MindScribe: Toward Intelligently Augmented Interactions in Highly Variable Early Childhood Environments

Early childhood is a period of critical development, with impacts that can last a lifetime. And inequities in the quality of care for this vulnerable population---especially for those at-risk due to disability, family income, or trauma---can perpetuate further downstream health and school-readiness effects. Technology-enabled solutions have the ability to bridge quality-of-care gaps by intelligently augmenting daily activities. However, many traditional computational approaches to natural language interactions are not yet feasible nor afford-able in highly variable and dynamic early childhood environments. Yet for rapidly developing preliterate young children, solutions are needed now. We present MindScribe, an interactive robotic object that leverages open-ended 'serve and return' natural language interactions to intelligently support reflective inquiry and school-readiness in highly variable and imaginative early childhood environments.