“The ‘iPhone Moment’ of AI Education, Is It Now?”
“Teaching students in accordance with their aptitude” is the dream of education. In the past, it was already very difficult to popularize education, and individualized teaching was even more of a luxury. In the past decade, with the popularity of the Internet, especially mobile Internet, great progress has been made in education popularization. And in today’s AI era, we see the dawn of individualized teaching.
Gaotu Group, founded in 2014, is a pioneer in the wave of education popularization. With its original “online, live-streaming, dual-teacher, large-class” model, it serves 90 million students and brings high-quality education to more people. Entering 2024, Gaotu has new thoughts on AI + education.
Luo Bin, co-founder of Gaotu Education Technology Group, shared his views on AI education at the Geek Park IF2025 Innovation Conference. In the past, although education has changed, the structure of real teachers teaching and students internalizing has not changed. Now in the era of ChatGPT large models, these two variables are undergoing new changes.
The essence of education is to bring changes to students, which may be changes at the cognitive, skill, test result, or personality level. Traditionally, it is difficult to help different people achieve their change goals. For example, in the online course model, students have a low completion rate. Gaotu has innovated the “online, live-streaming, dual-teacher, large-class” model. Through online means, it expands the coverage of high-quality teachers. Live streaming enhances the sense of immersion and interaction. Dual teachers help students develop learning habits. Large classes exert a scale effect.
In the era of large AI models, people are concerned about the changes brought about by the combination of AI and education. Many education enterprises cooperate with AI. Industry practitioners are worried that business models and product formats will be disrupted. Luo Bin believes that the key to the intelligent education reform in the next era lies in personalization, navigation, integration, and specialization.
In terms of personalization, intelligent education in the next era will shift from teacher- or school-centered to student-centered. Because students have great differences in learning ability, state, and rhythm. In the past, the focus of education was on popularization. In the next stage, we need to solve the problem of enabling everyone to have good learning results, including learning efficiency, experience, habits, and motivation. The ideal teaching state is to have an omniscient, omnipotent, and wholehearted excellent tutor who provides personalized teaching according to the real-time state of students and completes the PDCA cycle from curriculum development to testing. And in the future, CDP that collects data such as students’ mastery of knowledge points is expected to provide truly personalized teaching services.
In terms of navigation, taking the IELTS test as an example, by modeling the test, understanding the importance, scoring points, weights, and difficulty of knowledge points, and transforming it into a planning problem, students can spend their limited time on knowledge points with high scores and low difficulty to achieve high-certainty stable delivery.
In terms of integration, it is very difficult to move from state A to state B only relying on teachers and students. Taking medical treatment as an analogy to teaching, in addition to teachers, the supervision system and auxiliary system are also very important. The supervision system helps students manage goals, plan diagnoses, etc. The auxiliary system makes the process more efficient through personalized work and gamified design. Gaotu has proposed a famous teacher system force solution that integrates multiple scenarios to make each link more efficient. AI can play an important role here.
In terms of specialization, excellent teachers are scarce, and it is difficult to solve the problem of supplying high-quality teachers. Teaching services can be provided through the combination of AI and real people. Taking map navigation as an example, different AI invocation methods have different effects. AI Agent is a digital workforce that completes specific tasks end-to-end. In the education scenario, AI can also be used to solve different problems at different nodes. Different enterprises have different positions in the education ecosystem and different effects.
Gaotu’s mission is to make learning more beautiful. It hopes to use the power of AI not only to help students master knowledge but also to change their cognitive models and brain structures. In the next ten years, help students become better versions of themselves.
