Artificial Intelligence and Education: Lessons from China

Artificial Intelligence and Education
Facebook
Twitter
LinkedIn
WhatsApp
Email

China is going all-in on AI. From e-commerce to agriculture to law enforcement, few other countries have embraced artificial intelligence with the same speed and enthusiasm. In just a few years, it has turned up everywhere: whether opening a bank account or watching a shopping livestream, one can now easily find oneself dealing with a fairly convincing AI-generated avatar – and these are just the most visible manifestations of the new technology. Just as with mobile tech a decade ago, China appears to be leapfrogging the west, at least in terms of implementation. In the AI development race, Western firms may have access to greater processing power (for now), but China has applied the technology more broadly, gaining valuable insight into how AI functions in the real world. This is especially true in education, as discussed in a recent panel event hosted by the Lau China Institute at King’s College London. 

Education has long been at the heart of Chinese national strategy. As the world’s most populous country until earlier this year, China’s people are perhaps its greatest resource. The better educated they are, the better off the country will become (so long as there are safeguards against graduate overproduction, which can derail economic progress). AI has also been identified as a key area for development; it’s natural that the two strands be combined. 

Adapting education for the era of artificial intelligence

There are, broadly speaking, four ways China is adapting its education system to AI. The first is capital investment. Both government and private industry have poured huge sums into AI development and implementation across all sectors. While there is a lot of chatter right now about microchips, this is only part of the equation. It is one thing to have a phenomenally powerful AI – but how do you use it? Investment consequently focuses not just on chips and infrastructure but also on implementation and delivery. Education can be an area for both deploying AI, and an invaluable tool for training new generations to work with it.

The second is by building a talent pipeline. AI has upended the world, and nations will need to adapt – quickly. Although crude forms of artificial intelligence have been around for decades, the last three years have seen the first really significant tipping point in power and capability, and this will transform the future of work in ways we can barely fathom. As many as 65% of children in school today will work in jobs that do not yet exist – that is, if the AI leaves any jobs for us mere humans. But how can countries cultivate talent for a future none of us can yet predict? The answer is not to train legions of programmers and specialist engineers, who may find their skills made redundant by new technology, but instead to build AI awareness and basic skills into the curriculum. China is taking the lead on this. 

No one will be able to avoid AI in the future, just as few of us can avoid smartphones and the internet today. Soft skills will be particularly important, just as they are today. Take iPhones. It’s safe to assume that, however good we might be at using it, most of us couldn’t take an iPhone apart, explain how it works, or build one ourselves. But we don’t need to: we leave that to the experts, and instead learn how to incorporate the machine into our lives (or, sometimes, allow it to dominate them). 

The same will be true of AI. Tomorrow’s students won’t necessarily need to be able to code and program an AI, but they will need to learn the skills to co-exist with and benefit from it: how to write the most efficient prompts, how to apply it to their studies, how to spot AI-driven scams, and how to navigate the future workplace. (If you thought your annual performance review was stressful now, just wait until Skynet is doing it). Young people will also need to learn how to modify their daily behaviour for an AI world. When artificial intelligence decides your university admissions and your job applications, calculates your credit score, and brokers your mortgage, it’s inevitable that people will have to adapt their lifestyles to the system – a sobering thought indeed. If AI continues on its present development trajectory, then by mid-century dealing with it will be like talking to God. The sooner students learn to live with AI, the better they will fare in the future.  This all starts in the classroom. 

AI in the classroom

China is steadily rolling out AI across the education sector, from kindergarten to university, both as a tool for empowering learning and as a method of monitoring and data-collection, in the form of classroom behaviour-management systems. AI can help students learn, by identifying and adapting to individual strengths and weaknesses, but how AI impacts pedagogy must be carefully managed by teachers. It can easily teach students how to smash exams – ChatGPT could be a qualified lawyer in several US states; and while no one appears to have unleashed its Chinese equivalent Ernie on the gaokao, we could certainly expect similar results – but this may come at the expense of the student’s development. Education, after all, is not just about passing exams, but building a wide range of skills. We’ve all known students in our lives who ace tests but flounder when required to work outside the same narrow, prescribed limits within which they have become comfortable. AI could reinforce this problem: it thrives in rote-learning environments, but struggles at building deep literacy or critical thinking. It is worth remembering that programmes like ChatGPT and Ernie don’t actually know anything; they’re just exceptionally good search engines and passable text generators. Both students and teachers must be aware of the benefits and the limits of AI, and learn how to incorporate it into their work without becoming over-reliant on it. 

AI, private tutoring and language reform

Finally, AI is being used to manage a general shift in the Chinese education system. Many education-focused artificial intelligence companies cut their teeth in China’s enormous private tuition industry. However, the central government has taken an increasingly dim view of the sector, having become concerned that it was exacerbating social inequality and fuelling an educational arms race that was damaging young people’s health. After crackdowns beginning in 2020 – most private tutors were shut down overnight – many education-focused AI firms have had to move to new areas. One identified by panellists at the Lau Institute conference was Mandarin language education, both for foreigners and among China’s minorities. Since the First Emperor standardised Chinese script two thousand years ago, many have sought to unify China linguistically. AI may be the tool that finally makes it possible.

Conclusions

Overall, China is steadily building AI capacity. In industry, this means developing chip foundries, building infrastructure and training engineers. In education, this means rolling AI out in the classroom, both to help teachers, engage students, and also keep an eye on their behaviour. It has begun to implement AI across society faster and more broadly than in the West, and is gaining useful insights into how AI can be deployed in the real world. But, as with all things AI, we must be careful not to fall for the hype – a few breathless news stories are not always an accurate reflection of the situation on the ground. Last year it was the metaverse; this year it is AI. Perhaps next year it will be something else entirely. But whatever the reality, it is increasingly clear that AI will transform the Chinese education system. Exactly how remains to be seen. 

Sean Paterson

Sean Paterson