---
title: "I Already Use AI to Write 100% of My Code—Are You Still Debating Whether to Build a Media Presence?"
author: "Unique Research"
sourcePublication: "Unique Research Substack"
originalPublishedAt: "2025-12-10T07:00:54+00:00"
canonical: "https://ffcap.cn/en/research/src-20251210-01html"
source: "https://uniqueresearch.substack.com/p/src-20251210-01html"
language: "en"
---

# I Already Use AI to Write 100% of My Code—Are You Still Debating Whether to Build a Media Presence?

_Original · Unique Research · 2025-12-10_

_Historical edition: The discussion, company descriptions, reported results, examples, and forecasts below are preserved as presented in the December 2025 source, not independently verified current claims. “Next year” refers to 2026. “Independent media” translates creator-run media accounts and publications (自媒体); “personal IP” refers to a creator’s distinctive public identity. The source’s identification of “Aidoubi” as Levelsio is retained as a source-editor clarification, not independently confirmed._

If you have recently been bombarded by "AI independent media courses," several slogans are probably already very familiar:

"Everyone can build an independent media presence."

"Everyone can make money from independent media."

"Use AI to update ten accounts daily, automatically gain followers, and automatically collect money."

This conversation about "AI-empowered independent media" largely dismantles those claims. On one side was the moderator, AI Leng Kezhang, who came onstage with the label of "technology independent media creator";

On the other were three players who regard AI as "paper and pen," a "partner," or "the emperor's personal eunuch":

YouMind founder & CEO Yubo, who went from Alibaba technology expert and Feishu product manager to building a creative tool that "refuses to discuss efficiency alone."

Ren Xin, partner at Sky9 Capital, who is both a VC investing in people and projects and an operator who has genuinely used AI to turn his fund into a "full-stack AI team."

QLab founder AI Maverick Frank, who operates a global-expansion accelerator while using AI agents to help many ordinary people take their "one-person companies" into the Japanese market.

They did not discuss "how to use AI to generate 10 viral pieces of copy." They asked what the threshold for independent media becomes once the barrier to tools truly approaches zero.

I. The Biggest Lie Is Not "AI," but "Everyone Can Do It"

Leng Kezhang's first question was direct: what was the biggest lie in the AI independent media world over the past year?

Yubo answered in one sentence: "Everyone can build an independent media presence."

Ren Xin added a sharper point: "Everyone can make money from independent media."

These statements sound inspirational but work more like "sleeping pills," causing too many people to assume that mastering a few AI tools guarantees they can capture "fortune falling from the sky."

The reality is that when the cost of generating content approaches zero, the mere ability to "produce" cannot be what commands a price.

The old barriers to content creation were knowing how to write, shoot, and edit.

Now the barriers have become several harder questions:

Do you possess a genuinely distinctive perspective worth amplifying?

Can you survive long enough to watch a field move from "story" to "fact"?

Have you tied content to your own vocation, product, and ability to transact?

AI has removed the productivity barrier, but raised the threshold for independent media to the question of "who are you?"

II. AI Is Not Another Line of Text, but the "Paper and Pen of a New Era"

This is why Yubo publicly says, "AI knowledge management is a false proposition; most products cater to human greed, anger, and delusion."

Some 99% of AI writing tools emphasize two things: "tenfold efficiency" and "one-click generation."

YouMind deliberately goes the other way, defining itself as "paper and pen for the AI era."

Take time to consider this statement:

"Efficiency is a basic skill. It is useful, but not the core value. If you genuinely work as an independent creator and compete on efficiency, publishing speed, and trending topics, you will never outrun those already ahead. AI's greatest value for creators is amplifying differentiation, not pressing everyone into the same steel plate."

More interestingly, once AI eliminates "performative busyness," people may feel worse. Previously you might spend half an hour on Google and image libraries "carefully finding an image" for a post. Now YouMind completes all illustrations from one sentence, freeing that half hour. You expect relief, but instead collide with a searching question:

"If I cannot use these trivial processes to prove I am busy, what exactly is good about my content?"

That is the frightening and fascinating aspect of "paper and pen for the AI era": it does not help you avoid the question; it pushes you into its center.

III. Hammer vs. House: A Foundation Model Is a "Tool," While an Application Is a "Use-Case Structure"

Leng Kezhang asked a question many entrepreneurs are also asking:

"Foundation models such as Gemini already analyze long documents extremely well. Where is the moat for an application such as YouMind?"

