On Sept. 14, Slush Singapore 2018 was held in Singapore. 350 start-ups, 150 investors and more than 2,500 spectators from more than a hundred countries and regions gathered, discussing cutting-edge technologies and industries that will lead to global changes in the future. Yixue Squirrel AI Learning intelligent adaptive education Partner Yifang Liao was invited to attend this event. She introduced the AI+ education industry as a forward-looking field. In her opinion, the change brought by AI to education is a restructuring of the learning relationship. Squirrel AI simulates excellent AI teachers to let every child experience one-to-one personalized teaching. Her excellent speech aroused great interest in Squirrel AI among the experts, scholars and investors present.
Before the speech, Squirrel AI Learning’s Co-founder and CEO Wei Zhou attended Slush Shanghai 2018 on Sept. 7 and 8 on invitation, and gave a speech titled “Squirrel AI Learning Gives Every Child a Super AI Teacher”. He believes that the training of excellent teachers is a very long process, which requires continuous teaching practice and experience accumulation. At the same time, with limited energy and time, even excellent teachers cannot keep an eye on the whole learning process of every student. Based on this, the Squirrel AI intelligent adaptive learning system came into being. First, it analyzes the knowledge and skills of special-grade teachers based on advanced algorithms, forming a database with complete proprietary intellectual property rights. Next, by repeating the teaching process “early diagnosis focusing on cognitive blind areas – teaching intervention to make up for knowledge deficiency – assessment of current knowledge level – plan of the learning path for the next step”, it assists students with efficient learning, and has become popular with many parents and students. So far, the renewal rate of Squirrel AI has reached 80%.
Slush originated in Finland. With its focus on technology entrepreneurship and investment, Slush has grown from a small gathering of just 300 Finnish local tech entrepreneurs to one of the most popular and influential tech startup investment platforms in the world. So far, Slush has been successfully held in Helsinki, Tokyo, Beijing, Singapore and Shanghai, attracting 250 million people’s attention worldwide.
As a global tech conference committed to “helping the next generation of companies aiming to conquer the world to succeed”, Slush Singapore focused on a number of areas, such as transportation and education, which are in urgent need of change. It attracted a number of industry leaders to give passionate speeches, including Hyperloop Transportation Technologies’ Co-founder and Board Chairman Bibop Gresta, Golden Gate Ventures’ Founding Partner Paul Bragiel, Intrepid Ventures’ Co-founder and Chief Development Officer Zachary Piester, Entropical Labs’ Co-founder and CEO Tommaso Demarie, and Squirrel AI intelligent adaptive education Partner Yifang Liao.
At the conference, Squirrel AI intelligent adaptive education Partner Yifang Liao gave a speech titled “Squirrel AI Learning Will Surpass Human Teachers and Redefine Education” and gave the AI+ education industry a forward-looking interpretation.
In Yifang Liao’s opinion, for thousands of years, the ultimate goal of education has been “to put people first and teach students according to their aptitudes”; however, the actual condition is that school education is a kind of assembly line operation. In the K12 stage, a student is in a certain class of a certain grade; in college, a student is in a certain major of a certain school. The “talents” produced in this education environment are naturally standardized. It’s against the goal of cultivating individualized and innovative talents.
According to Liao, under the current education system, students not only fail to achieve the goal of efficient learning, but may be bored because of their unclear grasp of knowledge points. “Some students may need only a few minutes to master a knowledge point, while others a class period. The same student may be quick at learning some knowledge points, but slow at mastering others. However, it is impossible for teachers in the same class to slow down or speed up teaching for some specific students. It’s common for “fast-learners to be underfed and slow-learners to be overfed,” she said.
In the 1990s, the concept of intelligent adaptive education began to emerge in the United States. Back then, students were just simply classified into three categories: good, medium and poor, according to their learning levels, so as to match students in each category with the most suitable learning content and path. This education concept achieves rapid development with the support of AI technology. Through AI algorithms, the system acquires and analyzes the learners’ data and gives feedbacks to the existing knowledge maps, provides lessons and exercises at personalized difficulty levels and pace to the learners, thereby improving the learners’ learning efficiency and learning effects.
