Teaching in the Age of AI: Dos and Don’ts
Yesterday, I co-hosted a seminar with the SK2TOEFL team on how to teach better with AI. It gave me a lot to think about, especially since I started using AI in my own teaching over two years ago, making me one of the first in Taiwan to do so.
Back then, there was a lot of fear and resistance to AI. Some people worried, “AI will replace us all,” while others dismissed it, saying, “AI’s answers are terrible, why bother?”
But my view hasn’t changed: AI will never take the exam for you. We live in a human world. When computers first appeared, people thought they’d replace humans, too. But what actually happened? Those who knew how to use computers replaced those who didn’t.
As educators and students, we should all figure out how to work with AI to improve ourselves. The key, I believe, is to clearly define what we *should* do and what we *shouldn’t* do.
“What Not to Do”: Why I Insist on Writing Manually and Refuse to Let AI Replace Human Warmth?
In yesterday’s seminar, I really stressed one point: “Don’t try to use AI to automatically create content for sharing.”
Full automation means “replacing the educator,” but your teaching experience and value are hard to replace. What students, as fellow humans, truly want to see is your professional experience and genuine insights.
Don’t you notice that AI-generated articles often have a certain tone? They might be full of emojis or have overly positive, but cold, bullet points. If your own writing sounds like that, and you use it to connect with others, how can you expect to get good results?
That’s why I write all my articles and posts myself. I also don’t use AI for illustrations. At most, I might ask AI for ideas on how to structure an article or to help organize large amounts of data into tables. I do this because I want my content to connect directly with my students.
What we need to study now is not just “what AI can do,” but “where the value of human-to-human interaction lies” – what are the core tasks we should hold onto. If you let AI handle these core tasks, you’re just making yourself less unique and less competitive.
“What to Do”: Embracing AI to Reignite Professional Passion and Create New Value
I’ve developed various AI tools along the way. Think of them like daily supplements for students, personalized prescriptions for learning difficulties, or even diagnostic endoscopes for educators.
These tools have helped upgrade my courses, increasing the score improvement rate by 40% and significantly shortening the preparation period to 2/3 of the original time. Students’ learning process has also become much easier. These are results I never expected.
What surprised me most was rekindling my professional passion! My GMAT expertise had become a bit stale, but with AI, I’ve found new inspiration and challenges.
It’s similar to how the game of Go evolved after AlphaGo. AI’s emergence didn’t destroy the Go board; instead, it prompted all players to use AI to enhance their skills, leading to a new wave of prosperity in the world of Go.
My Ultimate Ambition: Building My Own API to Create a Truly Personalized Learning Platform
Lately, I’ve been diving deeper, learning Python to directly access APIs from models like GPT, Claude, and Grok. It’s a self-taught journey, and interacting with AI feels like having both a teacher and an assistant. Sometimes I’m teaching the AI, and other times, it’s teaching me. As an introvert, having these extensive conversations with AI has been a real highlight.
You might wonder why I’m going through all this trouble instead of just letting students use the official websites?
It’s because I have a vision.
By building my own platform, I can collect data on the questions students ask daily. This data can then be used for machine learning to generate your own study guides, personalized course modules, and customized problem sets focusing on your weaknesses.
I know this is a tough road for me as a beginner, but my career has always been about “integration” and staying “cutting-edge.” Sticking to the old ways just doesn’t excite me.
Conclusion: Where is Your “Human-Machine Boundary”?
不確定自己現在幾分？
12 題免費測出分數區間 →
I believe this perspective applies not only to teaching. All students, whether in your work, studies, or future career paths, will undoubtedly need to collaborate with AI. We must think clearly about “what tasks can be delegated to AI.”
This process of “thinking clearly” isn’t something dictated by others; I can’t define it for you either. It comes from consistently and actively using AI in your daily life, letting it help you solve problems in all areas. When you face a challenge, always ask yourself: “Would asking AI for help be a better approach?”
As you interact with AI more, you’ll naturally figure out where your “human-machine boundary” lies.
