ARTIFICIAL INTELLIGENCE: EDUCATIONAL TENSIONS BETWEEN TRUTH AND DATA VERISIMILITUDE

Elements of analysis that are operative for a critical incorporation of generative and non-generative artificial intelligences (AI) are pointed out. Approaching a balanced digital pedagogical realism in which AI can enhance individual and collective educational spaces instead of generating overproduction. The focus is on relevant pedagogical reflections, taking into account the regulatory needs of AIs. AIs are characterized, starting with a brief analysis of the social, educational and economic environment in which they are widely disseminated. The necessary framework is presented to examine various tensions that they incorporate into the educational ecosystem, which have their origin in the truth-verisimilitude-fiction value provided by data. Four spaces of tension provoked by the massive advent of AIs in education are discussed: authorship, verisimilitude spiral (or truth-value), fiction and literariness, datafication and agency. As a result, it emerges that it is appropriate to consider the impact of AIs on teaching-learning processes in terms of the more or less distant relationship they may have with material reality: their ability to create fictional narratives based on real data.