Scientific literacy is a key competency needed in the era of Smart Society 5.0. However, studies indicate that the scientific literacy of Generation Z remains low. This can be attributed to the implementation of learning models that are less relevant to the characteristics of their generation, commonly referred to as digital natives. Furthermore, an approach that can make in-depth learning meaningful yet enjoyable is also needed. One possible alternative is to develop an AI-based learning model using a deep learning approach. This study aims to develop a valid and practical Deep-AI learning model to improve the scientific literacy of Generation Z. The method used was research and development (R&D) with a model developed by Borg and Gall (limited to step seven). The model was validated by three education experts and tested on a limited sample of 25 Generation Z students. The results of the study stated that a) the Deep-AI model is valid based on an average validation score of 3.69, which is categorized as very valid, and b) the Deep-AI model is practical based on the average observation result of model implementation of 3.68, which is categorized as very good. Therefore, it can be concluded that Deep-AI is valid and practical. This model can serve as an alternative learning approach aimed at enhancing the scientific literacy skills of Generation Z students.
Copyrights © 2025