In the rapidly evolving landscape of education, a groundbreaking study spearheaded by Gao Qiong and Lu Jijian offers a compelling blueprint for the future of classroom instruction. Published in the distinguished journal Modern Distance Education, this research meticulously constructs and validates a human-machine collaborative teaching model specifically designed for junior high school mathematics. Far from being a theoretical exercise, this model is grounded in real-world classroom trials, demonstrating tangible improvements in student engagement, personalized learning, and overall academic performance. As artificial intelligence (AI) ceases to be a futuristic concept and becomes an integral part of our daily tools, this study provides educators and policymakers with a practical, actionable framework to harness its power, not to replace teachers, but to profoundly augment their capabilities and redefine the learning experience for a new generation.
The core thesis of this research is both simple and revolutionary: the most effective classrooms of the future will not be those dominated by machines, nor those that stubbornly resist technological integration. Instead, they will be dynamic ecosystems where the irreplaceable human wisdom of the teacher synergizes with the formidable computational and analytical power of AI. This “human-machine collaboration” is presented not as a compromise, but as the optimal path forward, addressing long-standing, intractable problems in traditional education. For decades, teachers have struggled with the near-impossible task of understanding the unique learning needs of every student in a crowded classroom, providing timely feedback on assignments, and differentiating instruction to cater to varying skill levels—all while managing administrative duties. The Gao and Lu model directly confronts these challenges, proposing a structured, three-phase approach—intelligent pre-class diagnosis, AI-assisted in-class learning, and precise post-class evaluation—that transforms these burdens into opportunities for deeper, more meaningful education.
The foundation of this model rests on four critical pillars: Teacher Wisdom, Machine Intelligence, Collaborative Teaching, and Student Wisdom. Each is indispensable, and their interplay is what creates the model’s efficacy. Teacher Wisdom, in this context, transcends traditional pedagogical skills. It is redefined for the AI era as the educator’s ability to interpret complex data provided by machines, to make nuanced, context-sensitive decisions that algorithms cannot, and to foster the social, emotional, and ethical development of students. The teacher is no longer merely a lecturer but a strategic facilitator, a mentor, and a conductor of a complex symphony of human and machine interactions. Their role shifts from delivering standardized content to curating personalized learning journeys, guiding collaborative inquiry, and nurturing critical thinking and creativity—skills that remain uniquely human and are increasingly vital in the 21st century.
Machine Intelligence, the second pillar, serves as the tireless, data-crunching engine that powers the entire system. It is not envisioned as a sentient being but as a sophisticated set of tools—intelligent platforms, data analytics engines, and adaptive learning systems. Its primary functions are to automate the mechanical and repetitive aspects of teaching: grading assignments, identifying common misconceptions across a class, tracking individual student progress in real-time, and recommending tailored learning resources. For instance, an AI system can analyze a set of pre-class diagnostic exercises and instantly generate a report showing that 60% of the class struggles with the concept of absolute value, while a smaller group needs reinforcement on number lines. This level of granular, immediate insight was previously unattainable for a single teacher managing dozens of students, making the machine an indispensable partner in instructional decision-making.
The third pillar, Collaborative Teaching, is the operational heart of the model. It is the structured process by which Teacher Wisdom and Machine Intelligence interact and complement each other. This collaboration is not ad hoc; it is carefully choreographed into the three phases of the teaching cycle. In the pre-class phase, the machine diagnoses student readiness, and the teacher uses that diagnosis to plan targeted lessons. During class, the machine might push adaptive practice problems to student groups while the teacher circulates, providing high-touch, personalized guidance to those who need it most. After class, the machine compiles performance data, and the teacher uses it to refine future instruction and assign personalized homework. This continuous, data-driven feedback loop ensures that teaching is never static but is constantly evolving to meet the learners’ needs.
The ultimate goal, the fourth pillar, is Student Wisdom. The entire model is engineered to cultivate not just knowledge acquisition, but the development of higher-order thinking skills, self-directed learning capabilities, and the ability to collaborate effectively. By leveraging AI for personalized pathways, students are no longer passive recipients of a one-size-fits-all curriculum. They become active participants in their own education, working in collaborative groups (or “learning communities” as the study terms them) to solve problems, present findings, and learn from one another. The AI supports this by providing just-in-time resources and challenges, while the teacher fosters the environment and provides the human connection that motivates and inspires.
The practical implementation of this model, as detailed in the study, is both inspiring and instructive. The researchers conducted a controlled experiment in a junior high school in Hangzhou, comparing a class taught using their human-machine collaborative model against a control group taught with traditional methods over a two-month period. The technology stack included Hikvision’s intelligent cameras for behavioral and emotional analysis and iFLYTEK’s “Zhixuewang” platform for academic diagnostics and resource management. The results were clear: the experimental class showed higher levels of student satisfaction, greater perceived teacher attentiveness, and, most importantly, superior academic gains on their final examinations compared to the control group. This empirical validation is crucial, moving the discussion from theoretical possibility to proven practice.
