The FAIR – Future Artificial Intelligence Research project aims to help address research questions, methodologies, models, technologies and also the ethical and legal rules for building Artificial Intelligence systems capable of interacting and collaborating with human beings, to perceive and act in constantly evolving contexts, to be aware of their own limits and capable of adapting to new situations, to be conscious of the perimeters of security and trust, and to be attentive to the environmental and social impact that their realization and execution may involve.

The research activities will take place through 10 spokes which will involve more than 350 researchers. Each spoke is characterized by a specific thematic area and its own set of research challenges, with the aim of addressing the challenges of FAIR from different points of view.

Prof. Nicola Gatti of the Department of Electronics, Information and Bioengineering of the Milan Polytechnic will coordinate spoke 4, dedicated to Adaptive AI.

Adaptability has been one of the primary goals of AI since the beginning and is now pervasively required not only in the virtual world but also in the physical world. In particular, adaptivity is linked to the ability of an entity to interact with the environment, to perceive the context and the related information that changes over time and to act promptly. In particular, so far machine learning has emerged as the main key enabling technology to design adaptive artificial systems. Despite its ubiquitous adoption, there are serious gaps in our theoretical understanding of how learning systems can provide assurances and how they can be effectively combined with other AI paradigms when designing adaptive AI systems.

Challenges in this area include: the foundation of learning theory for adaptivity, understanding non-convex optimization for deep learning models, and studying remarkable environmental structures that can be used to provide better assurances of adaptability.

CNR is the proponent of the project, in collaboration with the CINI AIIS (Artificial Intelligence and Intelligent Systems) National Laboratory. FAIR is a non-profit foundation and its President is Giuseppe De Pietro, director of the CNR Institute for High Performance Computing and Networks (CNR-ICAR).

The partnership, which for the first time brings together the national scientific network of AI experts also involving a first nucleus of companies in the sector, is made up of four research institutions (in addition to CNR, Fondazione Bruno Kessler, INFN, Istituto Italiano di Tecnologia), fourteen universities (Politecnico di Milano, Politecnico di Torino, Sapienza Università di Roma, Scuola Normale Superiore, SISSA, Università Bocconi, Università Campus Biomedico di Roma, Università di Bologna, Università di Pisa, Università di Trento, Università di Bari, Università della Calabria, Università di Catania, Università di Napoli “Federico II”) and seven companies (Bracco, Deloitte, Expert.ai, Intesa Sanpaolo, Leonardo, Lutech, STMicroelectronics).