Claude: You Want Me to Teach You How?

No.
100
Date
Read
9 min
Words
2,003

I first chatted with an AI on March 28, 2023, nearly three years ago. Back then I only dabbled and never went deep. Now AI is developing so fast that it has entered almost everyone's field of view.

AI also shows up in my life, and gets used, more and more often. These days the AI I use most is Google's Gemini Pro, which came free for a year through Google's student offer. I have to admit AI can now answer almost any question I have, and it gives me a lot of professional advice in many areas.

What amazed me most was the time I wanted to build a local blog on Hugo. All I knew about it was the sentence "I want to build a local blog on Hugo." I didn't know what Hugo was, and I didn't know what a local blog was. Even so, I got it built within three hours.

I started with "I want to build a local blog on Hugo, please give me a plan," and followed the process Gemini gave me. First I had to set up a repository on GitHub, then get a web address, set up Cloudflare, and so on. The steps themselves aren't actually that many. But for a complete beginner who knows nothing, even clicking into a web page and finding a button is something someone has to teach you by hand.

So whenever I didn't know where to click, or my action produced an error or something weird, I took a screenshot and sent it to Gemini right away. I didn't even need to describe it. It knew what my problem was. That's how I completed the whole process on my own.

That's AI: you can ask it anything. And to this day, Gemini is still the model I use most.

Today I opened a new model called Claude. From what my friend said, this model feels very human, which made me curious.

So I tried chatting with it about a few things, and I found that when it comes to discussing writing, Claude really is different from Gemini.

If Gemini and Claude were two people, Gemini would be a fairly earnest researcher, an answerer who replies to everything you ask. Claude would be a bit of a maverick, better at asking you questions and forcing you to think.

What I found most interesting and incredible is this. I talked with Claude about a post I'd written recently, all the way to a conclusion. When I looked back over the chat afterward, I realized I'd gone from the one asking questions at the start to the one answering them nonstop.

And in the course of answering, the problem I wanted to solve gradually became clear, and I ended up with a result I was quite happy with.

That blew me away. I think many people have a stereotype of AI: it should answer people's questions. But Claude is different. It also centers on solving problems, but instead of rigidly spitting out theories and methods, it keeps asking questions so the asker thinks deeply about their own problem, and the asker writes out their thinking in words. After a series of thinking and output, Claude gives the final result.

The benefit of this is that the asker feels deeply involved. After all, the problem to be solved is the asker's own. If you directly hand over a list of suggestions or doctrines, it makes the asker dizzy, because they never went through the process of thinking. With the thinking process, you have clarity when you face the final plan.

I have to admit, every large model has its own strengths. Knowing what you need and picking the right model for each job seems to be an important skill in the age of AI.

Here's my chat with Claude today:

Writing Chat Log

Discussion of the writing analysis and restructuring of "Jobs, Mao, and Writing"

You

I don't think this post achieves the effect I wanted.

Claude

Interesting. Where do you think it falls short? Is the argument itself not strong enough, or does the expression and structure not really bring out your thinking?

You

I think my argument is strong, but my writing isn't up to it, so it didn't come out with the effect I wanted.

Claude

Many writers feel this. What's in your head is three-dimensional, and on paper it goes flat.

Can you tell me which part makes you feel it falls short when you read your own post? Is there a spot where you know it could be better but don't know how to write it?

You

I think the post is a bit messy. There are some paragraphs that could be cut, like paragraph 10. Also, I think the logical structure isn't clear. There is logic, but reading it doesn't give a sense of building up layer by layer. For example, my core argument is that minimalism isn't about less... and then I extend that to personal branding. But the content I discuss doesn't stay tightly on that goal, so it's a bit scattered.

Claude

Your self-diagnosis is spot on. Essentially, the post has two "halves":

**First half:** the philosophy of minimalism (Jobs and Ive's design thinking)

**Second half:** the writer's personal brand (Mao, Dan Koe, Hecaitou)

But the bridge between the two halves isn't built firmly. Readers follow you through the first half, then suddenly find themselves somewhere new, not knowing how they got there.

Your core logic is actually very complete. One line can string it together:

Minimalism = digging deep into complexity and distilling the essence → a writing brand works the same way = dig deep into yourself and distill the essence → so write your real self over the long term

That line exists. It's just drowned in a lot of side material in the post.

Let me ask you one question: which sentence in the post comes closest to the core of what you really want to say? Just one sentence.

You

Dig into yourself

Claude

With just those three words, the problem with the whole post becomes clear.

**Your post's structure right now is:** Jobs → minimalism → personal brand → dig into yourself

But if the core is "dig into yourself," the structure should flip to serve it: What is "digging into yourself"? → Why is it hard? → How do you do it?

Jobs and Mao are only examples, not the main characters. The problem now is that you spend too much space explaining minimalism itself, which makes readers think that's the point. By the time "dig into yourself" shows up, the post is nearly over and you're out of steam.

