Theory-driven design of AIED systems for enhanced interaction and problem-solving

Successful AI learning environments are driven by effective designs for learning that incorporate what we know about how learners learn best. Theory-driven designs about the cognitive, metacognitive, affective, motivational, and behavioral components of learning need to be considered jointly in designing effective AI interfaces for learning and instruction. We provide examples of AI interfaces that embody specific theories. We discuss the benefits of AI-supported interfaces along with AI developments that enable new functionalities and modalities of AIED systems. We conclude with suggestions for future interface considerations.