Leveraging Artificial Intelligence for International Accessibility to Moroccan Linguistic Heritage: The Case of Translating Doukkala ’s Ancient Proverbs

Abstract

This validation study examines the role of AI, notably the ChatGPT-3 tool, in promoting Moroccan linguistic heritage. Specifically, the study investigates the usefulness of the ChatGPT-3 tool in accurately translating ancient proverbs from the region of Doukkala. The objective is to facilitate the understanding of proverbs on a global scale. The study experiments on 8 indigenous proverbs randomly selected. They are automatically translated through ChatGPT-3 into English using two major prompts (translation and explanation). Generally, the results reveal the potential of AI-powered ChatGPT-3 in facilitating accessibility to Moroccan linguistic heritage. ChatGPT-3 demonstrates proficiency in translating Doukkala’s ancient proverbs, enabling their understanding through prompted explanations. In this process, ChatGPT-3 contextualizes proverbs by analyzing their broader semantic framework and integrates cultural, historical, and social variables to contextualize the complexities of these expressions; thus, ensuring culturally resonant translations. However, the study also reveals limitations in ChatGPT-3's ability to preserve implied cultural concepts. This highlights the importance of continuous refinement and human intervention to ensure the preservation of cultural context and authenticity in AI-generated translations.

Keywords: AI, ChatGPT3, Moroccan linguistic heritage, Doukkala region, proverbs translation, cultural context