# AI and the engineering profession: above all, a human challenge **Authors:** Laurent Barlet **Categories:** Data & AI **Tags:** AI, Climate **Last Updated:** 2026-06-09T04:00:00.040Z **Reading Time:** 6 min read --- ## Summary AI is not just a technological shift: it is a human one. Laurent Barlet, engineering executive, argues that the real challenge lies in culture, trust, and method, not in access to tools. --- Twenty-five years ago, when I was a young engineer, data was physical: it lived in paper files stacked in cabinet drawers. Today, data is stored at scale on servers. That shift has moved data management into an entirely different dimension. AI now makes it possible to analyse that data at speeds and on a scale that would have been unimaginable at the start of my career. Just a year ago, AI was still largely a prospective subject. It has since entered our society and our businesses like a groundswell: the kind of wave that does not merely pass through a landscape but reshapes it entirely. AI is doing precisely that to our working environment. # A gap in culture, not just in access Today, while almost everyone acknowledges the value of AI, whether in the form of chatbots or sector-specific tools, actual usage patterns remain highly uneven. Unlike previous technological revolutions, AI does not simply create an access gap; it exposes gaps in culture and perception. What strikes me, beyond the sheer speed at which these tools are being developed and diffused, is **how differently they are understood across generations,** even between age groups that are not so far apart. I see this clearly within my own family. Between my eldest child, born in 2002, and my youngest one, born in 2007, the relationship to AI is already not the same. Both use these tools, but with different levels of understanding and appropriation, particularly in their awareness of what these tools can actually do. That gap, subtle as it may appear, is structurally significant. It tells us something essential: if the challenge of AI is certainly technological for those who build it, it is above all human for the organisations that deploy it. # A question of maturity, not technology Contrary to a widespread assumption, access to technology is no longer the primary determining factor. Over time, all actors will have access to the same tools. The differentiator will lie in the capacity to **align technology, organisation and data governance,** in the mastery of use cases, and in the trust built with clients. In industrial sectors where data is sensitive, that trust is foundational. Data security, confidentiality and sovereignty are far from secondary constraints; they condition the entire scope of what AI can be used for. This is why AI integration cannot be opportunistic. **It requires a framework, clear rules and a progressive approach.** That is precisely where the challenge for companies lies today: turning widespread access to AI into disciplined, coherent and value-creating usage. # Taking the time to structure use cases AI creates a sense of urgency, a feeling that one must move fast. But in complex organisations, **moving fast without method generates more risk than value.** The priority challenge is therefore diagnosis: identifying relevant use cases, assessing their value and prioritising accordingly. This must then be followed by building new processes, testing solutions and validating their robustness. That process may feel slow. It is, in reality, indispensable if the promise of AI is to be turned into measurable results. # The engineer as arbiter of an augmented system This shift is profoundly reconfiguring the role of the engineer. AI does not replace the engineer. It makes the engineer **a central actor in validating decisions and ensuring their reliability.** The engineer is no longer the person who analyses data, runs calculations and designs solutions. The engineer is now the person who writes the prompts, arbitrates between scenarios, orchestrates analyses, and stands as **guarantor of the relevance of the results produced.** This requires mastering not only one's professional expertise, but also the logic of the tools, how they work and where they fall short. This imposes a shift in frame of reference for companies, and that shift itself requires method. How do we integrate AI into our professions? How do we articulate its use alongside our sector expertise? These are questions that are reshaping corporate cultures. A concrete example: we use automated data analysis tools to produce diagnostics on infrastructure or industrial processes. These are powerful tools that very quickly generate a range of technical solutions, whereas until recently we would spend considerable time collecting data, analysing it and sorting through it. A question comes up repeatedly: **can we trust these diagnostics, given that they were not produced by a domain expert?** Trust in the new tools of artificial intelligence is a genuine issue that must be factored into any company's transformation strategy. # Managing the transformation: a collective challenge That trust cannot be decreed. **It is built over time,** through experience-sharing, training and the progressive acculturation of teams to AI-based ways of working. Professions are changing. The challenge is not to slow that change, but to manage it. That means bringing engineers along in this dynamic, through concrete use cases, experimentation and the development of AI-integrated solutions. We are squarely in the territory of human resources: supporting skills development, providing clear reference points and creating the conditions for lasting adoption. This transformation also raises a fundamental question: how will the next generation relate to these tools? The engineers of tomorrow will have grown up with AI. That is a genuine opportunity. Yet it should not lead us to treat these technologies as self-evident or as universal solutions. Being comfortable with AI is one thing; **understanding in depth the professions to which it is applied is another.** Knowing the operational stakes, grasping the realities on the ground, gauging the constraints: all of this remains indispensable if AI is to be used with real relevance. This is why a dual requirement must be developed: **being able to work with AI, while maintaining a critical eye on its outputs.** That means knowing when to step back, confronting the tool's proposals with actual needs, and drawing on solid technical expertise. AI does not replace that expertise. It raises the bar for it. # The question of limits: an industrial and environmental challenge Another question is now emerging with force: that of the physical limits of AI. The rapid scaling of digital infrastructure, data centres in particular, poses major challenges: **energy consumption, water requirements for cooling, and pressure on local resources.** These challenges become critical as the major technology players accelerate their deployments. For countries and territories, the question is no longer simply one of attracting these investments, but of making them compatible with environmental equilibria. # Closing the loop: how engineering makes AI sustainable This is where engineering reclaims a central role. Behind AI lies a physical infrastructure, with real resources and real environmental constraints. It is precisely on these dimensions that engineering firms can provide concrete answers, with the aim of ensuring that the growth of AI does not come at the expense of available resources. When a data centre project arrives in a territory, it brings with it enormous demands for electricity and water, comparable to the needs of a small town like Lançon-Provence, Le Pradet or Limoux in France. Engineering firms help these industrial clients match need with resource. My own company works, for example, on **finding water supply solutions to cool data centres in regions experiencing growing water stress.** We consistently favour solutions that integrate the reuse of water already drawn for other purposes, such as the discharge from a wastewater treatment plant. This approach, in reality, closes the loop. AI is transforming the engineering profession. And in return, **engineering is what makes sustainable AI development possible.** ## Key Takeaways 1. AI does not create an access gap: it exposes a culture gap, visible even between generations only a few years apart 2. The differentiator will not be the tool itself, but the capacity to align technology, organisation and data governance 3. Moving fast without method generates more risk than value: diagnosis and structured use cases come first 4. The engineer is no longer the person who analyses data; the engineer is now the person who arbitrates, orchestrates and guarantees the relevance of AI outputs 5. Behind every AI system lies a physical infrastructure with real environmental constraints, and engineering is what makes sustainable AI development possible --- *Article from [Albert's Deep Dive](https://deepdive.albertschool.com) - Albert School's Journal*