Although each country or region has its own history, language and cultural background, education concepts can be shared. The wave of intelligent adaptive education originated in the United States has quickly spread to places like China, India and Japan. For example, Knewton, ALEKS, Byju’s, Dreambox, Squirrel AI and other institutions have launched their own intelligent adaptive education products. Khan Academy, which developed MOOCs in the early days, has also turned to adaptive education, developing targeted contents for different regions and examination outlines, which has accelerated implementing the applications of AI in education.
For example, Squirrel AI system can “teach students in accordance with their aptitudes”. It can accurately identify the weak points in each student’s knowledge mastery and present learning content specifically, so that students can find the learning time suitable to them. Moreover, Squirrel AI can “trace the source”. It can help students to master the most basic knowledge points in the knowledge system and fill in the loopholes in their knowledge. In addition, Squirrel AI can “strategically give up some content”. It can give up some knowledge points that account for a small part and have a low occurrence rate in exams but are difficult to learn, and make students learn knowledge points that have the highest relative absorption and lead to the better scores in future exams. In a way, Squirrel AI makes education easier and learning more efficient. It frees students from the misery of excessive assignments, to help increase their interest and confidence in learning.
Why does Squirrel AI have such confidence? Yifang Liao introduced the superiority of Squirrel AI to the audience. The core of AI intelligent adaptive education lies in breaking down knowledge points. Squirrel AI is remarkable in the following aspects. First, Squirrel AI can break down knowledge points 10 times finer than its rivals do and reach the nano-scaled level. For example, the knowledge about rational numbers is broken down to 20 knowledge points. Second, Squirrel AI initiated the theory of correlational probability of uncorrelated knowledge points in the world. The correlations between knowledge points are also sorted out to form a network of knowledge, so that the intelligent system can test students’ knowledge levels with fewer questions. In real tests, Squirrel AI can test 10,000 knowledge points through 500 questions, reaching 90+% accuracy and improving the overall test efficiency and learning effects in education by 5-10 times. Third, while breaking down knowledge points, Squirrel AI also matches corresponding knowledge point explanation contents and examination questions. For one nanoscale knowledge point, it matches dozens of corresponding test questions at different difficulty levels. Aided by AI, the system matches the learning content with each student individually, solving 90% of the time and efficiency problems.
What’s more, Squirrel AI can also break down students’ learning abilities and learning methods, making them “definable, measurable and teachable”. It not only enables students to master effective knowledge points and get high scores in exams but also utilizes Squirrel AI teachers to cultivate and enhance students’ creativity and imagination, thereby fully stimulating students’ learning potentials.
In addition, Squirrel AI keeps developing the theory of AI-enabled education and putting it into practice. In recent years, academic papers published by Squirrel AI have been recognized by international academic conferences such as AIED, CSEDU and UMAP. Squirrel AI Learning has designed an AI-enabled education lab in New York, established a joint lab for AI in Silicon Valley with Stanford Research Institute (SRI) and increased its investment in AI-enabled education. In the meantime, Squirrel AI Learning has organized four “human vs. machine teaching competitions”, from the first competition in one city Zhengzhou to the fourth competition in 100 cities at different levels. Through competitions between the “machine-taught group” and “human-taught group”, the implementation of Squirrel AI system is accelerated.
As for whether AI teachers will impact the relationship between teachers and students, Yifang Liao also expressed her view. She said, “Instead, AI teachers will make the relationship between teachers and students more harmonious.” Because in the traditional education model, teachers’ tasks are relatively heavy. Besides teaching knowledge, they also need to care about students’ emotions and other conditions. To some extent, it affects the improvement of teachers’ professional levels. With the assistance of AI teachers, daily tasks such as teaching knowledge and checking homework can be handed over to AI teachers. Human teachers can focus more on cultivating students’ morality, interest and other aspects, combining “teaching” with “education” perfectly.
Up to now, Squirrel AI has opened more than 1,600 self-service learning centers in over 300 cities of 20 provinces (municipalities, autonomous regions) in China, providing more than 1 million users with personalized “one-on-one tutoring” services by AI teachers. “Our goal at the current stage is to subvert the traditional education model and make children learn with more confidence,” said Squirrel AI intelligent adaptive education Partner Yifang Liao.
SOURCE Squirrel AI Learning