Discover our series of conferences on AI in the construction industry
The meaning of AI
When AI and training are used in the same sentence, the first challenge lies in establishing correctly the definition. Many forms of artificial intelligence exist (like generative or non-generative, for example), but not all are created equally. What’s more, AI remains a continuation of traditional disciplines, largely integrated into education curricula or business-related hard skills. “There was a huge rush towards data science, then to Big Data, and now it’s all about artificial intelligence, but there is a certain continuity and technical overlap between these disciplines,” explains Ons Jelassi, director of Télécom Paris Executive Education.
When the stars align
The current craze for AI should therefore not be attributed to a fundamental break in knowledge. For François Lemaistre, CEO of Axians, it’s all down to a unique planetary alignment: “Algorithms have become infinitely more sophisticated, training data and computing power are available,” he explains. This situation is leading to a proliferation of uses for AI, which is itself at the origin of the skills race that has begun on a global scale. In France, the State has committed €360 million to nine universities and research centers, with the aim of training 100,000 people annually.
AI everywhere?
But despite political will, the issue of training and skills remains a thorny one. Indeed, AI is not a monolithic discipline. Like how digital technology brought about transformation in recent decades, AI is infiltrating all professions and transforming all training courses. “It’s clear that data sciences are permeating all sectors. Networks are now virtual, automation is the norm, and there’s a need for AI model security. There is no research chair within the university without an AI component,” explains Ons Jelassi. As it permeates all sectors, AI’s extremely porous nature throughout all sectors can also be witnessed on the hard skills side of things. In this context, all professionals need to advance and develop. This is one of the reasons that pushed VINCI to make the firm decision to tackle the issue in-house. Right now, a Leonard-incubated company like DIANE aims to develop AI projects across all VINCI Energies businesses. “AI has become a way to catalyze technical skills to accelerate costing or response time to tenders or research,” explains Stéphane Maviel, head of the company.
Agility the competitive edge
At a time of rapid development and extreme porosity, there is no ideal path for AI training. “What we’re expecting from our young employees is for them to have malleable minds and the ability to teach themselves skills,” explains Maviel. Similarly, education courses do not offer a best route and must be adapted to each situation. “A small proportion of our students choose to do a thesis. Some young graduates are interested in consulting firms, others opt to set up a business,” explains Jelassi This plurality is particularly conducive for a partnership mindset, enabling interesting synergies to emerge. Stéphane Maviel highlights the interest in associating fundamental research with cutting-edge topics. Meanwhile, Lemaistre stresses just how fruitful partnerships with higher education can be. “For higher education, concrete cases have value. For us, it’s a way of maintaining our skills. There is a consistency in the approach.”
Practical AI for the construction industry
In addition to general issues surrounding AI, there are also sector-specific challenges too, which raise a fundamental question: should there be a preference for pure AI specialists or would it be better to train business experts in artificial intelligence? While the conference speakers weren’t able to provide a definitive answer, they did all agree on one point: artificial intelligence is a distraction if it does not provide effective solutions to real problems. “At VINCI, we don’t do artificial intelligence, we solve problems,” explains Lemaistre, rather provocatively. Or rather, this translates to AI applied specifically to construction, more operational and used closer to work in the field.” The construction industry needs competent data scientists, but it also needs data scientists who are immersed in the problems they must solve, data scientists who must be aware that their time will be composed of 5% inspiration, 20% data science and 75% perspiration,” concludes Lemaistre.
