TeachMaster: Generative Teaching via Code

2025-12-07

AI Summary

This work addresses the high cost and lengthy production cycles of high-quality online educational content, as well as the limitations of existing video generation methods in simultaneously preserving pedagogical structure and enabling precise control. The authors propose a generative instructional paradigm that introduces a novel multi-agent collaborative framework using code as an intermediate semantic medium. This approach redefines the instructor’s role as a designer of pedagogical intent, decoupling intent specification from execution. Through a hierarchical planning and rendering coordination mechanism, the method automatically generates course videos that are structurally coherent, interpretable, and editable. Experimental results demonstrate a substantial increase in production efficiency, reducing costs to 0.3% of those associated with traditional online courses while maintaining strong structural coherence and visual fidelity.

Abstract

The scalability of high-quality online education is hindered by the high costs and slow cycles of manual content creation. Despite advancements in video generation, current approaches often fail to ensure pedagogical structure and precise control due to their pixel-level, black-box nature. In this paper, we propose Generative Teaching, a novel paradigm shifting educators from manual creators to high-level directors who focus on pedagogical intents while agents handle the execution. To realize this vision, we introduce TeachMaster, a multi-agent framework that leverages code as an intermediate semantic medium. Unlike traditional video generation methods, TeachMaster orchestrates a collaborative team of agents, spanning planning, design, and rendering, to automate the production of interpretable, editable, and curriculum-ready educational videos. Experiments validate that TeachMaster significantly boosts production efficiency without compromising structural coherence or visual fidelity, slashing production costs to only 0.3% of traditional online course videos and providing a robust solution for scalable education.

Problem

• online education
• content creation
• video generation
• pedagogical structure
• scalability

Innovation

• Generative Teaching
• multi-agent framework
• code as semantic medium
• automated educational video generation
• scalable online education