“SMEs often have the impression that AI is not made for them” (Jean-Philippe Desbiolles, IBM)

Artificial intelligence is the end of the code. The shift from a deterministic world – where when this happens, that happens – to a world of learning “Summarizes Jean-Philippe Desbiolles, member of the steering committee of IBM France and specialist in AI. A paradigm shift that has been taken over by large groups. But for small and medium-sized enterprises, on the other hand, the attitude is still quite wait-and-see.

They feel like it’s not for them thinks Jean-Philippe Desbiolles. Because they imagine these solutions too expensive, too difficult to implement, or because they don’t really see the point. An observation that is valid for France, but also for the South Region of course. And IBM intends to fix it.

Now, AI is quickly accessible from a technical point of view. SThe price has come down a lot compared to ten years earlier, all the more so thanks to the emergence of open-access resources”. Moreover, he thinks, companies are preparing for it. “In our 2022 AI adoption study, we observed that while 23% of representative companies surveyed are already using AI (compared to 30% globally), 37% are investigating the topic to deploy it within twelve month “.

Need for local support

Investigations which consist in particular in finding the right actors to support them. Hence the interest for IBM to strengthen its local roots. This is also the whole meaning of the partnership recently signed with the Riality lab of the Chamber of Commerce and Industry of Aix-Marseille Provence, as explained by Sybille Gérent, responsible for partnerships and in charge of the Provençal territory. ” This partnership is born of human relationships. When we discussed with the CCIAMP when setting up this AI totem place that is the Riality lab, it seemed obvious that as an expert on the subject and a player in the territory, we had all the our place “.

Player in the territory because IBM has R&D teams in Nice and a ” large commercial branch in Marseille “. It has counted local authorities in the territory among its clients. And maintains relations with several higher education establishments such as Kedge, Epitech, the Ecole des Mines in Gardanne, the Isen engineering school and Aix-Marseille University. With the latter, it is trying in particular to contribute to the rise of AI in certain strategic sectors at the local level. ” With AMU, we have targeted two main themes for the territory: health as well as transport and logistics”.

On the Nice side, projects could emerge in connection with the 3IA Institute, IBM already maintaining good relations with the House of Artificial Intelligence.

SMEs waiting for pragmatism

A proximity that rhymes with agility and accessibility for businesses. Within the Riality lab, IBM intends to provide its expertise and show use cases to SMEs so that they understand what AI can offer them. Informally, through moments of exchange within this third place, and above all pragmatically. “ SMEs don’t have a lot of time to devote to it. We must be able to find a solution for them in one day “.

Among the most anticipated applications: natural language with everything related to conversational robots, the digitization of sites, applications… A field that has reached maturity “Judge Jean-Philippe Desbiolles. The latter also observes a growing demand from companies wishing to use AI to profoundly transform their processes. ” This is the case, for example, when a company wants to extract of the data of its after-sales service letters to understand which products or services are malfunctioning”.

“Who says learning says pedagogy”

The challenge is also to overcome certain apprehensions in the face of a learning technology that some fear will see it “replace” humans. Faced with this, Jean-Philippe Desbiolles likes to say that “ AI will be what you make of it “, assertion which moreover titles the book which he published in 2019 with the Dunod editions. ” Who says learning says pedagogy. What model are we creating? What place do we leave to the man? “. For him, it is necessary to build a trustworthy artificial intelligence that is at the same time robust, transparent and explainable. “ And these three floors must be encapsulated in a system of well-defined values “. Clear and explicit values, rather than an ethical requirement by subjective definition.

Also, he thinks, it is essential to position oneself on the desired cooperation between man and machine, and to define upstream whether, in a particular case, it is preferable to refer to the decision-making power alone. of the machine, to that of the man or to bet on a collaboration of the two. ” Take a lawyer, artificial intelligence can analyze all the case law for it, collect different similar cases… But when you have all that, who decides on the pleading? It’s human! “.

There remains a challenge to be met: that of inclusivity in the world of technology: Inclusion is not just a dogmatic issue. If we don’t, if we don’t include people of different genders, sexual orientations, different ways of seeing things, the AI ​​will be inherently biased thinks Jean-Philiippe Desbiolles. Convinced that in this matter, it is essential to involve humans so that AI is not suffered by society, but is the desired reflection of it.