中文版：
Around early 2024, I realized that my work investment was gradually pulling me away from society. I was falling into an endless hole of overwork—manpower shortages, workloads piling up daily, my body declining without exercise, and my energy draining away to the point where I couldn’t keep up with further technical training.
This work experience gave me a deep realization: “The busier you are, the poorer you become.”
Having time means keeping good habits, maintaining peak condition, and boosting both learning and work efficiency, which compounds into long-term returns.
Conversely, being too busy harms your body; the worse your health gets, the more money you need to spend to fix it. Meanwhile, you lose the time and energy to study or build expertise. To avoid this bad ending and instead create a positive cycle, the first step is simple: don’t burn out.
Taiwan’s work culture is a global oddity. Taiwanese workers are strong in competitiveness, but in terms of final output and innovation, there still seems to be a gap compared to Europe and the U.S.
That’s when I began to wonder—
Should I switch tracks and explore healthier work environments?
Should I move toward international companies? If so, English is the first hurdle.
So, with that dream in mind, I quit my job without a safety net—XD.
---
Since childhood, I’ve never been particularly good at studying. Among all subjects, English was always the one that gave me the biggest headache. I tried many approaches:
Buying vocabulary books and flipping through them occasionally (but in the end, only a few words stuck).
Preparing for the TOEIC exam and pushing for a higher score (my score went up, but only my “test sense” improved—not my English).
Watching American TV shows or YouTube without subtitles (helpful, but far too time-consuming).
Attending English cram schools (expensive, and while my notes looked neat, I didn’t believe a few classes a week would fundamentally improve my English).
Other methods like hiring a tutor or attending language schools, but I passed on them due to budget limits.
With limited time and energy, I couldn’t afford to keep failing over and over again.
So I set a new plan: in three to five months, I would systematically build a full English learning environment for myself.
After all, my ultimate goal was to work in an international company. Beyond the vague goal of “improving my English,” I needed clear skill benchmarks:
The ability to interview in English.
The ability to communicate even with limited vocabulary.
With clear targets, I could then design matching strategies and act on them.
---
To make sure my plan was truly effective, I knew I needed a strong theoretical foundation. I decided to avoid unproven “methods” and start from academic research and papers.
Of course, as a newcomer in the field, I knew I couldn’t possibly read all related papers in a short time. So the key was finding the right research directions.
I divided my investigation into three parts:
The basic cognitive processes of learning – The Learning Brain
The brain’s process in language learning – Science of Reading
Whether there were mature learning products on the market – Speak
---
When I thought about “how to learn English more effectively,” I realized it wasn’t just a matter of methods, but about how the brain itself learns.
So I turned to The Learning Brain, a well-regarded course taught by University of Michigan psychology professor Thad A. Polk. It covers psychology, neuroscience, and educational technology, and explained things clearly and deeply.
Here are a few insights that resonated most with me:
💡 Learning ≠ Memory. Learning is the ability to change behavior through experience.
Learning is defined as the ability to change behavior through experience, while memory is the product of learning. For example, you might learn a fun fact about Berlin airport, even if you never use it in behavior. Conversely, some learning is unconscious—like habituating to the sound of a fan after sitting next to it for a while.
This points to two truths:
Learning requires memory.
If you’ve learned something, it means you can draw on past memory to influence current behavior.
💡 Memory happens in three stages: Encoding, Storage, Retrieval.
1. Encoding: Converting new info into a storable form. This decides if today’s input becomes tomorrow’s memory.
2. Storage: Strengthening neural pathways to keep info in long-term memory.
3. Retrieval: Accessing memory when needed (e.g., in decisions, conversation).
💡 The brain has multiple learning systems.
Different tasks use different regions and pathways. For example:
Type Feature Example
Explicit Learning Conscious, verbalizable Memorizing vocab Implicit Learning Subconscious, non-verbal Riding a bike Working Memory Short-term info processing Mental math Semantic Memory Knowledge of the world “Birds fly” Episodic Memory Personal events “Studied English at a café”
💡 Memory is reconstructed, not replayed.
“Our brain is not a video recorder, but a story editor.”
We rebuild memories from fragments, which makes learning strategies about connection and reconstruction more effective.
💡 Effective learning requires the right strategies:
Spaced Repetition – review over time, not cramming.
Interleaving – mix topics instead of block practice.
Retrieval Practice – recall actively instead of rereading.
---
After studying the biological mechanisms of learning, I turned to Design for How People Learn. This book takes a design-thinking approach, treating learning as a user experience.
💡 Know your learners. Different learners (novice, practitioner, expert) have different pain points. Many English learning methods are built for advanced learners, not beginners.
💡 Build mental models. Learning isn’t just remembering facts, but giving information “a place to live.” Organizing knowledge into clear frameworks improves recall and application.
💡 Curse of Knowledge. Experts often can’t empathize with beginners’ struggles. This explained why grammar teaching often felt so inaccessible—and inspired me to design my app around real usage contexts rather than just grammar rules.
---
This framework was a huge turning point. It showed me that not all struggles come from weak vocab or grammar—sometimes the underlying cognitive systems aren’t built yet.
For example:
Can’t sound out a word? → Weak phonics or phonemic awareness.
Can read aloud but no image comes to mind? → Weak vocabulary or comprehension.
The framework breaks learning into stages like phonemic awareness, phonics, fluency, vocabulary, and comprehension—each with strategies to target them.
---
I chose Speak as my learning platform. It’s a speech-practice-focused app that emphasizes: “You only really learn when you say it.”
Users practice speaking out loud and get immediate AI feedback, simulating real conversations. The best part is that Speak integrates principles from the Science of Reading—helping learners build strong phonemic awareness and phonics skills from the ground up.
---
Building My Own AI English Learning App
After analyzing theory and existing tools, the next step was clear: build my own English learning system.
My biggest challenges were in memory:
Storage: Hard to organize and maintain vocab, grammar, and examples.
Retrieval: Unsure when and how to flexibly use what I learned.
The first problem could be solved with spaced-review algorithms. But the second usually needed live tutors—until AI made simulated practice possible.
So I built an app where users could practice speaking naturally, with AI simulating real contexts and giving feedback, completing the learning cycle.
![App screenshots]
---
After building the app and combining it with Speak and ChatGPT, I felt my English ability take off like never before. This project directly helped me land my first job through a full English interview.
But since I had already built a proof of concept, I didn’t want to stop there. Last September, I visited a professor at NTUST just to seek advice—but the professor encouraged me to apply for graduate school. I did, and now I’m a master’s student in Digital Learning and Education.
My hope is to integrate these theoretical frameworks and technologies into a system that can truly deliver personalized, customized learning experiences.
The app is still in Alpha stage. Since I later secured a job, it hasn’t been updated for a while—but if you’re interested in the system or the topic, feel free to reach out! I’d love to discuss or co-create new possibilities.
