[Clinical intelligence and artificial intelligence: a question of nuance].

Current artificial intelligence (AI) in medicine has high performance, particularly in diagnostic and prognostic image analysis, but, in everyday clinical practice, evidence-based results of AI remain limited. In this forum, are analyzed the characteristics of clinical intelligence in medical practice, then the successes and promises of AI, as well as the limitations, reservations and criticisms brought to the introduction of AI in the front-line clinic. The importance of certain ethical and regulatory aspects is highlighted, including a "human guarantee" for AI, as suggested by the "Comité consultatif national d'éthique pour les sciences de la vie et de la santé" (National advisory committee on ethics for life and health sciences). Clinical intelligence could be this "human guarantee" of AI in medicine, and their complementarity could lead to a quality of decisions and, ultimately, of care, far higher than that provided, separately, by each of them.