A deep dive into the classroom reveals how transformative this model can be. In the pre-class “intelligent diagnosis” phase, instead of a teacher spending hours manually grading a stack of worksheets, the AI platform automatically analyzes student responses to a set of diagnostic questions. It doesn’t just score them; it identifies patterns. It might flag that while the class as a whole understands basic addition of rational numbers, there is a significant subgroup that consistently confuses the rules for multiplication. This insight allows the teacher, Gao Qiong in the case study, to restructure her lesson plan before class even begins, dedicating more time to the problematic concept and preparing targeted materials for the struggling students.
The in-class phase is where the magic of collaboration truly unfolds, broken down into three sub-phases: personalized guidance, intelligent assistance, and collaborative research. During personalized guidance, students might be working on individualized learning paths on their tablets. One student, identified as needing reinforcement, is reviewing foundational concepts with interactive tutorials, while another, who has mastered the basics, is tackling advanced problem sets. The teacher, Lu Jijian, is not at the front of the room lecturing; he is moving between groups, answering deep, conceptual questions that the AI cannot handle, and facilitating discussions. The AI, meanwhile, is monitoring progress, pushing new challenges to students who complete tasks quickly, and alerting the teacher if a student is stuck for an extended period.
The “collaborative research” sub-phase is particularly innovative. After working on problems, student groups are tasked with presenting their solutions to the class. But here’s the AI twist: the problems each group presents are not assigned arbitrarily by the teacher. The AI analyzes the questions each group struggled with during the “assistance” phase and generates a unique presentation task tailored to their specific learning gap. One group might present on common errors in applying the distributive property, while another focuses on visualizing negative numbers on a number line. This ensures that the presentations are deeply relevant and that the entire class benefits from a diverse exploration of the topic’s nuances. Furthermore, intelligent cameras can track student engagement during these presentations, providing the teacher with real-time data. If the data shows that the fourth group’s presentation is causing a dip in class-wide attention, the teacher can immediately intervene, perhaps by posing a provocative question or shifting to a different activity, thereby maintaining a dynamic and responsive learning environment.
The post-class “precise evaluation” phase closes the loop. Instead of assigning the same homework to every student, the AI generates a personalized set of practice problems for each individual, targeting their specific weaknesses identified during the in-class activities. The teacher then reviews and refines this AI-generated assignment, adding a human touch—for example, including a reflective writing prompt that asks students to explain their thought process, something an algorithm cannot assess. This ensures that assessment is not merely summative but formative, directly feeding back into the next cycle of learning.
The implications of this study extend far beyond the mathematics classroom. It provides a template for how AI can be integrated into any subject area. The principles of using AI for diagnosis, personalization, and automation of routine tasks, while reserving high-level cognitive and emotional labor for the human teacher, are universally applicable. For school administrators, it offers a roadmap for investing in educational technology—not as a replacement for teachers, but as a powerful tool to empower them. For educational technology companies, it provides clear design principles: build tools that seamlessly integrate into the teacher’s workflow, provide actionable insights, and enhance, rather than dictate, pedagogical choices.
However, the researchers are also clear-eyed about the challenges and ethical considerations. A significant recommendation from the study is the urgent need to enhance teachers’ “intelligent teaching literacy.” Educators must be trained not just to use new software, but to interpret complex data dashboards, understand the limitations of AI, and make informed decisions about when to trust the machine and when to override it. This requires a fundamental shift in teacher professional development programs. Furthermore, the study raises critical concerns about data privacy and algorithmic bias. Schools must implement robust data governance policies to protect student information. There is also the risk that over-reliance on AI could lead to a dehumanization of education. The study strongly cautions against this, emphasizing that the “soft intelligence” of human empathy, ethical judgment, and inspirational leadership must always remain at the core of the educational experience. AI should handle the “hard intelligence” of data processing, freeing the teacher to focus on what they do best: nurturing the whole child.
Looking ahead, the research team, which includes Wang Xiaojing, Shang Jiahui, and Zhou Yueliang, envisions a future where this model evolves. They predict a transition from “weak” human-machine collaboration, where AI acts primarily as a tool, to “strong” collaboration, where AI becomes a more proactive partner in the educational process, perhaps even taking on the role of a co-teacher for certain structured activities. This future, however, is not presented as a utopian inevitability but as a goal that requires careful, ethical, and human-centered development.
In conclusion, the work of Gao Qiong, Lu Jijian, and their colleagues is a seminal contribution to the field of educational technology. It moves beyond the hype and fear surrounding AI in education to deliver a pragmatic, evidence-based, and profoundly humanistic model for the future of teaching and learning. By embracing a philosophy of collaboration rather than competition between humans and machines, this model promises to create classrooms that are more equitable, more engaging, and more effective. It empowers teachers to be the inspirational guides they aspire to be and equips students with the personalized support they need to thrive in an increasingly complex world. This is not the end of the teacher; it is the beginning of a new, more powerful chapter in the noble profession of education.
Gao Qiong, Lu Jijian, Wang Xiaojing, Shang Jiahui, Zhou Yueliang. “The Construction and Practical Cases of Human-Machine Collaboration Teaching Mode in the Era of Artificial Intelligence.” Modern Distance Education, 2021, (4): 24-33. DOI: 10.13811/j.cnki.eer.2021.04.003