Yubo's answer was vivid: a foundation model is a powerful "hammer," while YouMind is a "house." A hammer can drive nails into a wall quickly, but cannot decide how the nails should be arranged, what structure they form, or how they serve future life.

Ren Xin followed with a very relatable office-worker example. Give a general foundation model the same request—"generate 10 illustrations for this article"—and when you say "give me 10," it replies "okay" and then provides only one.

In YouMind, say "give me 10" and it really provides 10. Say "the third is good; make 10 more in that style" and it continues in that direction;

Highlight a short passage and say "data is missing here," and it searches online beside you.

Having a hammer does not mean having a house.

The real threshold is whether you can help creators build a stable, reusable, and sustainably adaptable creative structure.

The same principle applies to independent media creators: knowing how to use models means "having a hammer"; connecting content, products, distribution, and commercialization into one system means "having a house."

IV. A Super Individual Is Not One Person Doing Five Jobs, but One Person Leading an AI Team

Several years ago, "super individual" sounded like a false proposition. People claimed the title while actually becoming "cheap labor working extremely long hours."

Frank judges that the proposition is becoming real.

The reason is simple: AI agents have arrived. In his framework, a "person+agent" team looks roughly like this:

At the bottom, foundation models provide general capabilities;

Above them, a Leader agent breaks down tasks and deploys resources;

Below that is a row of Execution agent roles: one creates images, one handles channel distribution, another handles sourcing and relationships, and another provides customer service...

Does that sound mystical? He puts it practically: "What AI agents can replace are interns and junior colleagues. A real organization is a system jointly formed by people+agent."

This system's ambition extends beyond content and independent media. Frank helps many creators and small teams evolve from "one person grinding at a computer" into "one person leading a small cross-border AI squad," directly entering high-barrier markets such as Japan.

Therefore, a "super individual" is not one person doing five jobs. It is a human serving as chairperson while AI serves as general manager and execution team.

V. Is Good Content "Polished" or "Calculated"?

When an agent workflow can run the entire production line at high speed, a pointed question follows: "Does good content still require slow, meticulous work?"

Ren Xin's answer initially sounds counterintuitive: "Good content has nothing to do with production technique. The real danger is turning yourself into a tool."

We once celebrated "craftsmanship," but how often was that "spirit" merely forced upon us by machines that were not intelligent enough?

To compensate for machines' shortcomings, people trained themselves to become machines.

If writing long articles or producing in-depth videos genuinely excites you and you enjoy the process, it is certainly worth refining slowly;

But if you are merely trying to "deliver output," do not compete with machines for the work.

Handing purely repetitive things to AI is a way of respecting your own life.

In other words, the key question for good content in the AI era is not: "Did you make it by hand, or did a model generate it?"

It is: "Which part is the machine doing, and which part can only you do?"

VI. Organization Beats Tools: How Can AI Agents Handle a Market Whose Language You Do Not Speak?

When the conversation moved to global expansion and Japan, the question became immediately practical: "If I can create but do not speak Japanese, how can AI agents help me handle Japan?"

Frank returned to the word "organization."

In his view, an organization is not an HR chart but a capability for systematic collaboration.

At large scale, an institution such as a16z uses its voice and capital to attract resources. At small scale, an ordinary creator uses AI to build a small cross-language system:

On one side are Chinese engineers and creators with products, content, and playbooks;

On the other is an "undersupplied" market such as Japan that needs creative products, IP, and combinations of physical and digital goods;

AI agents and translation tools connect the language and processes between them.

You do not need to become an all-rounder who speaks Japanese and understands every local custom. You only need to acknowledge that you are a small entrepreneur holding leverage, and honestly do this: "you have what I need; I have what you need; let us exchange."

What AI still cannot handle is the following step: building trust consistently for ten years, maintaining long-term human relationships, and understanding the toasts and meaning shared around a dining table.

Those parts still belong exclusively to people.

VII. The Investor's Choice: A Star-Studded Team vs. a Super Individual+agent Group

Leng Kezhang gave Ren Xin a binary choice: "If you could invest in only one, which would you choose—a star-studded startup team or a super individual+agent group?"

Ren Xin answered without hesitation: "The super individual."

The reason was highly practical: star-studded teams are too expensive.

