Abstract
Now that artificial intelligence is available to all, it has changed how we learn online. Now many excellent systems that use artificial intelligence are available to provide personalised learning paths, intelligent tutoring, etc. Learning stickiness is the long-term intention of students to use a learning platform and stay engaged in learning over time, so it has become an important indicator of the sustainability of online education. Based on the experience of flow and algorithm trust, a two-way theoretical model is put forward in this paper to explore how the empowerment of artificial intelligence affects the stickiness of students' online learning in different ways. According to the theories of flow and trust, a total of 426 students were selected as the subjects of the questionnaire survey on the AI online learning platform in this paper. According to the results of the experiment, both the flow experience and trust in the algorithm were found to be moderators of the effect of AI empowerment on learning stickiness. The road of experience and flow is relatively more popular among adolescent students, and at the same time, the rational path of algorithm trust is used more frequently by adults. Personalisation of the platform, interactive feedback and data transparency are the three reasons for the appearance of the two mediating states. This paper studies the two motives behind AI-enabled online learning and offers some practical optimization ideas for the Design of learning platforms, Algorithm improvement, etc.