News Center



Home News Center News

Practice and Convergence of AI Education Products

Jan 13-2025   



This event is co-organised by the AI Collision Bureau of the Digital China Alliance of the National School of Development at PKU, Gongchainhui and the SAIF AI Club of the Shanghai Jiaotong University.

Gerald Xu, an undergraduate from Fudan University, a master of Electrical and Computer Engineering (ECE) from Cornell University, and a full-stack engineer in Silicon Valley as well as a serial entrepreneur.

 

Gerald Xu's sharing is thought-provoking: Can education be merged with everyday life? If it can, then the purpose, means, and meaning of education will become one. However, education products are still mainly tools for learning skills, and language and subject education products still occupy the majority of the market. The main purpose of parents using these products is to improve their children's academic performance.

 

 

On the other hand, the product of Duolingo has better solved the problem of making teaching fun. Then, can we build a virtual game so that people can complete learning while experiencing the thrill of playing games?

 

It is widely believed that the biggest difference between AI Native education and the past should be a shift in the human state, from being driven by others to being self-driven. In the era of AI, the value of knowledge will become less and less critical, while the value of individual contribution and overall creative ability will become more and more important. Teachers or AI will turn into pathfinders, cultivating the ability to synthesize knowledge on a case-by-case basis.

 

Zhiyong Li, an alumnus of the EMBA Class 2021 of PKU NSD, the executive director of the AI Business Academy at Zhongguancun Zhiyong Artificial Intelligence Research Institute, and the founder of Zhineng Tiaoyue company.

 

Mr. Li mentions that the reality is that parents don't have that much time to spend with their children throughout the learning process. Even if they do have the time, they may not have the skills. So, one can see the need for companion products. But what will happen next? There will be AI-driven mums and AI-driven dads, and they will become more and more realistic and capable of solving problems. And then we'll suddenly find a paradox: Won't this really lead to an emotional detachment between parents and children?

 

If AI allows us to spend less time at work and more time at home, then it seems to be on the right track. But if it goes the other way, then it doesn't seem right, even though there is a short-term rationale for it.

 

Discussion Session

The guests then discussed a series of questions around "what can AI do now":

 

If AI has an illusion, what if we teach the wrong thing in education?

How do we control the accuracy of RAG, a technology that seems to alleviate rather than solve the problem? At this stage, the problem is intractable. Practically speaking, it can only be addressed through engineering practices. Meanwhile, we can reduce the impact of such illusions through product definition and positioning and then expect progress in the AI itself in the endeavor.

 

Is education itself more likely to progress offline or online at the moment?

All current technologies seem to point in the direction that the virtual online will eventually replace the offline, and increasingly so. Gerald Xu is neutral on this point, believing that in the long run both will have equal opportunities, depending on the realities of the individual.

Practice and Convergence of AI Education Products

Jan 13-2025   



This event is co-organised by the AI Collision Bureau of the Digital China Alliance of the National School of Development at PKU, Gongchainhui and the SAIF AI Club of the Shanghai Jiaotong University.

Gerald Xu, an undergraduate from Fudan University, a master of Electrical and Computer Engineering (ECE) from Cornell University, and a full-stack engineer in Silicon Valley as well as a serial entrepreneur.

 

Gerald Xu's sharing is thought-provoking: Can education be merged with everyday life? If it can, then the purpose, means, and meaning of education will become one. However, education products are still mainly tools for learning skills, and language and subject education products still occupy the majority of the market. The main purpose of parents using these products is to improve their children's academic performance.

 

 

On the other hand, the product of Duolingo has better solved the problem of making teaching fun. Then, can we build a virtual game so that people can complete learning while experiencing the thrill of playing games?

 

It is widely believed that the biggest difference between AI Native education and the past should be a shift in the human state, from being driven by others to being self-driven. In the era of AI, the value of knowledge will become less and less critical, while the value of individual contribution and overall creative ability will become more and more important. Teachers or AI will turn into pathfinders, cultivating the ability to synthesize knowledge on a case-by-case basis.

 

Zhiyong Li, an alumnus of the EMBA Class 2021 of PKU NSD, the executive director of the AI Business Academy at Zhongguancun Zhiyong Artificial Intelligence Research Institute, and the founder of Zhineng Tiaoyue company.

 

Mr. Li mentions that the reality is that parents don't have that much time to spend with their children throughout the learning process. Even if they do have the time, they may not have the skills. So, one can see the need for companion products. But what will happen next? There will be AI-driven mums and AI-driven dads, and they will become more and more realistic and capable of solving problems. And then we'll suddenly find a paradox: Won't this really lead to an emotional detachment between parents and children?

 

If AI allows us to spend less time at work and more time at home, then it seems to be on the right track. But if it goes the other way, then it doesn't seem right, even though there is a short-term rationale for it.

 

Discussion Session

The guests then discussed a series of questions around "what can AI do now":

 

If AI has an illusion, what if we teach the wrong thing in education?

How do we control the accuracy of RAG, a technology that seems to alleviate rather than solve the problem? At this stage, the problem is intractable. Practically speaking, it can only be addressed through engineering practices. Meanwhile, we can reduce the impact of such illusions through product definition and positioning and then expect progress in the AI itself in the endeavor.

 

Is education itself more likely to progress offline or online at the moment?

All current technologies seem to point in the direction that the virtual online will eventually replace the offline, and increasingly so. Gerald Xu is neutral on this point, believing that in the long run both will have equal opportunities, depending on the realities of the individual.