Sir Afilonius Rex

Bilbao, Spain. 13th December 2025

Oh, Europe, pronounced terminal yet again. This time, a Radically Human Ventures sage delivers the verdict. He is part probability theorist and part story-driven paterfamilias. He assigns the continent a 5% chance of civilisational survival. Meanwhile, he composes a moist-eyed farewell to the “Old Continent.” It’s an elegy made from familiar spare parts. It includes a dash of Peter Zeihan’s geopolitical fatalism and a sprinkle of Silicon Valley panel wisdom. There’s a belief that the future will be dominated by two AI giants. In this scenario, Europe is merely a polite museum gift shop.

It’s a very American way of looking eastwards: gazing across a virtual Atlantic and mistaking reinvention for decay.

The argument rests on a “perfect storm.” It involves collapsing hardware, paralysed software, and an energy crunch. This supposedly ensures Europe’s demotion to technological vassalage. The trouble is that, on inspection, the storm turns out to be little more than a puddle.

Take hardware. Zeihan’s long-running thesis has been circulating for years. It claims that cheap Russian gas has gone, German industry is finished, and that demographics are sealing the deal. It remains essentially unchanged and largely untroubled by Europe’s capacity to adapt. In 2025, the obituary still hasn’t arrived. Germany now generates more than half its electricity from renewables, with wind and solar continuing to scale as emissions fall. Nuclear power, once politically radioactive, is quietly being rehabilitated across multiple countries as a stabilising bridge. Industrial champions haven’t fled; they’ve pivoted. BASF, Siemens and their peers are retooling for green production and global supply chains. Volkswagen’s factory closures are less a retreat than a refit for electric vehicles. In this area, European firms are shaping standards rather than chasing them.

Demographics, too, are treated as destiny when they are better understood as policy challenges. Immigration pathways, skills programmes and labour-market reforms are uneven but real. And the EU itself, the supposedly dissolving glue, has rarely been more integrated, deploying hundreds of billions in shared investments in industry, energy, and resilience. Europe isn’t frozen in amber; it’s mid-transition.

Then there’s the software panic: the claim that Europe’s regulatory culture makes it incapable of competing in AI, doomed to rent compute from American hyperscalers or Chinese giants. The caricature is familiar, permits that take years, grids that can’t handle five gigawatts, exponential technology strangled by linear bureaucracy. It’s also increasingly out of date.

Europe is building sovereign AI infrastructure at scale. Under the EuroHPC Joint Undertaking, successive waves of “AI Factories” have rolled out across the continent—from Finland to Poland, Czechia to Romania, with more than 19 operational or incoming by 2026. Nine new AI-optimised supercomputers will roughly triple public compute capacity. Access layers are deliberately being broadened, not left to market gravity.

On the private side, Mistral AI has emerged as something rarer than a European unicorn: a globally relevant AI lab with a distinct philosophy. Valued at €12–19bn, it has released multilingual frontier models, efficient edge systems for robotics and defence, and open-source tools that emphasise diffusion over enclosure. Its partnerships, with SAP, Microsoft, CMA CGM and others, suggest less a vassal state than a hybrid strategy: sovereign capability, globally networked.

Regulation, meanwhile, is doing what European regulation often does best: turning constraint into leverage. The EU AI Act is no ghost-hunting exercise; it is a framework paired with sandboxes, sectoral guidance and an “Apply AI” strategy designed to accelerate real-world adoption. As with GDPR, Europe is once again exporting norms. In contrast, others scramble to respond, Washington is mired in fragmentation, and Beijing is pursuing a very different social contract.

Which brings us to energy, the supposed choke point where AI ambition goes to die. Here, the story flips entirely. Europe’s accelerating renewables build-out, coupled with nuclear revival, delivers exactly what large-scale compute requires: dense, reliable, increasingly low-carbon power. Data centres are proliferating accordingly. There are hyperscale campuses in Sweden and Spain’s Aragón region. Billion-euro investments are happening in Belgium. Mega-hubs are being developed in Wales. Grid upgrades and cross-border coordination are turning the energy transition into an infrastructure advantage, not a liability.

Seen this way, the intersection of hardware, software and energy isn’t Europe’s undoing; it’s its differentiator. The continent is constructing what may be the world’s most sustainable AI foundation. It is politically legitimate. This foundation is slower to ignite, perhaps, but it is far harder to destabilise.

The final verdict, then, is less dramatic than the eulogy would have us believe. This is not a civilisation in retreat, but one in the messy middle of reinvention. Europe isn’t “on the menu” of a US–China duopoly. It’s assembling its own table. It includes shared compute, green power, and a stubborn insistence that technology serve human ends.

The absolute risk of irrelevance does not belong to Europe. It is tied to a particular genre of collapse narrative. This narrative is confident, recyclable, and increasingly detached from reality. The Old Continent might not move at venture speed. But it endures, integrates and, when it matters, surprises.

Many thanks for reading.