This article explores the issue of language choice when users interact with AI, focusing on the differences in interaction experience between Chinese and English. The author shares personal observations: OpenAI may have exhibited “diminished intelligence” for Chinese users in its early stages, and asks if Gemini has similar issues. Based on actual usage, English Q&A has shorter thinking time and higher efficiency, possibly due to the relatively lower quality of Chinese language corpus. In community discussions, multiple participants shared similar experiences, triggering reflections on language bias in AI models and the quality of training data. Although this is a forum post, it provides valuable industry insights, revealing the actual challenges of AI language processing, and has reference value for optimizing AI interaction experiences.
