Hermeneutics of suspicion and affirmation: a study on the production of meaning from the film the secret agent in interaction with artificial intelligence and its educational implications

Abstract

This article investigates processes of meaning-making prompted by the film The Secret Agent (2025), directed by Kleber Mendonça Filho, taking as its point of departure an initial mistaken interaction between the author and ChatGPT. From this interpretive error—wherein Artificial Intelligence (AI) assigns the protagonist a role that the film deliberately refuses—the text advances a hermeneutic reflection on how meanings are constructed, stabilized, and may be placed under suspicion. Methodologically, the study adopts a hermeneutic-experimental approach in which interaction with a generative language model is understood as a concrete interpretive event, capable of revealing cognitive shortcuts, archetypes, and statistical biases mobilized in the production of narratives. The article connects this experience to the hermeneutic tradition, especially Paul Ricoeur’s hermeneutics of suspicion, in dialogue with Nietzsche and with the possibility of a hermeneutics of affirmation, linked to pathei mathos and formation through experience. The distinction between mathema and pathei mathos enables a critical examination of AI’s limits in the production of meaning, emphasizing that, although capable of simulating narratives and revising them, AI does not partake in the existential and formative dimension of human experience. The article concludes that both the film and the interaction with AI function as pedagogical dispositifs that invite the suspension of beliefs, distrust toward imposed meanings, and the affirmation of experience as a foundation of human formation.