I used to park my motorbike right in front of the entrance, a step away from the time clocks. I would pass through the turnstiles and, in the dazzling light of late winter, walk into the shipyard, among the enormous blocks of steel, the cranes towering over the dock, the dismembered hull waiting for its meeting with the water, that first day of floating life. Launching has not always been a day of awakening; it has not always given life. In the last century, if not quite from Fitzcarraldo\u2019s mountain, ships slid into the sea from high, dry slipways. On logs or on rollers, thousands upon thousands of tonnes slowly set off on a run that ended in the sea, a deep thud of spray and waves, and then breath stopped for a moment. People told of the famous pistol, hanging on the taut breath of the man who signed off on the fearsome launch, and of the instant in which the suspended hull dug its thrust into the water, searched for a first balance, like a baby\u2019s first cry, and then \u2013 life or death.<\/p>\n\n\n\n
A calculation wrong, a shipyard ruined \u2013 and that signature, with its terrible responsibilities, deciding the go-ahead.<\/p>\n\n\n\n
The world has changed since then; progress has extinguished the fear of the launch, but the signature remains the seal of that trust which leads to action, to the progress of the project, to the investment of capital; the signature seals the bond between knowledge and responsibility for the decision taken. And the larger the investment, the greater the trust placed in the signature, and the heavier the responsibility: on this value chain, companies like mine have built an industry over almost two centuries.<\/p>\n\n\n\n
What English callsassurance<\/em> \u2013 or, more in general, the judgement business,<\/em> an entire industry devoted to providing projects with the confidence needed to unlock large infrastructure investments.<\/p>\n\n\n\n And yet something seems broken; a disruptive energy momentum seems to be losing strength, slowing down. In the face of climate change, the transition is failing to gather pace: electrification, the reduction of CO2 production, new synthetic energies \u2013 they are not taking off. According to the Energy Transition Outlook (DNV, ETO, 2025<\/em>), net-zero CO2, which scenarios compatible with the Paris Agreement placed at 2050, is expected to be reached at least four decades late, beyond 2090. Large, complex decisions are struggling to crystallise. The drive seems to be missing, decisions seem slow to arrive, and three diverging forces explain a structure increasingly short of breath.<\/p>\n\n\n\n First, the experts \u2013 those who sign the myriad of complex technical decisions needed to change course firmly and decisively \u2013 are hardly a growing species. According to the International Energy Agency (IEA, World Energy Employment, Dec 2025<\/em>), the global economy sees 1.2 older workers approaching retirement for every single young person starting work. For the energy industry, in advanced economies, the recorded decline is actually double: 2.4 people for every new entrant aged twenty-five or under. Like beavers: if not endangered, then at least a protected species. This is the first diverging force in the industry.<\/p>\n\n\n\n The second is tied to the transformative nature of the energy world, where by now even simply maintaining production, let alone improving on it, requires ever more investment and complex decisions, from marginal fields to life extensions, the whole of deepwater. An IEA study (The Implications of Oil and Gas Field Decline Rates<\/em>, 2025)<\/em>, analysing the investment needed simply to maintain current production levels, concluded that the oil industry \"has to run much faster just to stand still\"<\/em>. Running faster, but on fewer legs. Diverging, and making the first force worse.<\/p>\n\n\n\n Third and finally, the progressive decline in the economic value of engineering, with commercial hourly rates often in the order of a few, or very few, tens of euros, on budgets that amount to single-digit percentages of billion-euro projects.<\/p>\n\n\n\n Three directions as precise as they are destructive, and the speed at which they diverge appears to be increasing, showing clear signs of fatigue in the industry.<\/p>\n\n\n\n In Europe, young people are increasingly committed to respecting the environment. In a meeting where I was integrating two new teams, many of them young, I broke the ice by asking everyone what their most desirable car or motorbike was \u2013 and among the answers, some had no driving licence, some preferred the bus. It becomes hard to find young people interested in spending sleepless nights on hydrocarbon development projects when the prevailing sentiment would like to eliminate CO2, and fossil fuels along with it, from our small, overheated world. Yet deepwater, today, is learned in oil and gas; a circle that does not close.<\/p>\n\n\n\n Meanwhile, the people who in the 1970s and 1980s invented and absorbed the great industrial change are now playing with their grandchildren and nursing the fatigue of very intense lives. Unreachable.<\/p>\n\n\n\n The glamour is missing too; the big digital companies attract talent in a way offshore engineering companies can hardly manage today. While a major infrastructure project runs through a single development loop over more than a decade, across pre-FEED, FEED and construction phases, other industries churn out a continuous stream of new Copernican revolutions, amid the enthusiasm of understandably fascinated young people \u2013 and the low prices of offshore engineering turn the key in the lock.<\/p>\n\n\n\n But while the gap is growing, and has been for a long time, the industry has long been on the move, using intelligence and technology to raise the productivity of experts. Let us call it knowledge<\/em> della<\/em> industrialisation<\/em>: the set of standards, calculation software, modelling and sensor technology. The use of synthetic data calibrated on real data from laboratories and full-scale tests is one example of how very modern technologies are already well-known solutions in use; neural models were being used in probabilistic calculations to solve offshore engineering problems as early as the 1980s, in Italy.<\/p>\n\n\n\n The industry has developed deterministic systems, written and codified, to respond to investors\u2019 many uncertainties, to unlock enormous capital, to give shape to the growing number of decisions needed to imagine, build and improve the world\u2019s energy infrastructure, even though the number of experts has been shrinking for a long time. Why, then, does the industry that knew how to feed the world now seem emptied of the pirates, the genius, the crazy explorers who made the history of Energy?<\/p>\n\n\n\n Could it be losing trust?<\/p>\n\n\n\n Between technologies that metabolise knowledge and a declining commercial value, the signature does not lose its value. It remains the seal of that trust with which the expert moves the industry, lets ships sail from port, lets foundations be installed at sea. The seal of that responsibility which weighs on the expert themselves, and directly on the legal entity. But mature calculation systems have long been absorbing the decline in experts, and in their mature state they are no longer enough \u2013 they are saturated. Like the existing infrastructure, the industrialisation of knowledge also needs new solutions.<\/p>\n\n\n\n What happens to a signature that keeps its value, that carries a growing weight of responsibility, as the number of people able to analyse, interpret and reasonably decide shrinks, in a technological landscape that is ever more complex and necessarily in motion? The digital world today is pushing generative and statistical solutions, with new and expanding cognitive systems. These stochastic systems, however, require skilful management of a mechanism that erodes knowledge, understanding and discernment \u2013 one that is not visible on the surface, and that is silent. The black box of AI, the Cognitive Monster (CM) we have all met by now, has not reached mature applications at the Governance level of large companies, has not reached any official role in decision-making in this industry, and yet it exists everywhere, spreading through the uncontrolled undergrowth, on individual machines, in small patches of silent programs. And the cognitive potential remains locked at the individual level of the Expert, far from structural implementations driven by Governance \u2013 as if blocked by the decay it can cause, as if Governance were unaware of the means to control it.<\/p>\n\n\n\n Use friction to retain ownership of decisions. Invest in the encyclopaedic preservation of knowledge, so that decisions can be controlled and the signature remains a seal of trust. Keep clear control over the purpose of knowledge \u2013 that harnessing<\/em> of the Monster that keeps it bound to the reasons of knowledge. Governance can harness the CM, but to date it has not done so.<\/p>\n\n\n\n