Bridging Embedded Intelligence and Multilingual Communication: A Case Study on AI and Human Translation of Wenzhou Asian Games Thematic Materials

This study investigates the comparative effectiveness of AI-generated and human translations of Wenzhou Asian Games Thematic Chinese (WAGTC), focusing on linguistic and cultural complexity. By using ChatGPT-4o and bilingual human translators, 45 samples were evaluated against four dimensions: accuracy, fluency, cultural appropriateness, and idiomatic usage. WAGTC features metaphor, rhyme, parallelism, and vocative functions, posing distinct challenges for both machine and human translation. Results show that AI outputs tend to be literal and structurally faithful but often lack idiomatic flow and cultural nuance. Human translators, meanwhile, favor interpretive strategies that enhance readability and contextual relevance. The study further examines the potential of embedded intelligent systems, leveraging AI-IoT integration and edge computing to improve adaptive, real-time multilingual communication in public spaces. By highlighting the strengths and limitations of each approach, the research proposes a hybrid model that combines AI speed with human creativity to optimize translation quality for culturally loaded, function-sensitive texts. This work contributes to the evolving discourse on human-AI collaboration in cross-cultural translation and introduces a framework for evaluating AI’s role in context-aware multilingual systems.