, politically messy, capital-intensive, NIMBY-challenged mining and refining operations. That's an uncomfortable fit for venture capital as traditionally practiced. It's a natural fit for patient capital, project finance, and strategic corporate venture arms.
Where the Asymmetric Opportunities Are
The value chain in critical minerals breaks into five layers, and the venture-scale opportunities sit in the gaps between them:
1. AI-enabled exploration. Geospatial machine learning, predictive drilling models, automated rigs. This is where Durin sits. The exploration phase is ripe for software-eats-hardware dynamics.
2. Domestic refining and separation. The biggest structural gap. Startups building modular, scalable refining capacity for lithium, rare earths, or battery-grade nickel. This is where DOE money is flowing.
3. Recycling and urban mining. End-of-life EV batteries contain recoverable lithium, cobalt, and nickel. Battery recycling is a circular supply chain play with strong unit economics once collection infrastructure scales.
4. Nuclear fuel cycle innovation. Uranium enrichment, fuel fabrication, and next-gen reactor fuel forms (HALEU for SMRs). With 75+ reactors under construction and uranium demand spiking, the fuel supply chain is a bottleneck within a bottleneck.
5. Mineral middleware. Trading platforms, certification and provenance systems, logistics optimization, supply chain transparency. Think of it as the Stripe-for-minerals layer -- the connective infrastructure that makes fragmented commodity markets function like modern supply chains.
The next trillion-dollar infrastructure buildout won't be in the cloud. It will be underground.