Although Hanhan hasn’t learned about programming, she has started to comprehend mechanics. There is a large amount of knowledge involved in the construction of LEGO. Jiang pointed out that the teacher didn’t teach the theory directly, but Hanhan has learned about basic physics while playing in the class, such as gears, lever principle and so on.

“The main reason for enrolling Hanhan in LEGO class is to develop her hands-on skills and spatial thinking. Besides, attending a training class that she has an interest in will provide her more opportunities to get along with others and broaden her horizons,” said Jiang.

Xiong Bingqi, Deputy Director of 21st Century Education Research Institute, told People’s Daily Online that parents must be rational, consider their child's circumstances and should not blindly compare with others. The main intention should be to develop a child's interests, rather than to pursue utilitarian goals.

Charging standard and optimized content are required

Training courses in programming which are highly sought after are a bit pricy for many families. According to a China News Service report in March, the price of children's programming training online is around 300 yuan (about 45 US dollars) per hour. If enrolling for a whole year, that can add up to over 17,000 yuan (about $2,528). If enrolling in an offline course, the price will increase.

The report also noted that the content of coding classes between different training institutions is similar, mostly relying on software from the MIT, such as scratch, and the teaching methods always include game enlightenment and graphics programming.

The lack of a unified charging standard and uneven teaching quality are also a concern to many parents.

According to a China Daily report in November last year, China's first AI textbook series designed for primary and secondary school students is expected to debut in 2019. The series covers the educational demands of students from primary school to high school on AI and provides a systematic framework of AI knowledge.

Currently, there are still some difficulties in the development of the AI education industry including lack of a place in the school curriculum, adequately trained teachers and teaching resources, student learning evaluation and other supporting teaching resources, monotonous teaching environment and teaching methods, to name a few. To this end, some experts have spent nearly five months launching a textbook series, to meet the teaching needs from primary through high school.

“The education sector is now carefully reflecting on whether learning coding should start with the children,” Xiong Bingqi noted.

According to Xiong’s suggestions, on the one hand, it is vital to carry out scientific argumentation on teaching content, and to achieve individualized education and to teach students by aptitude and talent, so that children can grow up and develop better. On the other hand, it is necessary to promote reform of the education evaluation system, to reverse the unscientific evaluation of education, such as assessing only by scores and school entrance opportunities.

“On this basis, new education courses like AI can enter the schools to nurture children better,” he said.