About me

I am Tree. I share my experiences with clear communication in English—not accent drills.

The goal of learning a language is to communicate. For me that means expressing your own ideas clearly—so someone else can follow, trust, or act on them. English is no exception. You need a logic framework, and everyone needs one that feels natural in their own mouth, not a template they memorized and forgot.

This site is the public home of how I do that: real use cases, output loops, AI only where it lowers friction, and frameworks you refine until they travel with you.

The moment I stopped blaming vocabulary

I had years of input—listening, reading, collecting phrases. I still failed a test I invented: introduce a singer you love. Blank stares, half-sentences, translation in my head until the moment passed. Around the same time I was working and squeezing in TOEFL prep. New prompts looked endless. What helped was not another list of pretty sentences. It was one reusable structure I could adapt: state a point, say why it matters, show one example with a result tied back to the point, then close.

When I started writing and making content in Chinese, the pattern repeated. Sometimes my native language was fuzzy too. That made the diagnosis simple: the bottleneck is expression—do I have a thought, and can I lay it out so someone trusts it? English is the same job in a different language.

What “good English” means here

Good English is clear communication in a real scene: explain an idea, pass an interview, publish a post, hold a meeting line. If you only say “I want to get better,” feedback stays vague. Name the conversation first; then build or adapt a framework for that conversation.

So the first move is always application scenario: name the situation, choose material that matches it, and measure progress by what you can say or write there—not by hours logged.

The four-part system (one formula, many notes)

Across my notes and public posts, the same equation keeps showing up:

Learn English well = use case + feedback loop + lower friction + framework modeling

• Use case. Do not wait until you feel “ready.” Pick a vertical, a job task, or a personal IP thread and learn through that lens.
• Feedback loop. Match input with output. Retell what you heard, review against the source, repeat until one idea sounds like you—not until you finished another episode.
• Lower friction. AI is a lever: practice when there is no partner, summarize the scenes you actually talk about, mark logic gaps on a transcript. It is not the learner; you still decide what you mean.
• Framework modeling. Collecting quotes without slots is like stacking wood with no house plan. I use a simple three-column habit—scene / how I said it / better native phrasing—and I tag lines by job (open, justify, example, close). One framework can cover many prompts once the skeleton is familiar.

I still believe less is more: one piece of material per week, squeezed through listening, speaking, reading, and writing, with your own decisions about what to keep, beats passive volume or letting AI shallow-parse a hundred files for you.

What I make

I write method pages, speaking practice routes, and an AI speech coach skill that follows the same rules—real task, real output, reusable frame. If you see yourself in the “studied forever, still freeze” story, start with understand but can't speak or the overview at how to learn English effectively.

— Tree · building in public while I share what I actually use.