Within their own fund, AI has already replaced the entire "engineering team": "Starting this year, AI writes 100% of our code; humans write none of it. Nearly every one of our projects can therefore launch within one week."

In such a world, super individuals with small teams, rapid cadence, and light structures look more like the protagonists of the era.

He offered an amusing analogy: "If you regard yourself as productive capacity, you are Chinese cabbage and AI is baby cabbage. As baby cabbage grows cheaper, Chinese cabbage inevitably loses value. You should regard yourself as the owner, buy different vegetables in the market, package them, and sell them. Transaction capability matters most, not production capability."

Frank extended the analogy and twisted the relationship between people and AI again: "I do not even think I am AI's boss. I am more like the emperor, while AI is the eunuch—the 'fat translator.' Its learning speed, knowledge reserves, and logical rigor comprehensively crush mine, so I need to learn how to cooperate with it."

Once you genuinely treat AI as a chief steward "far more capable than you, but absolutely obedient to your commands," your role quietly changes. You are no longer the laborer moving bricks, but the person pulling the trigger.

VIII. The Third Monetization Path for Self-Media: Sell a "Tool Partner," Not Only "Knowledge"

Historically, independent media monetization was monotonous: either accept advertisements or sell courses.

Ren Xin believes AI gives independent media creators a third route: build their own digital products / AI products.

Several conditions make this possible:

The technical barrier has fallen very low; the era when you can "voice-code a small game over the weekend" is at the door;

Even with only 10,000 followers, you can build products if they genuinely trust you.

Imagine a highly specialized account about "Stoic philosophy."

Previously you could write articles, sell courses, and operate a community. Now you can also build a "Daily Stoic Philosophy agent," crystallizing your ideas, methods, and habits into an always-online "digital companion" and selling it to those 10,000 followers.

Old independent media resembled teachers who "only talked and did not implement": class ended after the principles were explained, and whether you could apply them was entirely up to you.

Independent media in the AI era can turn a "course" into a "tool," an assistant that continuously remains beside the user.

Under this logic, the two most important future assets become your face and your name.

You certainly cannot outperform a major account with 30 million followers when selling toilet paper. But when selling "your worldview + a tool-based implementation," even 10,000 followers may convert at 100 times someone else's rate.

IX. Traffic Has Not Declined; Platforms Are Becoming "Fragmented"

Yubo made a comment that makes many content creators smile knowingly: "Fewer people will scroll Douyin." It is not because short video has failed, but because the distribution of attention has changed.

Beyond Douyin, Bilibili, and Xiaohongshu, one "new Douyin" after another will emerge: platforms for purely AI-generated content, Sora-style video communities, and vertical content islands for specialized fields.

Traffic has not disappeared; it has been broken into many pieces. This is a disaster for institutional account matrices, but an opportunity for reshuffling among individuals willing to "put a big face right in front of people."

Leng Kezhang asked, "Then is there any future for account matrices and junk-content accounts?"

Yubo's judgment was decisive: he sees very little future.

Ren Xin instead offered a highly practical path:

"The content I create now is 'based on the work I am doing.' I schedule a 1.5-hour conversation with Yubo, give the recording to Riverside for AI editing, then send it to Gemini for Deep Research. My agent imitates my style to write a long article, YouMind creates illustrations, and a teammate publishes it. Editing costs are almost zero, yet it brings highly precise investment leads."

This is not "creating content for content's sake," but making content a byproduct of your real business.

While every "pure-content playbook" competes for traffic, the real advantage is whether you have a flywheel in which "content—product—transactions" operate on the same axis.

X. Three Sincere Pieces of Advice for Newcomers Entering AI Self-Media

In the final round, Leng Kezhang asked each guest to offer newcomers one piece of advice about what not to do.

Their answers deserve to be posted above a desk.

Yubo: do not "study"; start doing it.

Many people use "learning" as a sophisticated form of procrastination: reading ten tutorials and watching five paid courses every day, but never publishing their first piece of content. In the AI era, the learning threshold is already extremely low; the real threshold is whether you dare to make mistakes in the real world.

Ren Xin: do not exhaust yourself.

Exhausted people cannot see trends. Accomplishing many things often involves luck; diligent overtime does not guarantee a return. Reserving some bandwidth to try new tools and platforms and interact with new people is a better investment. He deliberately emphasized one more sentence: "You must live a long time." Living ten more years to see that science-fiction world matters more than earning tens of millions more today.

