A Conversation with the Founder/CEO of Soy Sauce Culture | Inside a Thousand-Person "AI Manga Drama" Factory: Wild Growth and Content Ideals

All in AI manhua dramas: team breaks 1,000 people, revenue breaks 50 million.

👦🏻 Podcast interview: Koji

🥷 Edited by: Crossing

🧑‍🎨 Layout: NCon

🚥 This week's guest on Crossing is Jiangyou. In the chaotic early days of "AI manhua drama" — a brand-new species of content — he was one of the fastest movers.

Jiangyou graduated from a technical secondary school, made his first bucket of gold from web novels, founded "Jiangyou Culture," and achieved financial freedom. In September this year, after deciding to go all in on AI manhua dramas, his company grew explosively in just a few months: the team expanded to 1,000 people, revenue broke 50 million, and among the 22 manhua dramas that surpassed 100 million views on Douyin, five came from Jiangyou.

In this episode, Jiangyou breaks down the operational details of this thousand-person "AI factory," and shares his views on the investment value of AI tools, the "electronic pre-made food" controversy, the new job category of "card pullers," visual democratization, the strengths and weaknesses of large models, moats in the content industry, and "becoming the Pixar of the AI era."

Finally, Jiangyou shares his own remarkable story: as a "counter-current" in his generation, relying on instinct for content and extreme execution, he seized an opportunity for ordinary people to reverse their fortunes in the wave of technological democratization that AI has brought.

A fascinating conversation!

Listen on WeChat:

Listen on Xiaoyuzhou:

Since the full interview is quite long (13,810 words), here's the table of contents:

🟢 Lightning Round

Age, alma mater, zodiac sign, one-sentence description of what you're doing now, fundraising status, revenue and profit, team size, pre-entrepreneurship experience

🟢 The Thousand-Person "AI Manhua Drama" Factory

"He can never steal what's in my head, and I can never steal what's in his."

• Why did AI manhua dramas suddenly blow up?
• What does a complete AI manhua drama assembly line look like?
• In a thousand-person team, which roles have the most headcount? Which are hardest to hire for?
• New job categories emerge: "AI artists" and "AI card pullers"
• Even with a thousand-person team, script and storyboarding remain "soul" work that must be hand-polished by humans.

🟢 When the Whole World Is Resisting AI

"You can tell it's AI at a glance, too much AI flavor... They seem to use 'AI' as a synonym for shoddy work."

• Finding people willing to "swim against the current" and pioneer a new visual era.
• "Too much AI flavor" became shorthand for shoddy work — but those actually working with AI know: it improves every single day.
• A brutal fact: AI output has already surpassed that of 80% of animators with ten years of experience.
• Hayao Miyazaki said "lines drawn by computers have no soul," yet 90% of global animation is already computer-made.
• "My partner was making 300,000 a year a year ago; next month he's picking up his Rolls-Royce."

🟢 Content Methodology: Three Layers Behind the Hits

"Make a good work centered on life and death, human existence, society — and if it flops, so be it."

• Layer one: power-fantasy logic, giving 18-year-old boys an outlet to decompress.
• Layer two: visual spectacle, using AI to create god-demon war scenes on the scale of Creation of the Gods.
• Layer three: IP-worthy content, making a Coco that resonates with the whole world.
• Why "Chinese film and TV going global" is a false proposition: the culture is too deep, nobody understands it.
• AI's greatest value: giving you unlimited runway to experiment — 200,000 per production, no sweat if it flops.

🟢 Costs, Models, Good Content

"We don't want to make a quick buck on this trend, then wait for the next one. We want to keep building, keep pioneering at this trend."

• Industry average: 1,000 yuan/minute; we charge 1,500 — because others generate 40 images, we generate 160+.
• Cost structure breakdown: labor 60%, compute 20%, script 15%, voice acting 5%.
• Our most-used video model: Shengshu
• Keling AI, Dreamina, Sora... each has its strengths; we integrate the best of all.
• Why we don't cut costs: good things endure; cost-cutters chasing quick profits eventually get eliminated.

🟢 From Technical Secondary School to Financial Freedom

"You'll never get rich working for someone else. People like us probably only get one shot in a lifetime."

