More than a year and a half after ChatGPT’s release, the Ministry of Education has finally announced an initiative to overhaul Taiwan’s education system to accommodate artificial intelligence (AI), and develop AI literacy among teachers and students. However, the concept of “literacy” is misleading. Generative AI, unlike any other previous technology, is a “general” technology that can create any content, and because of this, we need to shift our thinking from “literacy” to “fluency” to truly appreciate its power and its risks.
AI literacy involves understanding the basic concepts of AI, such as how it works, its applications and the potential risks. This foundational knowledge is essential, but it only scratches the surface.
AI fluency, on the other hand, requires a deeper engagement. It means not only knowing about AI, but also effectively incorporating AI tools into workflows to solve problems, generate appropriate responses and critically evaluate outputs.
However, in the context of education, moving from AI literacy to AI fluency would transform students from mere consumers of information into active participants in their learning process.
This is an important distinction. It means students would shift their roles from creators to evaluators, from writers to editors, and from individuals to copilots in their learning tasks.
Instead of researching then writing a report independently, the workflow of an AI user might look like this: They get information with or without AI, give the information to an AI and tell it what to do with the information (such as write an e-mail or a report), and then revise the output repeatedly with or without AI.
This transition from creator to cocreator/evaluator is crucial, as it implies that students are equipped with the skills to generate AI-driven content and, more importantly, to critically assess its quality and appropriateness. However, this is the challenge educators would face: How do we equip students with these skills?
For example, my business student would use the e-mail structure, model and tips I give them to tell ChatGPT to write a sales e-mail to a potential client. The AI quickly generates the e-mail in one-third of the time it would normally take my student, and the English would be grammatically correct. This is great.
However, the English might be too wordy, stylistically inappropriate, or lexically complex and flowery (often the case with ChatGPT output) that the frustrated potential client does not even finish reading it and deletes the e-mail. The inappropriate language style sabotages the very purpose of the e-mail. Unfortunately, the student is often unable to evaluate these important language nuances, and is therefore unable to evaluate the quality of the AI output.
The ability to critically evaluate AI-generated content is paramount. Even many high-level English users I know, including coworkers and professionals in various fields, struggle to recognize different styles of English writing and their impact on communication. This gap in skills highlights the necessity of teaching students not only how to use AI tools, but also how to analyze and refine the outputs.
In many ways, education has always faced this challenge. So, the challenge of education that accommodates AI is not new, but perhaps slightly more obfuscated. Students still need to acquire the knowledge and skills to plan the desired output of AI, communicate that to the AI, and evaluate and revise the AI output.
The danger is that by adding AI, educators would get used to seeing AI-polished outputs and might have little way of knowing if students are learning what they need to learn to judge its quality.
Unfortunately, from research I heard at a recent AI education conference at Ming Chuan University, many teachers allow students to use AI tools, but few actually teach students how to use them, let alone properly evaluate the output. My guess is that many teachers have no idea what an effective AI workflow is and how to develop AI fluency.
To be honest, I do not think many people have really figured this out yet. This is a new age in technology, work and education.
Of course, if AI fluency becomes a major goal for Taiwan’s education, it would demand a revolution. It would necessitate a shift in education goals from rote learning reproduction in an exam-based system to creative learning generation in a project-based or inquiry-based system, such as in Singapore’s education.
Perhaps there is hope. Perhaps generative AI would provide the disruptive opportunity to finally revolutionize Taiwan’s outdated industrial education system to make it catch up with the realities of the modern ever-changing workplace and AI’s rapidly expanding role in all facets of society.
At the very least, the initiative of Minister of Education Cheng Ying-yao (鄭英耀) would need to include comprehensive teacher training to ensure that educators not only understand the difference between AI literacy and AI fluency, but are also equipped with the skills and resources to train their students on AI workflows to enhance rather than diminish learning.
However, if the examination-based system remains intact and unchallenged, teachers would have little incentive to resist the perennial “teach to the test” pressures. The creative learning opportunities that generative AI offers for Taiwanese students and the future workforce would also be ignored.
Nigel P. Daly is a business communications instructor at the Taiwan External Trade Development Council’s International Trade Institute. He has a doctorate in English from National Taiwan Normal University.
Chinese Nationalist Party (KMT) Chairwoman Cheng Li-wun (鄭麗文) on Tuesday returned from her visit to the US, where she had met with several US senators and representatives, think tanks, and overseas Taiwanese and Chinese communities, espousing her vision of promoting cross-strait dialogue as a way to evade war. On Monday last week, she sat down with US academic Orville Schell and Asia Society Policy Institute distinguished fellow Daniel Russel. Russel is a seasoned diplomat who has served in positions related to East Asian and Indo-Pacific affairs. At the conclusion of the discussion, Russel employed his considerable diplomatic experience and knowledge
I arrived in Taiwan when I was 18, fresh out of high school and trying to navigate university life. Everything felt unfamiliar. My broken Mandarin quickly revealed what I already knew: I was not from here. When I told people I was from Indonesia, many told me that I did not look Indonesian. This caught me off guard, not because it was cruel, but because of how casually it was said. Sometimes it came with curiosity, sometimes surprise, occasionally admiration for my “fair skin.” More often than not, the people asking meant no harm. It is understandable. I look ethnically
Nearly three decades after the 1997 handover, Hong Kong has become a paradox: a place where financial dynamism coexists with political repression — a reality Taiwan cannot afford to ignore. Despite a sharp contraction of civic freedoms since 2020, Hong Kong remains a major global financial center. It has recently overtaken Switzerland as the world’s largest offshore wealth hub, supported by steady inflows of capital from China and across Asia. The territory continues to serve as a key intermediary linking Chinese firms, multinational investors and global markets. Institutional continuity has played a role. The Hong Kong dollar’s peg to the
Five weeks since US President Donald Trump met with Chinese President Xi Jinping (習近平) in Beijing, his administration now faces two important tests in the Indo-Pacific. The first test is one of the president’s own making. After meeting with Xi, Trump described arms sales to Taiwan as “a very good negotiating chip” and announced that a previously approved sale was now “in abeyance.” The Financial Times later reported that Beijing is holding approval for a visit by Elbridge Colby, the undersecretary of defense for policy, pending the president’s decision on the sale. During Trump’s first term in office, he regularized the arms
RSS