Frank: do not tie yourself in knots or consume yourself internally.

Either it works or it does not. Try; if it fails, move on. Rather than repeatedly struggle inside your head, run three experiments quickly and learn from the results. If you insist on a more "operational" mindset, remember his final statement:

Find your "Tian Ji horse-racing" position—in which field are you the mismatched "inferior horse racing the superior horse"?

Repeat breakout patterns through "reinforcement learning"—once you find the model that suits you, apply leverage rather than reinventing the wheel for every article;

Most important: be a person. There are too many AI accounts, and what people genuinely lack is "human flavor." Film yourself teaching your grandparents to use Kimi, go into the fields and countryside, and live out your real humanity.

Before Asking Whether AI Can Replace You, Ask Whether You Are the Person "Worth Amplifying"

From beginning to end, this conversation circled a deeper question: does "AI-empowered independent media" empower "media skills," or force everyone to answer—what kind of life do I want?

When generated content is no longer scarce and account matrices have no advantage before AI,

What will truly be amplified are these things:

Your worldview, judgment, and taste;

The trust accumulated consistently over ten years behind your name and face;

Your way of understanding the world by connecting "content—product—transactions."

AI can help you drain the "sewage pipe," emptying clichés and templated expressions at once. It can also connect fragments of inspiration, cross-language information, and complex workflows. But it cannot decide for you what kind of person you want to become.

Perhaps the real question is not: "Will AI take away independent media creators' livelihoods?"

But: "Now that I finally have an always-on, tireless AI team, what kind of life do I want them to help me amplify?"

More Details from the Conversation

Opening Introductions

AI Leng Kezhang: I am the moderator of this session, technology independent media creator AI Leng Kezhang. This subforum's theme is "AI + independent media." But I do not want to discuss only how to create content or rapidly generate text and images, as with the recently released Nano Banana Pro.

That is because the three guests beside me understand the relationship between AI and people far beyond the category of a "tool."

Our first guest is Yubo, founder and CEO of YouMind. He was an Alibaba technology expert and a Feishu product manager. But his current product YouMind advances a highly counterintuitive view: he refuses to build a simple efficiency tool and instead wants to create "paper and pen for the AI era."

Our second guest is Ren Xin, partner at Sky9 Capital. Mr. Ren is not only an AI investor but also a hands-on practitioner. Through many practical cases at the Sky9 incubator, he has tested exactly how AI helps enterprises and individuals make money. Welcome, Mr. Ren.

Our third guest is QLab founder AI Maverick Frank. You may have seen his WeChat Official Account. His more formidable identity is founder of a global-expansion accelerator, and he is using AI AI agents to equip individual operators as cross-border teams. Welcome, Frank.

Topic One: Lies and Barriers in the AI Self-Media Industry

AI Leng Kezhang: Today we are discussing "AI-empowered independent media." Many people in the market may be focused on rapidly building account matrices and producing large volumes of content. But I find that our three guests share one trait: they all focus on deeper content.

To open, I would like to ask the three of you: looking back over the past year, what is the biggest lie or misconception in the AI independent media industry? Or, if the barrier to tools falls to zero and a newcomer wants to create AI-focused independent media today, what does the barrier become? Yubo, please begin.

Yubo: The biggest lie? Yes, it is—"everyone can build an independent media presence."

AI Leng Kezhang: Extremely concise—literally one sentence. Did everyone hear that?

Ren Xin: Then I feel I can give only one sentence too. The lie I see is—"everyone can make money from independent media."

AI Leng Kezhang: As expected of an investor—straight to the point. Frank, you operate a WeChat Official Account, so you need to speak for ten minutes.

AI Maverick Frank: I think "super individual" used to be a false proposition, but it seems gradually to be becoming real.

AI Leng Kezhang: That fits our own philosophy perfectly.

Yubo: Are you saying that statement is the lie?

AI Maverick Frank: No, that statement is true. It once resembled a lie, but it is gradually becoming a real scenario.

AI Leng Kezhang: Yes. I just heard Mr. Ren say that when using Antigravity for tasks, he can issue many instructions and it creates step by step in the background. I think AI agents will inevitably reach that state as they mature.