• No computer? Spend all night at internet cafes writing.
• First month writing web novels: five-figure income. Six months in: seven-figure annual income. Then opened China's first offline web novel incubator.
• Short dramas: 150,000 cost, 7 million revenue in one month, 50 million pre-tax annual income — financial freedom.
• Why pivot from short dramas to all in on AI manhua dramas?
• 1% talent matters more than 99% effort. Effort only makes you money; it doesn't make you successful.

🟢 Managing a Thousand People

"Who's actually happy working? But I need them to compare this place to others and think, this company is pretty decent. That's enough."

• This isn't talent management, it's factory management — staff, team leads, supervisors, tasks distributed downward, clock out when done.
• Regular staff never work overtime, only six actual hours a day; but team leads and above work 365 days a year.
• Why? People earning 5,000 a month have a right to their nightlife; high earners owe everything to the company.
• Biggest risk? Doesn't exist. Content drives production; as long as the content is solid, whoever produces it is the same.
• "There are no peers in this track" — are the companies behind Crayon Shin-chan and Detective Conan enemies?

🟢 Becoming the Pixar of the AI Era

"I want to go public. I want to become the world's AI application concept stock."

• If you're making this much money, why raise funding?
• Next year's plans: disrupt the domestic anime market, go global, build IP ecosystems, make games...
• "We're going to disrupt the entire domestic anime market, stir things up."
• We're working on a "top-secret project," announcing after New Year, that will send massive shockwaves through the entire AI market.
• "(Do we use AI tools?) No. Not a single one."
• "(Do AI tool startups have investment value?) None whatsoever. Zero value."
• Bold predictions for 2026?

👦🏻 Koji

This week's guest is one of the fastest movers in the chaotic early days of AI manhua drama, a brand-new species of content. He's from Nanchang, graduated from a technical secondary school, and first entered the content world through web novels. In September this year, he decided to go all in on AI manhua dramas. In less than three months, his team rapidly grew to 1,000 people, revenue broke 50 million, and among the 22 manhua dramas that surpassed 100 million views on Douyin, five came from Jiangyou. Hello Jiangyou, welcome to Crossing.

🧑🏻‍💻 Jiangyou

Hello Koji, thanks for having me.

Lightning Round

👦🏻 Koji

Our show has a tradition of starting with a lightning round to help everyone get to know you quickly. First, your age?

🧑🏻‍💻 Jiangyou

👦🏻 Koji

Alma mater?

🧑🏻‍💻 Jiangyou

Jiangxi Electronic Engineering Institute.

👦🏻 Koji

Your MBTI and zodiac sign?

🧑🏻‍💻 Jiangyou

I don't know my MBTI, but my zodiac sign is Aquarius.

👦🏻 Koji

Can you describe what you're doing now in one sentence?

🧑🏻‍💻 Jiangyou

Disrupting a visual era.

👦🏻 Koji

Wow, that's a cool line. Has the company raised funding?

🧑🏻‍💻 Jiangyou

Currently fundraising.

👦🏻 Koji

Can you share your current revenue and profit situation?

🧑🏻‍💻 Jiangyou

If we're talking annual, probably around 1 billion in revenue, with estimated profit of 200 to 300 million.

The Thousand-Person "AI Manhua Drama" Factory

👦🏻 Koji

Current team size?

🧑🏻‍💻 Jiangyou

1,000 people, expected to be 3,000 to 5,000 next year.

👦🏻 Koji

One more sentence: what did you do before becoming an entrepreneur?

🧑🏻‍💻 Jiangyou

Started with web novels, then short dramas, then pivoted to manhua dramas.

👦🏻 Koji

Let's go back to the birth scene of manhua dramas. It's been a period of explosive growth — monthly output from 10 to 100 productions, team from dozens to thousands. Last time you also mentioned daily revenue going from 10 million to 30 million. Jiangyou, can you explain in one sentence why AI manhua dramas suddenly exploded recently?

🧑🏻‍💻 Jiangyou

Mainly because short dramas arrived, getting all users hooked on this fast-paced format.

👦🏻 Koji

Any other changes on the platform or technology side that combined forces?

🧑🏻‍💻 Jiangyou

AIGC capabilities have made anime production incredibly fast. It used to take forever to get a new manga chapter; now you can watch dozens or even hundreds of episodes in a day. That gives viewers massive amounts of time to explore more content, unlike before when you'd wait a whole week for a single episode and just didn't pay that much attention to any one work.

👦🏻 Koji

With the recent explosion of manhua dramas, which AI breakthrough do you think is directly responsible?

