Given the rise of the AI era, learning programming languages is no longer exclusive to students in information-related disciplines. However, non-CS major students often encounter challenges when taking programming courses, such as unclear learning objectives, complex course materials, and a lack of learning motivation. The primary aim of this study is to examine the effectiveness of computational thinking in programming learning for non-CS major students by introducing innovative learning strategies. This study employs a quasi-experimental design involving 59 non-CS major students from a university in Taiwan. The findings indicate that the control group (CG) and the experimental group (EG) significantly improved post-test scores across all computational thinking units. Moreover, among the six units, the EG group outperformed the CG group in all but the algorithms unit, where no significant difference was observed. This study suggests that future liberal arts curricula should be tailored to non-CS major students and incorporate more forward-thinking courses and learning methods to align with the talent development needs of the AI era.