The Artificial Intelligence (AI) is transforming our daily lives, and this transformation raises fundamental questions about humanity and the goals of civilization. The AI&ART Pavilion is a multidisciplinary initiative that aims to reflect this emerging future in an interactive context by combining the industrial heritage of Belval with the scientific ambitions of the university. The Pavilion provides a space where scientists, artists, and the public can come together to exchange ideas and find a sense of community in the face of contemporary and forthcoming challenges.
Explainable AI (XAI) techniques are intended to explain machine learning models to humans or other intelligent virtual entities, but they face challenges in real-world settings such as distributed systems, heterogeneous data, bounded knowledge, and privacy concerns. To address these challenges, there is a need for personalized XAI that combines inter-agent, intra-agent, and human-agent interactions to integrate heterogeneous knowledge/explanations extracted from efficient black-box AI agents. The text also discusses a specific project that aims to develop and validate a personalized XAI approach in the domain of food and nutrition monitoring and recommendation.
