The Last Word: We’re not entering the age of artificial intelligence. We’re re-entering the age of energy, this time with better marketing.
Europe has spent much of April watching the Strait of Hormuz like a pressure gauge on the boiler room of the global economy. In recent days, Brussels warned that a prolonged Iran conflict could trigger a sustained energy supply shock, forcing cuts in fuel consumption and leaving the continent struggling to refill gas storage before winter. Industry voices have gone further, cautioning that even aviation could feel the strain, with localised jet-fuel shortages a plausible near-term risk.
It is a striking contrast. The public conversation remains transfixed by artificial intelligence: its applications, its promises, and the increasingly feverish competition around it. Companies are rushing to embed it into everything, investors are chasing the next winner, and governments are treating AI leadership as a proxy for future power. Yet the underlying anxiety is far more elemental. It is about fuel, flow and capacity. About whether the system holds.
For the past two decades, we have grown comfortable with the idea that the economy was becoming weightless. Software scaled without friction. Cloud computing abstracted away infrastructure. Value migrated from physical assets to intangible ones. The prevailing belief was simple: the future would be digital, and therefore detached from the hard limits of the physical world. That assumption is beginning to fracture.
The current energy tension is not an anomaly; it is a reminder. The digital economy never escaped the physical world. It merely sat on top of it, drawing from systems that were stable enough to be ignored. Energy was abundant, logistics predictable, infrastructure taken for granted. Under those conditions, abstraction felt real.
A race for megawatts
Artificial intelligence is now exposing the limits of that illusion. For all its apparent immateriality, AI is profoundly physical. It depends on vast data centres, dense clusters of chips, complex cooling systems and—above all—electricity. The more capable the system, the greater the demand. Intelligence, at scale, is not a software problem. It is an energy one.
The shift is already visible in behaviour. The world’s leading technology firms are no longer simply building models; they are securing power. Long-term energy contracts, direct investments in renewables, renewed interest in nuclear—these are no longer peripheral considerations but core strategic moves. The race for AI leadership is quietly becoming a race for megawatts.
And the constraints are no longer theoretical. In multiple markets, data centre expansion is running up against the limits of grid capacity. Projects worth billions are delayed not by a lack of capital or ambition, but by the inability to connect to reliable power at scale. Infrastructure, not innovation, is setting the pace.
This is where Europe’s position becomes more exposed. The continent has the intellectual capital to compete in AI, and the regulatory muscle to shape its development. But it also faces structurally higher energy costs, tighter supply conditions and slower infrastructure expansion. That combination matters. You can lead debates on ethics and governance, but if you cannot secure affordable, reliable power, you risk becoming a consumer rather than a producer of the next technological wave.
Elsewhere, the picture looks different. Across the Gulf, energy is not being treated as a legacy sector to be managed down, but as a strategic platform to be expanded. Investment is flowing simultaneously into hydrocarbons, solar, hydrogen and digital infrastructure. The logic is straightforward: if the future runs on electricity, then controlling its generation and distribution is a form of long-term leverage. AI is not separate from that equation. It depends on it.
An era of constraint
All of this points to a broader shift—one that sits beneath the headlines. We are moving away from an era defined by abstraction and into one shaped once again by constraint. Energy, materials, infrastructure and geography are reasserting themselves as decisive factors. The difference now is that they underpin far more complex and powerful systems than before.
This is the reinvention blind spot. Organisations are rushing to adopt AI, to experiment, to integrate, to signal progress. Far fewer are interrogating the conditions required to sustain it. Reinvention is not just about embracing new technologies. It is about understanding the system in which those technologies operate—and the limits that system imposes.
The mistake is not enthusiasm for AI. It is the assumption that it exists in isolation. Intelligence may be accelerating, but it remains bounded by the realities of power, capacity and supply. Those realities are becoming harder to ignore. The cloud has a footprint. The model has a cost. The system has a ceiling. We’re not entering the age of AI. We’re re-entering the age of energy—with better marketing.
Photo: Dreamstime.