🧑🏻‍💻 Jiangyou

I think it's mainly on the visual quality side. With AI-generated video now, you genuinely can't tell it's AI anymore — that's the critical point.

👦🏻 Koji

Inside your studio, what's the full workflow for producing an AI manhua drama from start to finish?

🧑🏻‍💻 Jiangyou

First we select a topic. Whatever we're making, manhua drama or anything else, we always start with topic selection.

After that comes content writing — a complete script — which then goes to the storyboard artist.

The storyboard artist breaks down all the shots, and after that it goes to the image generation lead, who distributes to various supervisors and team leads.

The team leads then assign to individual AI illustrators, who set the visual tone before handing off to the gacha specialists. They start generating corresponding video from the text prompts, and after everything's generated, it gets assembled and passed back up through team lead to supervisor, then to post-production for editing. After editing comes voice acting, then it goes live.

👦🏻 Koji

What role does AI play at each stage?

🧑🏻‍💻 Jiangyou

AI doesn't play a huge role in early-stage scriptwriting, because we feel text has warmth — AI-generated stuff is too formulaic.

The core moment is when the script reaches the gacha specialists. That's where AI becomes enormously powerful, generating character designs, story backgrounds — everything gets solved by AI.

Then at editing, AI doesn't do much again. AI only participates in image generation and video generation. For all post-production editing and voice acting, AI can't contribute much.

👦🏻 Koji

Among your 1,000 people, which role has the most headcount right now?

🧑🏻‍💻 Jiangyou

Post-production, by far. Because our output volume is so large — over 100 premium manhua dramas per month. Post-production can't be replaced by AI, it has to be done manually.

One person can only edit about one episode per day, roughly one to two minutes. That's their daily capacity. So in our case, one drama is about 60 to 80 episodes, that's 6,000 to 8,000 episodes per month, which means post-production editors are our largest team.

Because on the gacha side, one person can generate 10 minutes of raw footage per day, but it takes 10 people a full day to edit that into finished episodes.

👦🏻 Koji

For post-production people, are the requirements similar to traditional film and TV?

🧑🏻‍💻 Jiangyou

Almost the same. But definitely having some internet sensibility and understanding of anime helps, because anime's expressive style differs somewhat from traditional short dramas and film works.

👦🏻 Koji

Which role do you think plays the decisive, critical factor in content quality?

🧑🏻‍💻 Jiangyou

I'd say scriptwriting. Good content, even if produced somewhat poorly, will still find its audience. But bad content, no matter how beautiful, nobody pays for.

👦🏻 Koji

So even though it's called AI manhua drama, whether it becomes a massive hit still fundamentally depends on the human element in script creation.

Then which roles get paid the most?

🧑🏻‍💻 Jiangyou

Excluding our founders, three roles are paid roughly the same: scriptwriters, R&D, and technical.

👦🏻 Koji

R&D and technical are different?

🧑🏻‍💻 Jiangyou

Our R&D isn't about developing technology — it's about developing different visual styles. Photorealistic, 3D anime, 2D anime, more Japanese-influenced, more American-influenced, or fantasy, or Western. We research different visual styles because our understanding is that different styles enable different content. For romance novels, you wouldn't use 3D — you'd use something like Makoto Shinkai's style.

👦🏻 Koji

Within your 1,000-person team, what's the scale of technical staff? Do they build internal systems or workflows for everyone?

🧑🏻‍💻 Jiangyou

Our technical staff mainly help us build workflows and tools. Everything technical does is aimed at improving per-person efficiency across our company.

👦🏻 Koji

We can expand on technical later. But before that, I want to go back to something you mentioned earlier — some work still absolutely requires humans, specifically scriptwriting and storyboarding. Why is that?

🧑🏻‍💻 Jiangyou

The biggest factor is that AI-generated content is relatively fixed, very formulaic. It might just adapt a novel directly, pulling out everything that happens.

But we understand clearly that as a storytelling medium, manhua drama can't just film everything from a novel. It becomes incredibly drawn-out. Novels run millions of words — at minimum 1 million, often 3 million, 5 million.

Our manhua dramas are 120 minutes, about 60,000 words. We're condensing 2 million words into 60,000. What happens with AI in this situation? It pulls everything from those 2 million words, unable to distinguish primary from secondary.

But humans can. For example, recognizing when the author is padding, we cut that and keep the exciting parts.

