Chinese AI startups show glimpse of the future of education
Zhou Yi, a thirteen year old from Hangzhou, was afraid of math. She had tried traditional services, but one day a company called Squirrel AI came to her middle school. Instead of a human teacher, an AI algorithm would compile her lessons and by the end of the semester her GPA rose from a 5.0 to a 6.25. Two years later, she scored a 8.5 at her final exam (Hao, 2019). How you may ask?
Shanghai-based EdTech company Squirrel AI opened 2,000 (franchised) learning centers and registered over a million students in five years – equal to 85% of students in Dutch higher education (CBS, 2019). For every course it offers, its engineering team works with a group of master teachers to divide a topic into over 10,000 elements, while a textbook might divide the same subject into 3,000 elements. Every element is then supported by video lectures, lecture notes and practice problems to give the student the best “student-centered” education possible (Bourne, 2019). You can learn more about the technology behind Squirrel AI in this keynote speech from the O’Reilly AI Conference below.
This may sound like distance future, but Squirrel is already getting traction outside of China by collaborating with MIT, Harvard and developing an OpenAI platform to eliminate the need for institutions to develop their own intelligent systems. They plan to export their technology to the United States and Europe by the end of 2021 (Kong Ho, 2019). Squirrel AI plan to change universities as we know them by replacing rote tasks, but also the class room. Together with Alo7 they added visual and sound analysis to generate summaries, measure the accuracy of the student’s English pronunciation and assess basic indicators of their effort and elation, such as the number of times they opened their mouth to speak and laugh (Hao, 2019).
To achieve this goal they have raised over $150 million in funding and gained unicorn status, surpassing $1 billion in valuation while expanding rapidly (Peng, 2018). This raises several questions: How long will it take till lectures or teachers will be replaced by algorithms? And what are the (dis)advantages of this development?
References:
Bourne, J. (2019). ‘How Squirrel AI is looking to provide adaptive learning to revolutionize education through AI and big data’. Retrieved 6 October 2019, from https://artificialintelligence-news.com/2019/04/30/how-squirrel-ai-is-looking-to-provide-adaptive-learning-to-revolutionise-education-through-ai-and-big-data/
CBS (2019). ‘Onderwijs in cijfers’. Retrieved 6 October 2019, from https://www.onderwijsincijfers.nl/kengetallen
Hao, K. (2019). ‘China has started a grand experiment in AI education. It could reshape how the world learns’. Retrieved 6 October 2019, from https://www.technologyreview.com/s/614057/china-squirrel-has-started-a-grand-experiment-in-ai-education-it-could-reshape-how-the
Kong Ho, C. (2019). ‘AI education unicorn Squirrel targets foreign markets with plans for mathematics, Mandarin lessons’. Retrieved 6 October 2019, from https://www.scmp.com/tech/start-ups/article/3018297/chinese-education-unicorn-squirrel-ai-targets-foreign-markets-plans
Peng, T. (2018). ‘Adaptive Learning Startup Squirrel AI Raises CN¥1B‘. Retrieved 6 October 2019, from https://medium.com/syncedreview/adaptive-learning-startup-squirrel-ai-raises-cn-1b-df275cbce068
Hello Duncan,
Thank you for your post, it is very interesting. It is crazy to think how much technology is evolving and immersing into our lives; who would have ever thought a machine would teach us better than a human? However, my question is: for which subjects is the machine a better teacher? The data is clear, students had better GPAs with the AI teacher, but do you think this was for every subject? And does it convert all students into replicas of each other? I agree in math or physics for example, an AI teacher could really help you increase, because the answers are objective: either the numbers are right or wrong. However, what about arts, literature, philosophy? For these subjects feelings are essential. You need creativity, you need to express yourself, and correct me if im wrong, but I do not believe machines can develop feelings. Machines are more systematic, which works for quantitative subjects, but I do not believe they will be able to replace universities with human teachers, who have the expressive feelings and morality ethics engrained.
Another point that the video makes is that no student should sit together in the classroom as all students have different capabilities. However, the majority of people meet their friends in class, where they develop their social capabilities and expand their network which is very important in life.