Résumé

Morality plays a critical role both in our own sense of identity and how we form impressions and trust others. Traditionally, work in psychology has focused how trust is influenced by whether someone is moral. In this talk I will present some of my work over the last years showing how trust is not only sensitive to whether someone makes a moral judgment, but which specific moral judgment they make. Focusing on moral dilemmas that exemplify the tension between utilitarianism and deontology, in the first part of the talk I will provide a brief overview of my past work in different contexts and countries showing how utilitarian moral judgments can both erode and enhance trust. Yet while humans have long claimed the preserve of morality, artificial intelligence is increasingly used to perform tasks with a moral dimension, and may soon even act as artificial moral advisors. In the second part of this talk, I will discuss some of my in-progress, ongoing, and plans for future work during my time at the IAST focusing on trust in such “moral machines”.

Référence

Jim A. C. Everett, « Trust in humans and AI from (non)utilitarian moral decisions », IAST Lunch Seminar, Toulouse : IAST, 6 octobre 2023, 12h45–13h45, salle AUDITORIUM 4 (FIRST FLOOR - TSE BUILDING).