The article presents a framework for managing AI agents in workplace settings by comparing them to highly intelligent but socially unpredictable young children. The comparison suggests that AI systems possess advanced capabilities while lacking human judgment and ethical constraints, requiring specific management approaches from their human counterparts.