Persuasive Surfaces and Calculating Machines. A Rhetorical Perspective on Artificial Intelligence

Abstract This essay examines the communicative implications of generative AI through a rhetorical lens. Rather than asking whether machines can truly ‘think,’ this approach considers how generative AI’s probabilistic outputs construct persuasive textual ‘surfaces’ that interact rhetorically with human communication. Authorship and meaning are no longer localized to a single individual creator but emerge through the interplay of human and machine co-construction of ‘possible worlds.’ By viewing generative AI as a ‘techné’ in the classical sense, we examine how models ‘imitate’ prior texts through algorithmic recombination. As AI accelerates information flows, it becomes increasingly important for individuals to develop rhetorical literacy, to deconstruct the interests and values encoded in AI’s persuasive surfaces. In contrast to the influential but one-directional information-theoretical communication model proposed by Shannon (1948), the essay advocates a rhetorical model of communication that emphasizes the active role of readers in reconstructing possible worlds from textual offers. Rather than neutral vessels of truth, generative AI outputs are inherently contingent yet interest-driven, requiring enhanced communicative competencies to navigate their impacts on changing concepts of reality, authorship, and common ground.