Conoscere quello che non si vede : i trasformatori e l'eccezione informatica

Computer Science, because of methodological and factual problems, does not quite conform to the well-established scheme of an empirical science. Such an anomaly is contrasted with examples drawn from the development of some classical branches of knowledge such as microbiology, thermodynamics, and quantum mechanics. A problematic situation has recently surfaced in so-called Artificial Intelligence: its enormous success in applications is not backed up by a theoretical understanding of the attendant software. A new line of research, named Explainable Artificial Intelligence (XAI), has come into being and a brief sketch thereof is given. [Publisher's text].