In the first of a two-part series, Rosemary Chen looks at the state of readiness of Taiwan's educations sector for the sweeping changes that will engulf the workforce as governments and industry rush to embrace various forms of automation.

“I’m not worried about being replaced by robots because I am an emotional being and robots cannot replace the connection between humans,” says Lin Hung-yu (林泓宇), a junior student studying business at National Taiwan University (NTU).

But when asked about recent Artificial Intelligence (AI) developments, Lin is more circumspect. “It’s not really my field,” he says.

Lin’s confidence in the future is shared by many college students in Taiwan.

“I’m not afraid because creativity cannot be replaced,” says Hu Chun-lun (胡均綸) a former programming intern at AppWorks, Asia’s biggest start-up accelerator. He explains that the ability to think critically and create things are unique human qualities.

“I proactively think about how to improve things and solve problems — which is how I differentiate from robots,” Hu adds.


Credit: REUTERS/Phil Noble

Frontline service staff, such as those displaced by this robotic information panel at the Pyeongchang 2018 Winter Olympics, are among the sectors most under threat from automation in developed economies.

Kevin Tseng (曾慶元) a third-year mechanical engineering student at NTU is also unafraid of automation because robots are just tools. “After all, robots are created by humans and cannot replace jobs that require strategic thinking,” Tseng suggests.

Yet despite this confidence, students in Taiwan and around the world are about to enter a shifting and uncertain future that will demand unprecedented flexibility from employees.

A recent report by the global consultancy firm McKinsey & Co. warns that, “by 2030, 75 million to 375 million workers (3 to 14 percent of the global workforce) will need to switch occupational categories” as a result of various forms of automation, including AI and robotics, adding that right now, "60 percent of occupations have at least 30 percent of constituent work activities that could be automated."

Published in December 2017 and titled, “Jobs Lost, Jobs Gained: Workforce Transitions in a Time of Automation,” the report suggests that “all workers will need to adapt as their occupations evolve alongside increasingly capable machines" through higher educational attainment, as well as focusing more on roles that require "social and emotional skills, creativity and high-level cognitive capabilities and other skills relatively hard to automate.”

The axe is expected to fall more heavily in developed economies, and while Taiwan is not featured in the McKinsey report, the similar composition of its economy to regional peers like Japan, Singapore and South Korea suggest it will that will also likely see about 25 percent of roles automated between 2016 and 2030.

More alarmingly still, "The Global Future of Work Survey", recently published by global advisory firm Willis Towers Watson, suggests that almost half of employers in the Asia-Pacific region expect to need fewer employees by 2020, while respondents reported that "12 percent of work is currently being done using AI and robotics versus just 7 percent three years ago and they anticipate that this figure will rise to 22 percent in the next three years."

So what are Taiwan’s educational institutions doing to prepare students and the existing workforce for an increasingly automated future, and is the government acting fast enough?

Ministry of Education spearheads 'vanguard'

“Technology is changing so fast these days, we must constantly adapt our plans to keep up,” says Lan Man-chi (藍曼琪), Chief of the Technology Education Division at Taiwan's Ministry of Education.

Lan is director of the Artificial Intelligence Talent Cultivation Program (人工智慧人才培育計畫), a NT$292 million (US$9.7 million) four-year plan scheduled to start next year.

The program has already been green-lit by the Executive Yuan, the executive branch of Taiwan’s government, and awaits review by the Legislative Yuan. Once approved, the program will focus on increasing industry-academia collaboration and creating Problem-Based Learning education platforms, where students learn by solving open-ended questions, to schematize education resources in the field of AI.

An “AI course map” will be created, integrating various open-source online courses from universities in Taiwan and abroad. There are also plans to host student competitions in collaboration with businesses and create internship opportunities.

Stimulating this kind of direct collaboration between industry and academia is a reoccurring issue in Taiwan. Academia is often heedless of the profit motive; creating a mismatch between graduates and employers, explains Professor Kang Shih-Chung (康仕仲), the Deputy Vice President of Academic Affairs at NTU.

The aim of the ministry’s program is to work with manufacturers, firms and research institutes. “Anyone with enough resources to provide data,” says the education ministry's Lan. But she anticipates challenges ahead: “Industries in Taiwan are afraid of having their data made public, so it’ll take a bit more effort to have them share it with us.”

“Most government programs [in] this AI push are underfunded and don’t include enough cooperation with real enterprises,” adds Martin Hiesboeck of Geber Consulting, who also writes regularly about AI, machine learning, and automation for his Futurist blog.

The education ministry’s plan is designed to be a “vanguard” – to lay foundations and build networks with businesses and local governments, explains Lan. While the Ministry of Science and Technology focuses on research and development, the Ministry of Education looks to build foundations that underpin the education environment in schools.

“We will test the water first and if we find problems then we can quickly adjust – the goal is to figure out a model that works. So, in the future, schools can expand on the foundation we’ve built,” she says.

The planning team currently consists of professors from top-performing universities, such as NTU, National Chengchi University and National Tsing Hua University, among others. The 10-strong team will expand when full-time administration assistants join later this year.

NGO founds AI Academy to retrain workforce

Outside the higher education system, other groups also see an urgent need to arm people with knowledge and skills in working with AI and machine learning.

The Taiwan AI Academy (台灣人工智慧學校) established by the Taiwan Data Science Foundation (台灣資料科學協會), launched its first three-month AI training programs on Jan. 27.


Credit: Taiwan AI Academy (台灣人工智慧學校)

Taiwan's AI Academy opens for business in January, 2018.

The “Technology Leaders (技術領袖)” course is an intensive Monday to Friday program to train technical workers, while the “Managers (經理人)” program offers a Saturday class for people in leadership positions. Both programs cover statistics, programming languages, machine learning, deep learning, computer vision, and natural language.

Funded by five Taiwanese corporations, the academy has admitted 210 students to its technology leaders program and 330 to its managerial counterpart.

“Our goal is to lift economic growth. We want people with professional skills or domain know-how to learn AI and bring it back to their industry to increase competitiveness,” says CEO Chen Sheng-Wei (陳昇瑋) who is also a Research Fellow of the Institute of Information Science at the Academia Sinica.

During the three-month training, students will solve real-world problems provided by local business. “The businesses have approached us because they can’t find AI talent in the current market,” says Chen.

“AI is nothing new, but for it to have its full effect, it must solve important problems, and if it doesn’t solve any problems – its only as good as a calculator,” adds professor Kang, who also teaches robotics and automation in the Department of Civil Engineering.

The long-term goal of the academy is to break the barrier in the shortage of AI talent in Taiwan so the country can become less reliant on foreign technology. As more AI-trained talent emerges, Chen expects more local service and solution providers to appear; a halcyon scenario that will see engineers turn into entrepreneurs themselves.

The foundation plans to open up parallel branches in Hsinchu by the middle of the year, a Taichung branch by the end of the year and a southern branch next year, either in Tainan or Kaohsiung, with each admitting a total of 2,000 students each year.

“The good thing about being an NGO is that we are flexible. We’ve jumped to take initiative because we know when the government finally steps up it will be too late,” says Chen.

The preparation race

In Singapore, similar retraining programs are instead government-led. Under its national program, AI Singapore, the city state's AI Apprenticeship Program is scheduled to start in April this year. The nine-month full-time training program targets young professionals within the first three years of graduation and expects basic computer programming abilities.

The program intends to train 200 professionals over the next three years and will focus on cultivating candidates’ employability as AI engineers, developers, and consultants.

Singapore also has SkillsFuture, a program under the city state's Future Economy Council that specifically aims to help Singaporeans re-skill as the working environment evolves around them. The organization, which claims to have benefited 285,000 Singaporeans, recently launched its Education [Training and Adult Education (TAE)] Industry Transformation Map. Taiwan has yet to initiate a similarly workforce-wide retraining effort.

Moreover, while such plans and retraining for already well-educated young professionals appear promising, it remains an open question whether those at the sharp end of the so-called automation cliff – the workers and laborers first in the firing line when businesses upgrade their infrastructures – will be able to take part.

Tomorrow, part 2 will look at Taiwan's efforts to arm its kids for a future that expects coding skills as standard.

Listen Next: PODCAST: Can the Universal Basic Income Save Asia from the Automation Cliff?

Editor: David Green