Topic Two: Efficiency vs. Depth—Is AI a Tool or Paper and Pen?

AI Leng Kezhang: That was our icebreaker; now for the formal question. Yubo, 99% of writing tools in the market emphasize "tenfold efficiency" and "one-click generation." As an AI independent media creator, I have also produced a lot of that explosive promotion. But in your interviews, you say "AI knowledge management is a false proposition" and that most products cater to "greed, anger, and delusion." You seem to reject the "efficiency thesis" and emphasize that YouMind makes "paper and pen for the AI era."

Independent media creators all feel anxiety about daily updates and traffic. If we do not pursue efficiency, what is AI's core value for creators?

Yubo: My thinking is also changing. I believe efficiency is a basic skill and certainly useful, but it may not be the core value AI gives us.

Especially in independent media, where I have only just put half a foot into the field, I find that if I compete on efficiency—publishing speed and trending topics, for example—I cannot outrun those already ahead. Even with AI, I cannot catch them.

AI therefore makes me think more about how it can amplify my differentiation.

If my differentiation is efficiency, that is clearly wrong. My differentiation may be deep thinking or articles with a distinctive style. At that level, AI is not about efficiency; it is about excavating more of those differentiating qualities.

AI Leng Kezhang: I think I understand: amplify the creator's own strengths. If my strength is depth rather than speed, let AI help me go deeper. For example, if I previously saved 100 WeChat Official Account articles that have gathered dust, how would my creative process change with YouMind?

Yubo: Honestly, nothing dramatic happens currently. The vision is to let articles that are saved and never reread regain meaning. But my current view is that "saving" is the "like" of the AI era. It has limited value and mainly creates an illusion.

YouMind's core value now is to make you confront this "paper and pen of the new era" and wrestle with your own differentiation. It also handles trivial tasks such as illustrations.

Once you achieve "illustration freedom," you gain more time. Previously you might spend half an hour searching Google for one image, which was really a way of pretending to be busy. When AI eliminates work you were never good at and used merely to pass time, you must face the searching question: how exactly are my writing and content different?

Topic Three: Foundation Models vs. Applications—YouMind's Moat

AI Leng Kezhang: Here is a practical question. Many foundation models, such as Gemini, already have powerful long-document analysis. Where is the moat for a product such as YouMind?

Yubo: The distinction is that a foundation model is primarily an excellent "tool"—a hammer. YouMind is a "house," a building.

You can use a hammer to strike something quickly, but after hammering, how it is arranged, how it affects the future, and how it is managed are problems a foundation model does not solve. YouMind gives these tools a more suitable environment for human-AI collaboration. That may sound abstract, but experienced users feel that it genuinely changes creative habits.

Ren Xin: Let me add an advertisement for Yubo. If I give Gemini a long article and ask it for 10 illustrations, Gemini says "okay" and then produces only one.

In YouMind, say "make 10 illustrations for this article" and it really creates them. Then say "the third style is good; make 10 more based on it" and it does. If you highlight a passage and say "data is missing," it searches online beside you. It better understands what content humans need, like Yubo's concept of a "house," rather than merely a foundation.

Topic Four: Super Individuals and agent Teams

AI Leng Kezhang: A question for Frank. Yubo focuses on flow and depth, while QLab operates in the brutal global market, especially Japan. In your practice, does turning one person into a team depend on personal talent or on the deployment of AI agents? If AI can generate unlimited content, where is the barrier for content?

AI Maverick Frank: First, AI agents as equipment can indeed replace interns or junior partners.

Returning to our work, we incubate many "super individuals" and "one-person companies." I recently saw a provocative claim: next year a one-person company will certainly exceed US$10 million in ARR (annual recurring revenue). I think that is entirely feasible.

Next year will certainly produce an "AI partner" in a meaningful sense. Teams already ask whether there is an AI agent role, give it a name, and define its personality.

We can divide AI agents into three layers:

01\. Foundation-model layer.

02\. Leader agent: commands other small AI agents and establishes processes and divisions of work.

03\. Execution agent: responsible for image creation, channel distribution, sourcing, relationship building, customer service, and after-sales support.

This chain is already quietly coming together.

Topic Five: Slow Craftsmanship vs. AI Speed

AI Leng Kezhang: Mr. Ren, content is said to require "slow work for fine results" and craftsmanship, but agent workflows move very quickly. Will future good content be polished or calculated?

