LOCAL CONTEXTS IN ARTIFICIAL INTELLIGENCE ADAPTATION: HUMAN AND MACHINE UNDERSTANDING OF KAZAKHSTANI NEWS
ЛОКАЛЬНЫЕ КОНТЕКСТЫ В АДАПТАЦИИ ИСКУССТВЕННОГО ИНТЕЛЛЕКΤΑ: ЧЕЛОВЕЧЕСКОЕ И МАШИННОЕ ПОНИМАНИЕ КАЗАХСТАНСКИХ НОВОСТЕЙ
Akhmetov S.S. Akhmetova L.S.
2025MSU Publishing House
Vestnik Moskovskogo Universiteta. Seriya 10. Zhurnalistika
2025#2025Issue 424 - 49 pp.
The ability of generative language models to correctly adapt texts and recognize deep cultural and historical local contexts remains poorly studied. Despite the high quality of generating and translating texts, and analytical outputs, the models still demonstrate limited capabilities for interpreting socially significant meanings and values. These limits are clearly visible in current news items, where the model can produce distorted or downright erroneous interpretations. The purpose of this study is to diagnose the practical applicability of the model in adapting Kazakhstani news for an English-speaking audience unfamiliar with local realities, as well as to classify errors and inaccuracies. During the two-level analysis, the generated ChatGPT texts were compared with experts’ interpretations. In addition to assessing the adapted surface meanings, the tasks included a two-stage analysis of the current deep contexts of Kazakhstani public discourse. Based on the results of comparative analysis, the ChatGPT output data revealed a number of persistent errors, classified into a typology of interpretative distortion vectors. At the same time, authors have recorded unique interpretations, which allows us to assume the prospect of practical interaction with generative models as a tool for an alternative view of phenomena or a source of non-obvious ideas.
artificial intelligence (AI) , ChatGPT limitations , cultural meaning interpretation , Kazakhstan , news adaptation , practical journalism , prompt-engineering
Text of the article Перейти на текст статьи
Doctoral student of the Printing and Electronic Media Department, al-Farabi Kazakh National University, Almaty, Kazakhstan
Doctor of history, professor of the UNESCO Chair in Journalism and Communication, al-Farabi Kazakh National University, Almaty, Kazakhstan
Doctoral student of the Printing and Electronic Media Department
Doctor of history
10 лет помогаем публиковать статьи Международный издатель
Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026