AI tutoring as structured one-to-one voice interaction rather than a chatbot changes the educational definition of the category

AI tutoring being spoken one-to-one math tutoring with a curriculum-aligned program where human teachers maintain control rather than a chatbot or instant answer generator is the definition that distinguishes genuine educational AI from AI study assistance.

Third Space Learning's AI tutor Skye being shown as the reference implementation covers voice interaction, real-time feedback during problem solving, curriculum alignment and human teacher oversight as the combination that makes it educationally effective rather than just technologically interesting.

The distinction between a student getting instant answers to homework problems and a student engaging in structured tutoring dialogue that guides them toward understanding is the pedagogical difference that determines educational outcomes. The former builds answer-retrieval skills. The latter builds mathematical reasoning.

The human teachers maintaining control being emphasised as a feature rather than a limitation tells you something important about the design philosophy: AI as an educational force multiplier for human teachers rather than a replacement.

For educators evaluating AI tutoring tools: what is the specific learning outcome evidence you would need to see before recommending an AI tutoring tool to students or parents?