The Algorithmic Turn: Assessing AI's Role in Shaping Intercultural Narratives and Semantic Transfer

Abstract

The rapid development of Neural Machine Translation (NMT) and Large Language Models (LLMs) marks what can be described as an Algorithmic Turn in translation studies. This shift fundamentally reconfigures translation from a primarily human-centered act of linguistic mediation to a process increasingly shaped by computational systems and algorithmic logic. This paper critically examines the growing influence of Artificial Intelligence on intercultural narratives and on the process of semantic transfer in an interconnected global context. While AI-driven translation technologies offer unprecedented speed, scalability, and accessibility, their widespread adoption also raises significant concerns related to cultural authenticity, linguistic diversity, and the preservation of meaning. Drawing on a critical-theoretical framework, the study argues that AI systems function not merely as neutral tools but as active co-authors in the construction of cross-cultural meaning. The analysis focuses on two central issues: first, the ways in which intercultural narratives are shaped by AI models trained on culturally imbalanced datasets, often privileging dominant linguistic norms and contributing to cultural homogenization; and second, the inherent limitations of AI in handling semantic nuance, particularly in relation to irony, implicit power relations, and culturally embedded meanings. The paper ultimately contends that the Algorithmic Turn produces a paradoxical outcome—a translation that is linguistically fluent yet culturally hollow. Addressing this paradox requires moving beyond accuracy metrics toward a critical evaluation of AI’s ethical, cultural, and epistemological implications, reaffirming the essential role of the human translator as an intercultural mediator whose expertise complements, rather than competes with, technological innovation.