Cognitive and discursive models of the ambivalent representation of artificial intelligence in American media discourse

Abstract

This article examines the ambivalent representation of the concept ‘artificial intelligence” in American media discourse. The relevance of this study stems from the need to identify the mechanisms underlying the formation of an ambiguous axiological and ambivalently interpreted image of artificial intelligence amid rapid technological development and its active mediatization. The aim of this study is to identify and describe the main cognitive and discursive models of the ambivalent representation of artificial intelligence, as well as to determine the linguistic means through which the positive and negative components of its media-discursive image are conveyed. The research material consisted of authentic texts from contemporary American media discourse devoted to artificial intelligence, its technologies, capabilities, limitations, and the social consequences of its application. The study employed a set of methods, including descriptive, contextual, discursive, cognitive, semantic, component, axiological, and interpretative analysis, as well as elements of quantitative analysis. The study found that the ambivalent representation of artificial intelligence is formed through stable binary oppositions that differ in the degree of polarity and mutual exclusivity of their value-based characteristics. Four main cognitive and discursive models of the ambivalent representation of artificial intelligence in American media discourse were identified: opportunity – threat, helper – competitor, control – independence, and enhancement – displacement. It was found that these are primarily expressed linguistically through lexical units with positive and negative connotations, predicative constructions, anthropomorphic verbs, and metaphorical usages. The syntactic position of artificial intelligence plays a special role: presenting it as the object of an action emphasizes the controllability of the technology, whereas functioning as the subject of a predicate contributes to its conceptualization as an independent actor. The results obtained allow us to view ambivalence as a systemic property of AI’s media representation in American media discourse, which is significant for further cognitive and discourse studies, as well as comparative corpus studies in various languages.