Last November, the Observatory of the Technological, Economic and Social Impacts of Artificial Intelligence (OTESIA) was launched in the Alpes-Maritimes. A first in France and Europe. The first four projects were officially launched and cover areas such as health, education, cyber-hate and cyber-harassment, as well as business skills.

1st project: “AI & Health” survey

This project is led by Jean-Marc Bereder (CHU Nice) and Frédéric Prate (Université Côte d’Azur / CHU Nice). It involves researchers Lise Arena (Groupe de recherche en Droit, Economie et Gestion – Université Côte d’Azur/CNRS), Gerald Gaglio (Groupe de recherche en Droit, Economie et Gestion – Université Côte d’Azur/CNRS), Jean-Pierre Merlet (INRIA Sophia Méditerranée).

The medico-social sector, which plays an important role in caring for the users of our health care system, is intended to be integrated within an overall vision of health, well-being and aging.

As a major mechanism for supporting the elderly and their families, residential institutions for the elderly (EHPAD) could be the main beneficiaries of digital technology, both for the reorganisation of the care offer (exchanges with health structures and skills) and for the quality of life of residents and the management of human and material resources.

However, as illustrated again very recently by the COVID-19 epidemic, the situation of HPAHs is very heterogeneous and attests to a digital divide. It is necessary to decipher the multiple dimensions and impacts of the use of digital technology in medical and social institutions: levels of acculturation and training of professionals, obstacles and levers to the adoption of digital tools, including artificial intelligence, the link between care and digital technology, cooperation and coordination with other actors in the health system, identification of risk situations and anticipation of needs.

2nd project: Prevention of cyber-harassment and cyber-hate

This project is led by researchers Serena Villata and Elena Cabrio (I3S / Inria team) and Catherine Blaya (Université Côte d’Azur- CAPEF-INSPE).

Cyber-hate on racist, xenophobic or religious grounds is on the increase and 6% of young people in Europe are victims of cyber-harassment. A study in Europe shows that more than one in three young people are exposed to hate content online, with victims ranging from 5% to 13% (Machackova, Blaya, Bedrosova, Smahel & Staksrud, 2020). Cyber-spaces are important places of exchange, but also places of risk and violence, with consequences for individuals but also for communities in terms of social cohesion.

There are 4 types of prevention: legal, educational, technological and communicative. No single method implemented in isolation gives convincing results and it is important to combine approaches for greater effectiveness.

There are already tools that use AI and allow the automatic detection of violent, racist speech, insults, etc.; but most of these tools are in English, do not integrate cultural and contextual dimensions, do not detect allusions, implicitness, humour.

The project will cross the approaches of mediation / remediation: it is planned to develop a software to detect hate messages based on an analysis of natural language, but also to understand their argumentative structure (not a simple detection of isolated words, insults) and to develop the critical spirit of the victims and therefore, to do this, to develop a counter-speech. Hence interventions in 6 secondary schools for role-playing games that will serve as a basis for data collection. This data, once analysed, will support the development of software to detect hate and violent speech online. As part of a restitution of the work carried out, the institutions will participate in a collaborative manner in the development of counter-speech.

The research will take place over a period of one year with data collection planned between September 2020 and January 2021, to be modulated according to the health context. The analysis of the data as well as the production of counter-speeches are scheduled throughout the year. The research team will carry out a restitution and demonstration to schools. Finally, a presentation of the research results is planned for June 2021.

3rd project: Artificial Intelligence Devoted to Education (AIDE)

The project is led by researchers Margarida Romero (director of the Laboratory of Innovation and Digital for Education (LINE), University Côte d’Azur) and Thierry Vieville (LINE, Inria) and by Lisa Roux, PhD in computer science, who is carrying out a postdoc on the creation of an ontology on the CreaCube task, who has just joined the team.

The AIDE project has the ambition to study the development of computer thinking in a double approach: creative and critical thinking towards digital. The objective of the AIDE project is the modelling of creative problem solving tasks such as CreaCube. Under a computational learning sciences approach, a modeling of both the learning subject and the task is developed. The modeling will allow an analysis of the behavior of the subject of the experiment in front of the task to be executed through the capture of his gestures. The project uses an ontology of the tasks to be performed and the model of the brain in computational neurosciences (INRIA Mnemosyme team).

The first action in the framework of the project is the constitution of a digital working group (GTnum) within the framework of the call made by the Direction du Numérique Educatif (DNE) of the Ministry of National Education. The GTnum Scolia: Renewal of digital practices and creative uses of digital and AI aims to identify and develop practices and creative uses of digital for an education adapted to the challenges of contemporary societies by putting the future of the critical and creative citizen at the center of its reflection. The activities should strengthen the capacity to engage and develop the attitude, strategies and social skills contributing to problem solving (at the individual, collaborative and social levels).

In this perspective, the WGnum is part of an emancipatory approach, aiming to develop the knowledge and know-how of students, teachers and trainers. Critical thinking, creativity and computer thinking are key competences both in the education of students and in the training of teachers.

In 2006, Wing named “computational thinking” the ability to use computer processes to solve problems in any field. To develop it, learners (starting in kindergarten and at all ages) can combine learning the computer concepts and processes that are the subject of “digital literacy” (object, attribute, method, design pattern, etc.) with creative problem solving using computer concepts and processes (Romero, Lepage, & Lille, 2017).

On the other hand, the OECD (2013) emphasises the importance of problem solving in technology-intensive environments, which represents the interface between digital skills and problem-solving abilities. Taking into account the current educational challenges and the emergence of AI, the Villani and Torossian report (2018) points out the need for a more critical and creative education, but also for an approach more oriented towards computer thinking so that citizens (young and old) can acquire digital literacy by considering the human factor in the modelling and creation of artificial systems, the basic functioning of algorithms and machine learning, or by adopting a critical judgement in the face of the limits of the solutions produced by AI (Romero, Alexandre, Viéville&). Giraudon, 2020).

For enlightened citizenship in the digital age, we need to continue sharpening our critical, creative, collaborative problem-solving thinking, while adding a new key competency in the AI era: the development of computer thinking.

The AIDE project will result in a modeling ontology for both the learning subject and the task, but will also deliver a public document that will help to raise awareness of the importance of task modeling for the advancement of computational educational sciences.

More information :
CréaCube Task: https://creamaker.wordpress.com/
Action AIDE: https://team.inria.fr/mnemosyne/fr/aide/
GTnum Scolia : https://frama.link/GTnum-Scolia

4th project: Impact of Machine Learning on Skills

The project is led by researcher Edward Lorenz (Groupe de Recherche en Droit, Economie et Gestion – GREDEG – Université Côte d’Azur – CNRS) with a team of 3 people, Cécile Cézanne, Nathalie Lazaric and Eve Saint-Germes also from GREDEG.

This project has the ambition to go beyond the limits of current research. Case studies are planned in three sectors in the Nice and Sophia-Antipolis region: the health sector, transport and logistics, and aerospace.

A survey with open-ended questions will be addressed to all types of staff: why and how do machine learning and related new technologies fit into the agent’s usual tasks? Do they replace skills within the company? Do they diminish them? Are they increasing them? Conceptually, the project aims to contribute to an understanding of the learning processes by which companies develop new technological capabilities.

The preparation and testing of the interview guide will take place until September 2020. They will be followed by stakeholder visits and interviews between September 2020 and April 2021 before feedback is given to participating companies, local decision-makers and members of the academic community.

You can follow the evolution of the projects on the OTESIA website as well as the news of the observatory: https://otesia.eu

Translated from L’OTESIA lance ses 4 premiers projets IA dans la santé, la prévention du cyber-harcèlement, l’éducation