As a text medium, novels let audiences imagine. But manhua drama is video directly in front of you — no imaginative space.

For example, "a bright moon" — four words, but the moon in your mind differs from mine. Once made into manhua drama, that moon is right there, identical for both of us. We try to make it something broadly acceptable. Novels are "a thousand readers, a thousand faces," but anime is different — a thousand viewers, one image. So we believe AI can't replace scriptwriters.

But it can assist you in quickly consuming the book. Normally a scriptwriter adapting a novel would need to read 2 million words. Now AI can extract an outline, organizing all the story elements, letting you essentially finish that 2 million-word book in 30 minutes. Then you can think about how to adapt it — a massive efficiency boost for our writers.

👦🏻 Koji

When you mentioned the roles earlier, one seemed entirely new — AI illustrator. This seems like a completely new position?

🧑🏻‍💻 Jiangyou

The so-called AI illustrator isn't the gacha specialist. The AI illustrator's core function is: when a drama comes in, who defines the protagonist, who defines supporting characters, who defines the settings. Because all the gacha specialists don't know. So before we produce a drama, we have AI illustrators lock everything down — all male leads, female leads, supporting cast, including environments, fixed and finalized. That's the AI illustrator, responsible for generating these through AI methods, then handing off to all the gacha specialists below.

👦🏻 Koji

Sounds like the AI illustrator defines the key asset package. Could you also explain the other role you mentioned, AI gacha specialist — what do they typically do?

🧑🏻‍💻 Jiangyou

They follow the text to generate all effects, including scene dialogue and everything else.

👦🏻 Koji

What skills does someone need to be a good gacha specialist?

🧑🏻‍💻 Jiangyou

This is the simplest — no barrier to entry. You just need to understand text and know how to operate the AI tool.

👦🏻 Koji

So the hardest parts are actually the creative work up front — topic selection, scriptwriting, storyboarding — and AI illustrator is relatively harder too, but gacha is probably the easiest.

What's the hardest role to hire for across the entire chain?

🧑🏻‍💻 Jiangyou

First, scriptwriters — hardest to find. Second, R&D is hard to recruit. Gacha specialists, AI illustrators, post-production editors — these are common roles in daily life, easy to find.

Our philosophy is: top talent doesn't circulate on the market. We've always believed this. When you encounter real talent, they're definitely not on 58.com — that's not talent, either someone eliminated or just an ordinary job-seeker. Because talent is always employed elsewhere; you have to figure out how to get them to come to you.

👦🏻 Koji

So how do you recruit such people?

🧑🏻‍💻 Jiangyou

I'm very direct. First, money. I think to get someone from a familiar environment to an unfamiliar one, you use money to move them, not pie-in-the-sky promises.

Second, shared vision. If someone is purely about money, they have many options. I remember reading a story where a company invited Coca-Cola's chairman to join, asking if he wanted to change the world.

Our logic: do you want to join our company in pioneering a new visual era? If you have this idea, let's work together. If you don't, if you just want to make money, we won't place you in a senior position.

Because we're a group swimming against the current of our times. Right now there's still widespread resistance to AI — from audiences, markets, platforms. Like when hand-drawn animation transitioned to digital — how many people resisted? But we all know how that ended.

We don't control the final outcome, but in this process we're definitely going against the current. I need to bring people willing to swim against that current with me, to pioneer or lead a new visual era. There needs to be a shared goal. Without that, it's meaningless to me. Because what we want to do isn't simply about making money.

When the Whole World Is Resisting AI

👦🏻 Koji

Can you think of a specific detail or story that suddenly made you feel "I really am swimming against the current"?

🧑🏻‍💻 Jiangyou

Why do we feel like we're going against the current?

First, the market feedback on AI currently — many people give you negative reviews. Not cursing you, but saying "obviously AI," "too much AI flavor." They seem to treat AI as shorthand for shoddy work, assuming AI-generated content must be low-quality.

But those of us working in it know AI's terror lies in its daily progress — what you saw yesterday differs from today. We also clearly know AI output has already surpassed 80% of elite practitioners.

A traditional animator, ten years in, already considered highly skilled — their current work may not match AI's capabilities. That's just reality. But because people are unfamiliar with this, it's like pre-made meals — using AI feels like you're using pre-made meals.

👦🏻 Koji

What do you think of the "pre-made meal" comparison?

