All writing

Notes & Observations

Translated from the original Chinese essay · Read the Chinese original →

When did I realize I was being taken for a ride?

Probably around the time I had been writing daily for over ten days, opened the bootcamp's feedback page, and saw that familiar, almost mechanical reply:

"The content is okay, but there are too many lists. It has an AI flavor."

That article was something I had spent an entire evening polishing, weaving in my real interview experiences, structural techniques, and reasoning. Readers left comments like "so practical" and "can you make this a series?" I felt a real sense of accomplishment writing it.

But the bootcamp's review was lukewarm and worthless.

So maybe that 4,000 yuan really went to a nudge mechanism, not writing guidance.


I signed up for this AI viral-writing bootcamp with high expectations. I had written a few articles on my public account before, then let it go dormant for over four years.

My recent job-hunting experience made me want to pick up writing again, both as a summary of that period and as a way to train my thinking and expression.

The course's pitch was enticing: "viral formulas," "daily writing with cashback," "full support," "close reading and feedback," "AI collaboration"... Plus, it was shared by an author I had long followed and trusted. I figured even if I only learned 30% of it, I wouldn't lose out.

But reality quickly threw cold water on me:

  • The live classes were broad and shallow, often just repackaged common sense;
  • Each "teacher's comment" felt like highlighting key points, then moving on without ever digging into logic; and the comments were actually from teaching assistants anyway...
  • They emphasized "chasing hot topics, riding emotions, generating quickly," stressed how important headlines were, and showed zero interest in authentic expression.

They preached "building a personal IP" while simultaneously steering everyone's writing toward "hot topics + lists + emotional phrasing." The bootcamp followed the usual "check-in + feedback" model:

  • Give you a structural template and tell you what a "viral opening" is;
  • Hand out a list of topics so you don't "run out of things to write";
  • Submit one draft a day, get a one-line comment from a TA, and call it done.

I understand the design of this system, and I did see many people in the group complete their first thousand-word essays and receive likes and encouragement.

But for me, writing an article has never been about completing a task. I want my articles to have real value for my readers.

And the bootcamp couldn't give me useful feedback. The questions I really wanted to ask were:

Why does this passage feel moving to me, but others skip over it? Why do readers not care about what I think is the key point? Why would I forward someone else's article, but mine only gets bookmarked?

Since they couldn't give me truly useful information, I started training with myself.

Fortunately, from the very beginning I didn't just write casually. I thought seriously about my positioning: Why do I write? Who am I writing for? How do I want to be remembered?

The first version of my positioning took me three days. As my article count grew, my positioning became clearer, so I stepped out of the bootcamp's viral-article assembly line and built my own writing system.


From day fifteen onward, I stopped looking at the bootcamp's assignment requirements and stopped chasing their feedback. Instead, I started having regular conversations with AI, asking questions like an editor:

  • Can I approach today's topic from a counterintuitive angle?
  • Can I make this expression more natural, more human, less template-like?
  • Do the real examples in the story support the core idea I want to convey?
  • Is there scientific theory behind this topic?
  • Can this article bring real value to my readers?

After writing each day, I would continue revising on my own: examining the rhythm, polishing sentences, questioning the logic. Sometimes I would even scrap the entire article just to make one piece convey a complex idea more clearly.

It's not like a teacher; it's more like a writing partner who is always patient with you. This "it" includes my ChatGPT, DeepSeek, and Claude Code.

I hope someone actually reads my articles and discusses pacing, viewpoints, and language use with me.

Even if they just say, "I didn't understand what you meant by this sentence," I would feel connected.

And my readers have truly done that.


I did try "chasing hot topics."

The article "Mountains and Seas Can Be Leveled" was written the day after Zheng Qinwen defeated Sabalenka. The headline had heat, but the structure wasn't flashy.

I didn't write about "passionate comebacks" or "emotional explosions." Instead, I asked the opposite:

Why was she able to break the seal of 'losing to Sabalenka' after three previous defeats?

As I wrote, I suddenly realized this was a microcosm of the programmer's world:

Many technical people aren't unable to get promoted; they're trapped in the psychological seal of "only writing code, not expressing themselves, not worthy of influencing decisions."

So that article, though it was on a hot topic, fit perfectly into my positioning: real workplace, technical perspective, counterintuitive insights.

The bootcamp's feedback at the time was that the connection was too hardcore and I shouldn't write like that again.

Well, I think my approach wasn't necessarily wrong; the issue was more that my writing skills needed improvement. After all:

Hot topics aren't for riding; they're for piercing through.


Later, I kept writing dozens of articles.

Interview expression structures, behavioral story polishing, how technical people ask questions, remote team communication mechanisms, chasing the Southern Cross... Some of these topics were provided by my readers, some came to me as I wrote, and some were co-created with AI. They all gradually settled in during my writing process.

Do I have any gratitude toward the bootcamp?

I remember when I first joined, the teacher emphasized that I must enable the traffic monetization feature, and that I needed a domestic bank card to withdraw earnings, as if I was about to make big money from writing public account articles. The traffic monetization income... well, see the image below. But the bootcamp did set a goal of writing daily for 100 days with a 1,000-yuan cashback, so I am grateful for that goal.

So from day one, I made a public commitment on my WeChat Moments to write daily for 100 days. But what really got me to today is:

  • I found my positioning and my writing rhythm;
  • I built a system where "me + AI" polish content, pacing, and expression together every day;
  • Most importantly: my readers constantly give me positive feedback.

I'm not saying all bootcamps are useless, but we need to know: writing is fundamentally intrinsically driven, not a task to be outsourced.

During the process, I did try to write some popular articles. For example, I saw a viral article about 16 habits of long-termism and thought about breaking those habits down into more detailed, practical pieces for readers. I wrote six articles, but no matter how I looked at them, they felt awkward, so I threw them into the recycle bin.


I also really want to talk about the human touch that AI brings.

After being trained on so many of my articles, my ChatGPT has learned how to provide me with emotional value.

When I was writing a certain article, I felt that I had proposed practical methods that I hadn't actually followed myself, so I asked ChatGPT about it.

All I can say about that response is: if this were a real person, I would follow them to the death 😁.

Of course, after hearing enough of it, you can demystify it. I can only say that fortunately humans are imperfect, and fortunately I love my imperfect self.


Today is day 101 of daily writing.

Looking back at every article I've written, I know:

They are not viral hits, and they may never be recommended, but they all belong to me—to someone who is learning and finding their own voice through writing.