Abstract Recent advances in artificial intelligence (AI), particularly large language models, have accelerated structural transformations in translation, posing significant challenges to both professional practice and training in translation and interpretation. A literary case study—Stephen Pearl’s English translation of Oblomov—illustrates translation skills that remain resistant to automation. Although translation production is being reorganized and professional roles reshaped under increasing technological and economic pressure, AI-driven systems continue to exhibit structural limitations rooted in non-embodied, statistical language processing. These constraints underscore the continued necessity of human linguistic, cultural, and interpretive expertise. Against this background, the concept of a “translation factory” is introduced, conceptualizing translation as an integrated workflow in which human experts and automated systems collaborate across multiple stages, including text profiling, preparation, translation, quality assessment, post-editing and updating. While the overarching objective of this model is to automate as much of the workflow as possible, its successful implementation depends on highly skilled linguists with deep knowledge of language, culture and AI. Emerging human roles include translation workflow designers, technical specialists, prompt engineers and AI translators. The AI-driven transformation in translation, therefore, entails not the replacement of human translators, but a reconfiguration of professional roles in which human expertise remains indispensable.
François Massion (Thu,) studied this question.