National Science Foundation AI Institute for Student-AI Teaming (iSAT)

• In Event: Cross-Cutting Research Across Three NSF (National Science Foundation )-Funded National AI Institutes on Education

Abstract

Objectives or purposes

The NSF AI Institute for Student-AI Teaming (iSAT) addresses the challenge of how to promote deep conceptual learning via rich socio-collaborative learning experiences for all students. iSAT’s vision is to develop AI Partners – intelligent technology which collaborates with teams of students and their teachers to promote development of AI literacy, STEM competencies, and 21st century skills including collaborative problem solving and critical thinking while empowering diverse youth to become participants and leaders in the AI-driven workforce of the future.

Perspective(s) or theoretical framework

Our research is guided by a range of theoretical perspectives including academically productive talk (Michaels & O’Connor, 2015), collaborative problem solving (Sun et al., 2020), cultural-historical activity theory (Lecusay et al., 2008), roles of ideologies in/of learning (Philip & Sengupta, 2021), interactive team cognition (Cooke et al., 2013), and situated grounding (Clark & Brennan, 1991).

Methods, techniques, or modes of inquiry

We adopt the framework of responsible innovation, which “means taking care of the future through collective stewardship of science and innovation in the present” (Stilgoe et al., 2013) to guide all research and development. Accordingly, we use co-design as a broad strategy (Severance et al., 2016) to engage K–12 teachers, diverse students, parents, and other community stakeholders in the design of AI Partners for the classroom.
Data sources, evidence, objects, or materials
We held a Learning Futures Workshop (LFW) that engaged 30 diverse high school-aged youth from California and Colorado over 10 days to deliberate the role of AI in supporting collaborative learning. We followed up with a three-day immersive design sprint to identify metaphors for the AI Partner that were responsive to students’ needs, desires, and concerns.

Results and/or substantiated conclusions or warrants for arguments/point of view

The LFW indicated that the youth consistently expressed their desire for affirming interactions from an AI Partner. However, they did not want the AI Partner to report any poor behavior to their teacher. The youths also wanted agency over the AI Partner, but they felt they would be willing to “give up” data about themselves in exchange for features they value. Building of these insights, the design sprint converged on three metaphors for an AI Partner: (1) Community Builder, which supports students in developing trusting collaborative relationships with each other; (2) Augmenter, which support teachers and students by helping relay information exchanges through an interface; and (3) Interactive co-pilot, which supports task progress and collaboration by providing guidance and facilitating small group interactions.

Scientific or scholarly significance of the study or work

The LFW and subsequent design spring provided a major reconceptualization of the possible roles of an AI partner based on youth input. These new ideas helped to shift the Institute’s focus away from the idea of an AI partner a top-down “monitor” of collaborative learning and informed the new student-driven metaphors for the AI Partner. Our work illustrates how inclusive co-design processes can empower stakeholders with diverse identities to envision, co-create, critique, and apply AI learning technologies for their schools and communities.

Authors

• Thomas Breideband, University of Colorado - Boulder

• Quentin Biddy, University of Colorado - Boulder