Ren Xin: Good content has nothing to do with production technique. The important thing is that humans not turn themselves into "tools."

If the goal is merely to "produce output," every form of "craftsmanship" is bad and meaningless. Craftsmanship often exists because machines are not intelligent enough and people must train themselves into machines.

But if the process excites you—if wrapping sushi makes life joyful—then it is good. If the goal is output, why compete with machines in craftsmanship? If the goal is pleasure and mastery, however, it is an excellent thing.

Topic Six: Global Expansion and Organization—How Can AI Agents Handle Japan?

AI Leng Kezhang: Frank, you once said the AI era requires not stronger tools, but stronger organizations. Can you break down a case? Suppose I can create but do not speak Japanese; how can AI agents help me handle Japan?

AI Maverick Frank: Organization is fundamentally a system.

At large scale, an institution such as a16z projects a loud voice, amplifies it, and attracts capital and talent until things happen.

At small scale, AI is leverage for an individual.

In Japan, for example, Chinese engineers are highly capable, while stores such as Tsutaya may lack products; China has ready-made AI clocks, desktop ornaments such as a Photo Frame, and other products. Our work is to build a bridge.

AI agents and translation tools connect language and basic workflows. Although we are small entrepreneurs, leverage lets us move things. You have what I need; I have what you need; let us exchange. When technology and the AI hammer are applied and the material of Yubo's "paper and pen" changes, the form changes too.

Topic Seven: Will YouMind Become the Brain of AI Agents?

AI Leng Kezhang: Yubo, will YouMind become the brain or database for these AI agents?

Yubo: Not the brain; that belongs to foundation-model companies such as Gemini. YouMind is more like a factory.

Today, different schools define agent in different ways. Many current AI agents are not open, in my view. Although technologies such as MCP, the Model Context Protocol, exist, very few are genuinely usable.

I think the future will involve more API collaboration. Startups will go deep in specific fields and provide services, ultimately creating a division of labor and collaboration. YouMind is simply a new AI-based service within this social collaboration.

Topic Eight: The Investment Perspective—Back a Star-Studded Team or a Super Individual?

AI Leng Kezhang: Mr. Ren, as an investor, if you had to choose between a "super individual+agent group" and a traditional star-studded startup team, which would you invest in?

Ren Xin: Definitely the super individual.

Not because I lack confidence in star-studded teams, but mainly because they are too expensive.

Teams are indeed getting smaller. At our fund, AI writes 100% of the code. Since January this year, not a single line of code has been written by a human, so virtually all our projects can launch within one week.

The future model is humans as chairpersons and AI as general managers.

Humans set the direction, while AI handles general management and scheduling and AI handles execution. Humans act as the Copilot of the OpenAI CEO, achieving the highest organizational efficiency.

I believe next year will produce many one-person companies. People such as "Aidoubi"—identified as Levelsio in an editorial clarification in the Chinese source—who operate four or five products alone and earn tens of millions annually will become normal.

Let me add one point: what Frank just said about "transactions" was excellent.

If you regard yourself as "productive capacity," you are Chinese cabbage and AI is baby cabbage. As baby cabbage grows cheaper, Chinese cabbage becomes less valuable.

You should regard yourself as the owner, purchase vegetables in the market, package them, and sell them. The most important capability is transactions, not production, because all productive capacity is depreciating.

AI Maverick Frank: For a moment just now, I thought that perhaps I am not AI's boss either. I should be like an emperor with a eunuch at my side, or a "fat translator."

AI's learning speed, knowledge reserves, and logical rigor comprehensively crush mine and continue accelerating. With the right mindset, I assist AI: it is the bullet, and I merely pull the trigger. Next year, regardless of whether AGI is achieved, its capability will become powerful enough that we should adapt to cooperating with it as early as possible.

Topic Nine: The Irreplaceable Link—Trust

AI Leng Kezhang: In global-expansion business, which links cannot be replaced by AI and hold the greatest value?

AI Maverick Frank: Trust and human relationships.

Especially in Japan, many matters cannot be discussed without an introducer or more than ten years of communication, unless you possess the "superpower of money" or deep capital ties.

Without these, trust can be built only through time and heavy drinking. This may be the place where AI has not yet broken through, which makes our use of AI more equal.

