Key points are not available for this paper at this time.
This study examines how the integration of horizontal and vertical generative AI (GenAI) tools—such as ChatGPT and its specialized extensions—can enhance educational outcomes. Using structural equation modeling on data from 419 U.S. students, we find that both types of GenAI synergy positively affect perceived reliability and flexibility. Horizontal synergy has a stronger impact on reliability (β = 0.279, p < 0.001), while vertical synergy more strongly influences flexibility (β = 0.202, p < 0.001). Perceived flexibility emerges as a stronger predictor than reliability for both learning outcomes (β = 0.394 vs. β = 0.111, respectively) and students’ willingness to pay for membership (β = 0.303 vs. β = 0.214, respectively). The model accounts for 35.4% of the variance in learning outcomes and 29.2% in payment intention. These findings offer practical guidance for educators and institutions in balancing general-purpose and specialized GenAI tools to foster reliable and adaptable learning environments.
Lee et al. (Thu,) studied this question.