The notion of truth put to the test by artificial intelligence

Can we apply traditional notions of truth and error to the artificial production of statements? Starting from a tripartition proposed by Umberto Eco, the article questions the contemporary relevance of the criteria of truth used to evaluate human statements. It is a question of noting that artificial statements are not true or false like those produced by humans, which calls into question our conception of language. The article identifies the specificity and the limits of automatic language generation in the inability of machines to anchor themselves to reality through perception. It concludes by indicating the potentialities of mechanical enunciation and names "humility", the attitude, opposed to the current concept of explicability, that one could adopt towards artificial intelligences.