🧑🏻‍💻 Jiangyou

Personally I don't think they're the same thing.

Pre-made meals definitely have downsides for your health, but AI-generated animation doesn't harm the human body. For example, if you're making top-tier work, go ahead and do your thing. We're here to fill market gaps — below the S-tier premium stuff, there's the A-tier, B-tier, C-tier market that AI can completely cover, and do it better than traditional production.

I watched an interview with Hayao Miyazaki recently, and afterward I realized what the problem is.

It's a clash between two eras. Miyazaki still draws lines by hand, with a pen. His argument is that computers have no warmth, that hand-drawn lines have soul. But over 90% of the global animation industry today uses computers. So we've already gone from the hand-drawn era to the computer era, and now from the computer era to the AI era. This transition is bound to be difficult.

That's why we call ourselves the Countercurrent — because this transition period is when you're hardest to understand.

But I believe next year and beyond, AI will become the absolute mainstream. Maybe an entire era will be powered by AI. In 10, 20, 30, 40, 50 years, over 50% of all visual content you see could be AI-generated. We're going to live in an AI-generated world. It's unavoidable.

But I don't think that's a bad thing. Is computer-generated content really so different from AI-generated content? In the end, both satisfy the audience's visual appetite. You can't hear "AI" today and immediately assume there's something wrong with it, or approach it with bias, worrying about what harm it might cause.

It won't. When the computer era arrived, everyone thought video games would destroy a generation. Was that generation destroyed? No.

So AI won't destroy anyone either. It will only help everyone, and it will give every young person a chance — whether in money, wealth, or status. The next breakthrough opportunity for young people is right here in AI.

👦🏻 Koji

Go back 100 years — when the camera first came out, there was plenty of backlash too. People felt photography was a desecration of painting.

But today, most of the images we see in daily life are photographed, not painted. They satisfy our visual needs just the same, and we've embraced photography fully as an art form. The most expensive photographs still sell for astronomical prices.

You mentioned AI can help people achieve this kind of turnaround — in money, wealth, status. Do you have people around you who've actually done this? Any stories?

🧑🏻‍💻 Jiangyou

One of our partners is a textbook example.

I was already financially free before I got into AI anime series — I was already set from short dramas and web novels.

One of my partners was making less than 300,000 RMB a year. Next month he's picking up a Rolls-Royce. That's one year. It was a massive transformation for him.

👦🏻 Koji

For anyone, really.

🧑🏻‍💻 Jiangyou

You know that saying online — if you weren't born with a Rolls-Royce, you'll basically never have one in this lifetime. He made that leap.

👦🏻 Koji

From his previous income level to picking up a Rolls-Royce in a year — do you feel the impact this kind of dramatic change has on a person, or on the industry?

🧑🏻‍💻 Jiangyou

I think money is just a step along the way, because we still have ideals. We genuinely want to change an era through AI.

Not the Steve Jobs "change the whole world" kind of thing — but change a visual era. Once you've satisfied your desire for money, what you want more is recognition.

We're benchmarking ourselves against Pixar, built with AI. We don't want our work to just be popular in China — we want people around the world to cheer and applaud our work. It's not just about the money.

Content Methodology: Three Layers Behind the Hits

👦🏻 Koji

Pixar made Toy Story. You've also produced numerous mega-hits with hundreds of millions of views. Do you think there's something these successful pieces share in common?

🧑🏻‍💻 Jiangyou

We approach content on three levels.

First is meeting commercial market demand — straightforward, satisfying, made for specific audiences. For example, content for 18-year-old guys to decompress from life's pressures. That's one approach.

Second, we do technical showcase pieces — relatively heavy on flexing technique. There's a domestic film called Creation of the Gods — tons of special effects, a story everyone already knows, but we blow up the scale and hit you with visuals. Like Avatar — no real story substance, just the world's first 3D film. That's another angle for us: using AI to create large-scale stories. We're planning something called Battle of the Gods, using AI to render true divine-demon warfare at massive scale, top-tier spectacle. Think of it as a visual feast.

Third is the IP-driven content we want to make, something like Coco. Honestly, I think Chinese film and short drama "going global" is a false proposition — it's hard to export because the culture runs too deep. Who's going to engage with your culture? Like American culture — we don't really get it either. But American superhero films, The Avengers, Titanic, Catch Me If You Can — simple, no heavy cultural baggage, and we all love them because they're straightforward stories.

