When Donald Trump, now president of the United States, returns to talk about Greenland as a strategic territory, the media debate often focuses on geopolitical balances. But beneath the surface of the Arctic ice – literally – lies an even deeper story.
This is the story of rare earths, chemical elements little known to the general public, yet indispensable to the functioning of the contemporary world. They are not seen, they don’t make noise, they don’t attract tourists. Yet they are everywhere: in data centers, in supercomputers, in factories that train artificial intelligence.
This search for rare earths is a journey that does not follow classic routes. The journey to their discovery crosses deserts, steppes, mountains, and Arctic regions. A journey combining geology and innovation, extreme places and artificial intelligence.
More valuable than oil: what rare earths are and why they are so important
Rare earths are a group of 17 chemical elements that include unfamiliar names such as neodymium, dysprosium, europium, terbium, yttrium and lanthanum. They do not glitter, they do not appear as nuggets, they do not attract the traveler’s eye. Yet they are everywhere, hidden in the devices and infrastructure that hold up the modern world.
Their value lies not so much in their absolute rarity as in the complexity of their extraction and refining. Separating them requires long, expensive and technologically advanced processes. This is why they have become strategic materials, often referred to as “more valuable than oil“: not because they produce energy, but because they make possible the technology that governs the present and the future.
Without rare earths, there would be no high-performance magnets, industrial lasers, semiconductor components, advanced cooling systems. Today they are worth more than oil not by unit weight, but by strategic centrality: those who control rare earths control the infrastructure of the digital world.
Bayan Obo, Mount Weld, Mountain Pass and the North Pole: the invisible treasure map
Where the world’s rare earth deposits are found. The geography of rare earths does not follow the traditional boundaries of tourism, but those of geology.
The global heart of rare earths beats in China, which controls much of themining and especially the refining. The key name is Bayan Obo, in Inner Mongolia: an almost lunar landscape of sand, dust and open-pit mining. A huge share of the globally used rare earths comes from here.
In Australia, the Mount Weld site-one of the few outside of China that is truly operational-emerges in the heart of theoutback as one of the purest and most important deposits, making the Australian continent a key point in the global balance of resources.
In the United States, the Mountain Pass mine represents an attempt to reduce external dependence and rebuild an autonomous supply chain.
Then there is the Arctic. Greenland in particular, where reserves are immense but difficult to exploit for environmental, climatic, and political reasons. Greenland is a land of ice, silence and immense space, where potential deposits have been identified that are still largely unexplored. Here the journey becomes extreme, and the link between pristine nature and global technology seems more evident than elsewhere.
At the heart of artificial intelligence
Today, rare earths are not primarily needed for consumer products (e.g., smartphones and wind turbines), but for AI Factories: large technological infrastructures where artificial intelligence is trained, managed, and made operational.
These factories do not produce traditional objects, but computing power. Inside them are supercomputers, data centers, GPUs, industrial robots and automated systems. All of which depend directly on rare earths.
Neodymium magnets move motors and robotic arms, dysprosium enables stability at high temperatures, europium and terbium enable computer vision, lanthanum and cerium make electronic components efficient. Without these materials, Artificial Intelligence would remain a theoretical concept.
AI Factories are thus the meeting point between algorithms and matter, between software and the subsurface. Rare earths come into play quietly but decisively:
- in the very high-efficiency magnets that drive the cooling systems;
- in the advanced electronic components of GPUs and AI-dedicated chips;
- In optical systems, lasers, and precision sensors.
Without rare earths, AI factories simply would not work. This is not an exaggeration: they are the physical bottleneck of artificial intelligence.
Where the AI factories are and where they will be located.
Energy, climate, and new technological geographies. AI factories cannot spring up everywhere. They need huge amounts of energy, advanced infrastructure and favorable environmental conditions.
- North America is home to some of the most powerful computing hubs in the world.
- Northern Europe-with Sweden, Finland, and Iceland-offers ideal cold climates for cooling servers and a growing focus on advanced technologies.
- East Asia, particularly China, South Korea and Japan, concentrates a key part of hardware production.
In this scenario, Arctic regions take on an increasingly central role: not only as a reservoir of resources, but as the potential home of the AI factories of the future.
AI factories are concentrated where three factors converge: abundant energy, favorable climate and political stability.
Today we find them mainly:
- United States, between Texas, Virginia , and the Pacific Northwest;
- in northern Europe, between Sweden, Finland and Iceland, where the natural cold helps with cooling;
- in East Asia, with China, South Korea and Japan.
It is no coincidence that many of these regions are looking with interest at Arctic routes and northern resources: the distance between mines and data centers is becoming a strategic variable.
From the dunes of Bayan Obo to the ice of Greenland: a journey into the invisible future
The journey to discover rare earths is not about sights to photograph, but connections to understand. It starts from the dusty dunes of Bayan Obo, crosses deserts and continents, and reaches the frozen lands of Greenland, where silence hides crucial resources.
This is where the future of AI factories is born. A future that depends on distant, fragile, and often forgotten places. Telling the story of rare earths means telling the hidden side of progress, looking at the world not from what appears on the surface, but from what sustains it.
For further study:
