Today’s AI is inherently expensive. Frontier AI model builders have projected that a single training run could cost over a hundred billion dollars within a few years. Already this year, annual spend on AI training of some model builders is rumored to be over 100 billion dollars. No model builder in Europe can match that kind of spending. It’s hence needless to say that Europe will not win by doing more of the same.
Today’s models are wasteful. Brute scale has been the fastest path to capability when capital is abundant. Without the capital, Europe must be resource efficient. Constraint and necessity have oftentimes been the mother of innovation and better engineering. To have the pressure to engineer and innovate is not necessarily a disadvantage, but a call to action.
AI waves have historically followed paradigm shifts in compute. Quantum holds promise in delivering the next one. It accelerates frontier AI development in multiple ways, ranging from training data simulation to solving optimization problems. At its core, frontier AI relies on the kind of computational complexity that quantum systems are built for. The obvious opportunity is that quantum is far cheaper than classical brute scale. This is the paradigm that Europe should take a bet on.
While full quantum advantage remains a technology for the future, a shift from classical to quantum-inspired approaches is happening today. Quantum-inspired methods—techniques derived from quantum mechanics that run on today’s GPUs—are already delivering real gains. While they deliver an advantage today, the advantage further deepens as quantum hardware matures: first hybrid, then fault-tolerant full quantum, with each stage extending the gain. The world is no longer waiting for a single breakthrough.
Europe has spent a decade watching American and Chinese companies capture the value of cloud, mobile, social, and AI waves. Quantum is different. It is early enough for no one to hold an insurmountable lead, as error correction and scaling remain unsolved for everyone. And it rests on the kind of fundamental physics where European research has long been world-class. Europe has two of the leading quantum hubs in Cambridge and Helsinki, billions in committed public funding, and a strategy to lead the field by 2030. The companies are being built, the foundations are being laid. Quantum is a race Europe actually has a chance to win.
AI is the North Star, and quantum is the compute paradigm that will change the AI race. Europe’s science, engineering depth, and capital discipline make Europe well positioned to win. Europe just needs the ambition and action to take the lead.