Painting with Words: AI-Assisted Learning in the Assessment of Vocabulary in ESP

Since the rise of artificial intelligence, AI tools have been widely used in language learning and teaching. In ESP courses, the assessment of vocabulary comprehension and application is crucial; however, students are hardly motivated in the learning process. The research aimed to explore the application of AI image generators in assessing the comprehension of technical terms in ESP programs and facilitating vocabulary learning. The action research was conducted in an elective ESP course with eight college students. The data were collected from questionnaires, in-depth interviews, class observations, and text analysis. The results indicated that the students spent more time communicating with AI image generators to depict the image they were required to generate. They gradually became immersed in the flow while viewing more generated images and were motivated by their curiosities for further revisions. Besides, more technical terms were mentioned for picturing the precise images, and longer sentences were constructed for revising the detailed textures. By presenting their artworks in class, the students explained and illustrated their understanding of technical terms with the final generated images, and the teacher assessed their vocabulary comprehension and application visually. However, because AI image generators tried to understand the typo errors and fragments of incomplete prompts so that some students made grammatical errors and did not realize their mistakes in English writing. Further research is recommended for developing formatted prompts which could help students write in correct sentence structures and scaffold them into independent writers in English.