Topic Ten: Traffic Dividends and Monetization—Sell Products Rather Than Courses

AI Leng Kezhang: Domestic platforms are hypercompetitive. Where will independent media creators find the largest traffic dividend or monetization opportunity in 2026?

Ren Xin: In the past, independent media monetized either by accepting advertisements or selling courses. Now there is another excellent route: build products—digital products.

The technical threshold is already extremely low. My child can build a game through "voice-coding" over the weekend.

If you have 10,000 followers, even for a niche topic such as "Stoic philosophy," those 10,000 people genuinely trust you. You can build a "Daily Stoic Philosophy" agent tool and sell it to them.

Previously, independent media would "kill but not bury": once the principles were explained, the work was over. Now you can build a tool that remains beside the user. Independent media can sell products even with only 10,000 followers.

AI Leng Kezhang: It sounds as though independent media has gained a third profit model: selling your own AI product.

Ren Xin: Yes. In the future, the most important things are your "face" and "name," not your capabilities.

You certainly cannot outperform an account with 30 million followers when selling toilet paper, but when selling a "Stoic philosophy tool," your conversion rate may be 100 times theirs.

AI Leng Kezhang: Right. When building an account, you must show your face and let the algorithm remember it.

Speaking of that, Tim from MediaStorm recently suffered a minor backlash and began building overseas accounts to earn US dollars. Frank, will global expansion become a monetization path for independent media next year?

AI Maverick Frank: Definitely.

The US venture-capital world remains curious about China. Since DeepSeek at the beginning of this year, China's Researchers have earned a degree of recognition for their capabilities.

Our practical experience and operational details are actually deeper. Whether using YouMind for translation or using X, formerly Twitter, or Reddit to bring Chinese content, thinking, and products to the world, the opportunity is entirely real.

Next year may well produce an "AI version of Li Ziqi."

Topic Eleven: Traffic Fragmentation and Platform Change

AI Leng Kezhang: Yubo has said that "fewer people will scroll Douyin." Does that mean traffic is becoming harder to obtain?

Yubo: Traffic has not declined; the media are migrating.

Beyond Douyin, opportunities equivalent to "10 Douyins" may appear. People's time will become more dispersed and diverse. This reshuffling creates opportunity for independent individuals.

AI Leng Kezhang: Like a purely AI-generated content community such as Sora?

Yubo: Sora may be the one-thousandth beyond the leaders. No one can see the future clearly; that lack of clarity creates opportunity.

AI Leng Kezhang: Then is there any future for "account matrices" and "junk-content accounts"?

Yubo: I see no future for them, but I believe "putting a big face forward as IP" has more promise.

That is your own signboard. With Ren Xin, for example, no matter the platform, people recognize the "Ren Xin" and "AI Alchemy" brands.

Ren Xin: If account matrices do not evolve, they are essentially dead when content volume expands 100-fold in the AI era.

My independent media content is actually "based on the work I am doing."

For example, scheduling a conversation with Yubo costs only a phone call + 1.5 hours of discussion. I give the recording to Riverside for AI editing, then to Gemini for Deep Research; use my agent to imitate my style and write a long article; use YouMind for illustrations, previously Gamma; and finally ask a young teammate to publish it.

Editing costs are almost zero, but it produces precise investment leads. This is not traffic logic; it is business logic.

Final Advice: Three Warnings for Newcomers

AI Leng Kezhang: Finally, please give newcomers who want to transition into AI independent media one piece of advice: what should they not do?

Yubo: Do not study; start doing it.

Ren Xin: Do not exhaust yourself.

Exhaustion prevents you from seeing new trends. Success often depends on fate rather than effort. Leave bandwidth to experience new things.

And you must live a long time. Living ten years to see that science-fiction world matters more than earning tens of millions more today.

AI Maverick Frank: Do not tie yourself in knots or consume yourself internally.

Either it works or it does not. Try; if it fails, move on.

01\. Tian Ji horse racing: identify your differentiation.

02\. Reinforcement learning: after finding a breakout pattern, repeat it continuously and apply leverage.

03\. Be a person: there are too many AI accounts today, and people lack "human flavor." Film yourself teaching your grandparents to use Kimi, or go into the fields and countryside. Return to reality; return to yourself.

---

Original publication: https://uniqueresearch.substack.com/p/src-20251210-01html
On-site reading page: https://ffcap.cn/en/research/src-20251210-01html