Ethical Consideration for Designing AI to Support Dynamic Learning Transitions
15th International Conference of the Learning Sciences
International Society of the Learning Sciences
Since individual and collaborative learning have different strengths, one way to leverage the benefits of both is by allowing students to dynamically switch between the two modes based on their needs at the moment. Managing dynamic transitions as students simultaneously work on different tasks is an ambitious learning practice because it entails high orchestration load for teachers and careful personalization of learning transitions (Uttamchandaniet al., 2021). To support teachers, we investigated how AI and students can have a voice in deciding when these transitions occur and how (Echeverria et al., 2020). However, designing orchestration tools with shared participation requires explicit reflection on the equitable and ethical considerations regarding how these parties’ agency, bias, and privacy are accounted for in the tool (Holmes et al., 2021). The contributions of our paper include findings regarding how teachers want to share control with students and the AI during dynamic transitions.We share challenges to consider when leveraging AI in ambitious learning practices including navigating tensions of agency, considering privacy and transparency, and reflecting on biases.
Lawrence, L., Rummel, N., & Aleven, V. (2022). Ethical consideration for designing AI to support dynamic learning transitions. Symposium conducted at the 15th Annual International Conference of the Learning Sciences. International Society of the Learning Sciences. Hiroshima, Japan.