Artificial intelligence and big data are two subjects that are being discussed simultaneously. Along with robotics, they fire the imagination of our contemporaries. Of course, the cohabitation of humans with artificial systems has long been a favorite theme of science fiction. As for economics, in the past, these disciplines have alternately aroused exaggerated hopes and disappointments, in an almost cyclical fashion. But a number of recent events, often skilfully staged, suggest that an era of technological breakthroughs is dawning once again.
This is why A.I. and big data are now the subject of a real social debate. In his new white paper, Jean-François Marcotorchino*, aided by Christophe Faurie**, a specialist in change management, takes a look at the real and reasonably foreseeable impact of A.I. and big data on the economy and business.
From a scientific point of view, current A.I. and Big Data technologies increasingly rely on high-level mathematical algorithms and increasingly powerful machines. But the novelty and disruptive prospects are far from being as radical as some would have us believe. So it's important to distinguish between progress in the process of being validated, and the rediscovery of pre-existing, little-used techniques, or between real, validated scientific breakthroughs and the "makeover" of already-known approaches. It is this concern that drives this reflection: what are the benefits of A.I. and Big Data for the economy and business?
It leads to a practical conclusion: what we currently hear about A.I. and big data does not reflect reality.
On the one hand, if there is a case for radical transformation, it's not where we're looking for it. And it won't have the impact we're talking about. It is, in a way, "democratic". It smiles on human or social talent, and doesn't cost much.
Secondly, we've sold the bear before we've killed it. A.I. and big data don't have the capabilities we're giving them credit for. They face obstacles. More precisely: 5 "antinomies". However, a breakthrough is not impossible. It's a mathematical problem. In France, we have excellent mathematicians and statisticians. It's just a question of keeping them here...
*Jean-François Marcotorchino is Associate Director of Research at the Laboratoire de statistique théorique et appliquée (LSTA) at Paris VI/UPMC. University professor, lecturer at Paris VI and ISUP (Institut de Statistique de l'Université de Paris). Former vice-president, scientific director of Thales SIX, and Thales TechnicalFellow. Previously, he was Director of the IBM Scientific Center in Paris and of IBM Europe's European Centre for Applied Mathematics (ECAM). He is the creator of a new branch of mathematics: mathematical relational analysis.
**Christophe Faurie is a specialist in change management, a subject on which he has published 5 books (most recently: J'ai pensé à tout... et pourtant ça ne marche pas, published by Kawa). A graduate of the Ecole Centrale de Paris, Insead (MBA) and Cambridge University (M.Phil), he began his career in the strategy department of Dassault Systèmes, then went on to become Marketing Director of the Norisko Group (Dekra), Director of Consulting at MV2, and partner at management consultancy GM2.
Learn more: