Artificial intelligence (AI) has become an extremely influential field, and as this technology has advanced at the speed of light, the closely interlinked semiconductor industry is thirsting for talent, driving various industries into stages of intelligent transformation and development, and thereby AI has created urgent needs for talent that understands AI. Human resource experts predict that up to 70% of future job openings will be related to AI, gradually blurring the skill boundaries between humanities and science majors.

The 2025 Semiconductor Industry Talent Report, jointly released by the Industrial Technology Research Institute (ITRI) and 104 Job Bank, revealed nuanced changes in job market demand. In response to the semiconductor industry’s globalization strategies, companies now prioritize candidates with foreign language proficiencies, adaptability to overseas workplaces and cultures, as well as breadth and resilience in professional expertise.

In addition, job-ready skills have long been a key metric for recruitment. How to create a learning-friendly environment on campus and gradually build interdisciplinary AI abilities for students outside these related fields has become a major issue in higher education. In line with these trends, the Ministry of Education (MOE) has promoted and facilitated the establishment of the Taiwan Artificial Intelligence College Alliance (TAICA).

Fifty-five Universities Collaborate to Foster Interdisciplinary Talent

From the initial 25 schools, membership has doubled to 55 universities and colleges. TAICA, operating as an alliance, breaks down barriers between universities and universities of science and technology. Using National Taiwan University’s NTU COOL online course platform, it has launched a new model of large-scale cross-institutional course enrollment and mutual credit recognition, helping students enter into or deepen studies in AI while addressing the shortage of faculty in related fields within the R.O.C., Taiwan’s universities, a breakthrough in the R.O.C., Taiwan’s educational history.

With over 20 years of teaching experience in AI, Professor Yi-Shin Chen from the Department of Computer Science at National Tsing Hua University's College of Electrical Engineering and Computer Science, took on the role of project office host without hesitation. Having spent more than seven years promoting AI courses under the guidance of the  Department of Information and Technology Education of MOE, she worked with the MOE to quickly design a comprehensive project execution framework. Pointing to the course map, she explained the project’s original intent and her hopes for sharing AI education resources, noting, “This is an important step to put educational equity into practice.”

Over the past 20 years, the job market in the R.O.C., Taiwan has undergone major changes. In AI teams that emphasize human–machine relations and technology ethics, humanities major students have emerged as standout contributors alongside science major students. “Humanities majors focus on the close ties between people and society; which can motivate team growth, support international engagement, and coordinate management. This echoes the view of many AI experts that interdisciplinary talent is the true core of AI teams.” Current talent demand trends confirm this.

Taking the semiconductor industry as an example, in May this year (2025) of the 34,000 AI-related job openings, alongside traditional R&D and software/hardware engineering roles, many were in business management and project operations. Notably, salary increases for managers from non-STEM backgrounds ranked highest among various managerial positions, rising 14.9% and 19.9% year-on-year. Salaries for non-managerial employees and professionals with interdisciplinary expertise, such as engineers in technical development, green energy, and reliability, rose by 21.1%. This indicates that, under the influence of global competition and geopolitics, companies urgently need “hybrid” talent with diverse areas of expertise in order to effectively enhance critical competitiveness.

Building an AI Education All-Star Team with Curated Teachers and Courses

TAICA began with a nationwide inventory and integration of university AI courses and faculty. Four progressively advanced credit programs were planned: Applied Artificial Intelligence Exploration Program, Artificial Intelligence in Industrial Applications Program, Artificial Intelligence for Natural Language Technology Program, and Artificial Intelligence for Computer Vision and Imaging Technology Program. Concurrently, universities and instructors with abundant teaching resources and widespread acclaim from experts and students, were invited to offer “master courses” via livestream for universities and college students, with some courses registering up to 1,600 students.

“We designed the course progression based on the learning map and selected representative courses from various universities to include in TAICA.” These courses included Professor Yi-Shin Chen’s “Data Mining: Concepts, Techniques, and Applications” and Professor Hung-Yu Kao’s “Natural Language Processing” at National Tsing Hua University, Professor Hsuan-Tien Lin’s “Machine Learning” and Professor Jyh-Shing Roger Jang and Assistant Professor Chun-Ming Chen’s “Introduction to FinTech” at National Taiwan University, and Professor Wei-Ta Chu’s “Introduction to Artificial Intelligence” at National Cheng Kung University. Starting from the second phase, the very important required course “AI Ethics” was gradually added to every program; other courses were added to correspond to the teaching and learning needs of the four credit programs, including “Introducing Generative AI for the Humanities”, “Robotic Navigation and Exploration”, “Generative AI: Text and Image Synthesis Principles and Practice”, and “Deep Learning.”

At the same time, Professor Yi-Shin Chen faced an even greater challenge: “Some instructors were not very familiar with livestreaming teaching and felt somewhat hesitant, worrying that fully remote instruction might weaken the effectiveness of in-person classes.” Through one-on-one communication and in-depth discussions, they explored how to use interactive teaching designs to increase student engagement and, in consideration of the varying student achievement levels across universities, adopted rolling adjustments. Master courses were divided into “mirror courses”, in which the instructor graded and assessed assignments of alliance university students, and “satellite courses”, where assessment authority was delegated to each university. In the latter courses, collaborating instructors worked either according to the assessment methods designed by the course instructor, or customized suitable assessment approaches and standards based on student levels and the characteristics of their universities. This allowed each institution to seek commonality amid diversity, enhance student learning outcomes, and integrate grading standards across different universities.

Sharing Educational Resources, Nurturing Talent Across the R.O.C., Taiwan

TAICA’s master course instructors are all renowned professors in the AI field at their respective universities, with some courses so popular that even their own department students compete to register. In addition, Professor Chen noted that these instructors share a common trait: “a sense of mission to help students recognize that artificial intelligence, which emphasizes human–AI collaboration, is not only about technology that is advancing by leaps and bounds—it is ultimately people who are the most important factor.”

Facing the challenge of outstanding instructors, TAICA has expanded the channels for AI learning. Professor Chen further explained, “It’s undeniable that relatively abundant teaching resources are one reason students compete to enter the top universities, so our core principle is resource sharing, in order to create fair learning opportunities. Students are no longer excluded from AI education due to weak fundamentals, low past achievement levels, or limited school resources.” Any student who wants to learn can enter top university classrooms and, combined with remedial instruction designed by each university, improve their learning outcomes. “By following the progress and earning the program’s credit certificate, anyone can show that their abilities are on par with top university graduates, breaking the traditional mindset of classifying students by exam scores, and thereby realizing true educational equity.”

In the early days of the program, some courses saw high withdrawal rates and unsatisfactory student assessment outcomes, sparking external doubts. “We found that one reason was that students were registering for courses without fully understanding them and then withdrawing due to mismatched expectations. Another key factor was that some schools couldn’t provide sufficient learning support.” Observing high-pass-rate schools revealed that highly engaged assistant instructors and teaching assistants could more easily identify students’ learning difficulties and bottlenecks in time and then provided assistance, which directly correlated with better learning outcomes.

In addition to teaching “Data Mining: Concepts, Techniques, and Applications”, a master course conducted entirely in English, Professor Chen also serves as an assistant instructor for other courses. She often encourages students with her own saying: “Grades are only a momentary evaluation; the knowledge you accumulate is what lasts a lifetime.”

Learning AI cannot be crammed at the last minute, as when an application gets stuck, you can’t move forward. “Everyone has gone through this,” Professor Chen wants students to understand that courses that are not challenging do not help you learn. Only by actually attending a course can you find the gaps in your learning.

“Completing assignments and running models every week, you’ll inevitably get stuck and spend a lot of time without being able to find the problem. When that happens, there’s no need to obsess over it, go do something else, rest and return later to do it again. After repeating this a few times, inspiration can often strike.” She believes that some students limit themselves by overthinking the course or assignments, focusing on how far they are from the final step and creating unnecessary “difficulties” for themselves. But learning has no shortcut, like overcoming challenges in a game step by step, only gradual progress can lead to improvement.

Increasing Student Course Participation to Cultivate Future Talent

“Artificial intelligence is a very special field, where theoretical research and practical application are equally important; almost all new technical advances are aimed at solving practical problems.” Therefore, Professor Chen particularly praises the approach of some universities of science and technology and has even invited their faculty-student teams to share their experiences. These instructors take the time to observe their students and, after accompanying them through the master course livestreams, provide remedial-style support to strengthen fundamentals. They are also willing to further explain and clarify difficult ideas to fill gaps. “Students at universities of science and technology generally have stronger resilience. If they are given the opportunity to learn and be recognized, they'll achieve impressive results.”

Currently, most TAICA courses are offered at the graduate level. Professor Chen stated that this approach aims to expand its influence and allow more students to enjoy the use of these educational resources. “Most of these courses are suitable for juniors and seniors, but offering them at the undergraduate level wouldn’t count toward graduate-level credit requirements, so offering them as graduate-level courses open to undergraduates can increase students’ willingness to enroll in these courses.”

In the future, Professor Chen hopes to find suitable instructors and incorporate more specialized areas of AI applications into the TAICA teaching platform. “We hope that students in the R.O.C., Taiwan can learn the latest AI technical developments and applications without going abroad or searching for foreign resources. This not only promotes educational equity but also cultivates more related talent and future instructors right here in the R.O.C., Taiwan, creating a positive cycle.”