Abstract With the assistance of DeepSeek, the “Chemical Snake” game was developed based on the cognitive load theory, flow theory, and zone of proximal development theory. The game content covers chemical terms, chemical concepts, chemical laws, and chemical reactions. In the game, students control the “snake” to move via the keyboard and make it grow by devouring food that meets the requirements. The dynamic feedback mechanism caters to personalized learning needs. To achieve the serialized development of game knowledge, a method for game development by building game systems is proposed, which avoids the randomness in the quality of games generated by GenAI to a certain extent. A survey was conducted among 30 eighth-grade students in Nanjing who had experienced the “Chemical Snake” game, the game significantly stimulated interest and received high ratings on design, content, playability, usefulness, and scalability.
MO Zhi-Rong, LING Yi-Zhou, HENG Hui, REN Hong-Yan. Theory and Method of Generative Artificial Intelligence DeepSeek-Assisted Serialized and Efficient Development of Chemistry Games[J]. Chinese Journal of Chemical Education, 2026, 47(3): 78-83.
