The development of generative artificial intelligence provides new technical conditions for mathematical culture to enter the high school mathematics classroom. Aiming at the problems of scattered mathematical culture, difficult to reveal the source of concepts, and insufficient classroom interaction in traditional teaching, this paper constructed a teaching framework integrating semantic retrieval, intelligent question answering, dynamic graphics and learning behavior collection, and carried out a 12-week quasi-experiment in two parallel classes of senior grade one. The results showed that the post-test score of the experimental group was 84.83, which was higher than 78.14 of the control group. The score of mathematical cultural understanding was 17.42, which was higher than 14.83. The score of transfer word problems was 25.37, higher than 21.98. The frequency of classroom interaction was 3.84 times, higher than 2.46 times, and the task completion rate was 91.60%, higher than 83.70%. The research shows that artificial intelligence can enhance the visualization, interactivity and implementation effect of mathematics culture teaching under the leadership of teachers.