In other words: you spend 80% of the space on setup and 20% on the core.

Do you think that's accurate?

You

Yes

Claude

Then the next step is clear.

You face a choice: reorganize the existing material, or rewrite the post?

The difference between the two: if you reorganize, you'll find some paragraphs hard to place, because they were written around "minimalism," not around "dig into yourself." Forcing them in only makes it messier.

Rewriting sounds expensive, but your thinking is already clear, so writing may actually go more smoothly.

Which way are you leaning?

You

Rewrite

Claude

Good. Before you start writing, let's build the skeleton.

Your core is "dig into yourself." Around it, there are three questions you need to think through:

  1. What does "dig into yourself" mean for a writer, specifically? How do you define it?

  2. Why don't most writers do it? Is it that they don't know how, they don't dare, or their method is wrong?

  3. What does each of the Jobs and Mao examples illustrate for you?

Think about these three questions first, and then we'll build the structure.

You

  1. "Dig into yourself" means that an individual writer should go deep into their own life and understand their own experiences and thinking. These are a writer's treasure trove.

Claude

That definition is good in itself, but I'd like to push one level further: why are these experiences and thoughts a "treasure trove"?

Because the last line of your post actually already answers that: "only well-matched readers have value." And the premise of being well matched is being real, and the source of being real is your own experiences and thinking.

So behind "dig into yourself" there's a deeper logic: your uniqueness exists only in you, and you can't find it anywhere else.

Do you agree with that addition? Or do you have another layer in mind for "treasure trove"?

You

Because for an individual writer, their own experiences and the thinking they do in those experiences can't be copied by others. And finding the points worth writing about in them is what makes them one of a kind.

Claude

Right, that logic is solid. Distilled into one sentence:

Uniqueness isn't designed. It's dug out.

That line could actually be the core sentence of the whole post. Jobs and Mao both serve it. One dug out "simplicity" through products, and the other dug out "seek truth from facts" through essays. Neither's "brand" was a label designed in advance. It emerged naturally after long, deep digging.

Now the first question has an answer. Let's move on to the second: why do you think most writers don't "dig into themselves"?

You

I think most individual writers share one flaw: they always want to write something profound or revolutionary, and they lack a down-to-earth, hands-on spirit. Honestly, I used to be like that too. Only later did I realize that to write, you have to do things unrelated to writing and think within your experiences, and only then can you write real things. So I think most writers fail to dig into themselves because they don't realize how valuable their own thinking and experience are.

Claude

That passage is itself material for a good post, because it comes from your own experience and doesn't preach.

**"To write, you have to do things unrelated to writing"** is a powerful line, and counterintuitive too. Readers will stop and think.

You've answered two of the three questions. Let's look at the third, and the most crucial: what does each of Jobs and Mao illustrate for you?

You

With Jobs, the idea that simplicity is the ultimate sophistication made me think that a person can write ten thousand posts, but those ten thousand may all be saying the same few things. I understand this as: all of a person's posts should use different angles and subjects to write, again and again, the few views, theories, and truths that belong to the writer. For example, although Jobs created Apple and designed the iPad, iPhone, and iCloud, he was really using these products to say one thing: simplicity is the ultimate sophistication, keep a Zen spirit. And with Mao, I want to show that building a personal brand, and the views, theories, and truths that are uniquely yours, take long accumulation to form.

Claude

Now the skeleton is completely clear. The logical chain of the whole post is:

Individual writers need to dig into themselves → because your experiences and thinking are one of a kind, and uniqueness is dug out, not designed → but most people don't, because they always want to write "big things" and ignore down-to-earth accumulation → how do you dig? Jobs tells you: in your whole life you're really only saying a few things that truly belong to you, and you present them again and again through different works → Mao tells you: those few things can't be found today. They take long accumulation to surface

The two examples have a clear division of labor: Jobs covers "what to dig," and Mao covers "how to keep digging over the long term."

This line is much clearer than the original.

Now you can start writing. I suggest you don't look at the original. Just rewrite it along this logical chain, then look back at the original and add in any details that are still useful.

Licensed under CC BY-NC 4.0. Please credit the author and link to xiluluke.com/en/ai01

Xiluluke

A dialectical minimalist writing about life, reading and thinking.

Comments

0 / 2000Plain text · Line breaks preserved
Comment as

Loading comments…

    Your browser remembers your name and email once you post.

    Related writing

    WeChat · Xiluluke

    WeChat QR code for Xiluluke

    Scan with WeChat, or long-press to save the image.

    Open original ↗

    Buy me a coffee · Xiluluke

    WeChat reward code for Xiluluke
    WeChat
    Alipay code for Xiluluke
    Alipay

    Scan with WeChat or Alipay, or long-press to save the image.