The "Forum on the Impact of AI Technology on Foreign Language Teaching," jointly organized by Tamkang University's Department of English and the Taiwan ESP Association (TESPA), and co-hosted by the College of Foreign Languages and Literatures, was held on November 20 at 2 PM in I501. Nearly 30 faculty and students participated. The forum featured keynote speeches by Tunghai University Chair Professor of Foreign Languages and Literature One-Soon Her, National Taiwan University Professor of Foreign Languages and Literature Zhao-Ming Gao, Tamkang University Professor of Japanese Chiu-Kuei Tseng, and Tamkang University Assistant Professor of English Chia-Chien Chen, who explored the philosophical and practical classroom and translation applications of generative AI and its impact on teachers and translators.

The opening ceremony began with speeches by Dean of the College of Foreign Languages and Literatures Yi-Ti Lin and TESPA Director Chien-Lung Yang, welcoming participants and emphasizing the university's development strategy of "AI+SDGs=∞" while encouraging a collective focus on integrating AI into teaching. Professor One-Soon Her, delivering an online lecture, began with a quote from Jensen Huang, noting that with the advent of AI, there is no longer a need for programming languages. He pointed out expert predictions that the job of translation would be fully replaced by machine translation within 20 years. "Although the translation industry won’t disappear, only the best will survive," he stated. He criticized Taiwan's overemphasis on English education, which he sees as the greatest crisis in foreign language education. This overemphasis, he argued, crowds out subject learning and other foreign languages, leading to misguided educational directions. He advocated for students to take responsibility for their autonomous learning, deciding what to study and effectively using AI tools to enhance their learning.

Prof. Zhao-Ming Gao offered a different perspective, emphasizing the practical application of AI tools. He noted that 93% of machine learning data comes from English, and with precise prompts, AI tools can significantly improve language learning. He also explained that ChatGPT, as a neural network, can produce varied outcomes depending on the refinement of instructions, unlike tools like DeepL and Google Translate. Guiding students to craft accurate prompts in English can lead to improvements in sentence structure and semantic accuracy.

Prof. Chiu-Kuei Tseng shared her hands-on experience with ChatGPT and the metaverse in Japanese teaching. She emphasized that cultivating "Foreign Language Talents 2.0" requires not just AI but also solid language skills, as "human judgment is needed when AI generates incorrect information." She illustrated this with a video created by AI, featuring a Japanese introduction of Distinguished Professor Chih-Yung Chang from the Department of Computer Science and Information Engineering, which mistakenly had him refer to himself as "sensei" (a term not used for self-reference in Japanese). She also shared her experience of guiding students through iterative writing revisions with ChatGPT, where students praised her final corrections, saying, "Ma’am, you understand me better than AI!"

Asst. Prof. Chia-Chien Chen, with 20 years of experience in subtitle translation and over 800 translated films, provided insights into the current state of the translation industry. Discussing her topic "AI Technology and Subtitle Translation: The Synergy of Practice and Teaching," she advocated for human-machine collaboration. While AI cannot yet fully replace translators, she quoted Jensen Huang, saying, "Translators who don’t use AI will be replaced by those who do." She explained the emerging method of Machine Translation Post-Editing (MTP), where translators edit AI-generated translations. While this approach increases translation speed, with deadlines compressed to 3 days for an eight-hour workday, it reduces fees. She emphasized that human proofreading skills have become even more critical in this new workflow.