But Chinese film and TV, because it has to serve the domestic market, ends up carrying too much of its own culture.

What we want to do is build around universal stories — around life and death, human existence, society — and make a great piece. That's our third direction, the IP route: making a series for that sense of resonance.

Even if it flops, it doesn't matter. The thing AIGC gives us is infinite room to trial and error.

Today, developing a film costs tens of millions or over 100 million RMB minimum. Now I can make one for 200,000 or 300,000 RMB. A hundred pieces costs a few tens of millions.

And it's not like all of that goes down the drain. With low-cost trial and error, I can endlessly produce what I want to make. As long as one topic hits, you get all your money back and then some.

Cost, Models, Good Content

👦🏻 Koji

You mentioned a piece costs around 200,000 to 300,000 RMB to produce. Can you break down the average cost per piece? Where exactly does the money go?

🧑🏻‍💻 Jiangyou

The current market generally runs about 1,000 RMB per minute on average.

Why are we more expensive? Because most people rush to get in quickly and produce crude work — that's a problem too. They don't care about final presentation, they just want to enter the赛道 fast and trial-and-error cheaply. That creates the impression that "AI output is garbage" — it's these people who created that perception.

Our average cost is about 1,500 RMB per minute. We're noticeably pricier because while other companies might generate 40 images and call it done, we require at least 160+. Our compute and labor costs run higher because we demand TOP 1 image quality in the entire industry — number one in visuals, number one in visual effects. In this赛道, I want to be first.

We're not comparing ourselves to domestic animation or global studios — just others in our赛道. So our prices run higher, and that's a quality requirement. We believe good things endure.

Cutting corners to save costs might earn quick money now, but you'll get eliminated from the赛道 immediately. The more you invest, the higher your returns — not the other way around. Low investment with high returns is just a temporary or occasional exception.

Where do our costs mainly go? Labor takes the biggest share, followed by compute, voice acting, script, and editing. Labor is over 60%, compute about 20%, voice acting under 5%, script around 15%.

👦🏻 Koji

What AI tools or models does your current production system use?

🧑🏻‍💻 Jiangyou

We mainly use Shengshu Technology the most, plus Keling AI, Dreamina, Nano Banana, and Sora — we're always using all of them.

👦🏻 Koji

Didn't expect Shengshu to be your most-used — why is that?

🧑🏻‍💻 Jiangyou

Because I like Shengshu's art style. That's it.

👦🏻 Koji

You have an internal tech team. What tools are they developing for internal use?

🧑🏻‍💻 Jiangyou

Our internal tools are mainly focused on efficiency gains. Nothing particularly special — every company has different workflows and efficiency needs. These tools are basically custom-built productivity boosters for our specific environment.

👦🏻 Koji

You mentioned using Shengshu the most — why use other models simultaneously?

🧑🏻‍💻 Jiangyou

Every AI tool has its strengths.

For example, when we're doing big scenes, we use Keling AI more because it handles large-scale detail better.

Dreamina has strong imaginative image generation.

Each has its advantages, and I consolidate all of them rather than locking into one and ignoring the rest. It's about bringing the best together.

👦🏻 Koji

So do you care about each provider's cost?

🧑🏻‍💻 Jiangyou

Cost is certainly core to a company, but I think if saving on cost degrades quality, it's not worth it. The extra profit doesn't mean much because we're in this for the long haul.

We don't want to finish this year, make a windfall when the赛道 is hot, then wait around for the next风口.

We want to keep building in this风口, keep pioneering. Today it's AI anime series, tomorrow AI film, the day after AI animation.

I'll keep doing more through the logic of AI, rather than finishing AI anime series and showing up at the next赛道风口 having already pivoted to games.

I don't want that. I want to walk this赛道 all the way to the end.

From Vocational School to Financial Freedom

👦🏻 Koji

You've done many other types of work before — web novel writer, short dramas. Now that you've found the AI anime series赛道, do you see it as a long-term endeavor?

🧑🏻‍💻 Jiangyou

Although I've done short dramas and web novels, I've always been in the content business.

The reason we want to keep doing AI is that AI is the only tool that can democratize the visualization of all text.

All my content needs to be visualized, and with AI, I definitely want to take it all the way. Only AI can serve as the carrier for all our content — whether it's film, animation, anime, short dramas, or some new medium in the future, whether the story is told through video or visuals, AI can handle it. If you keep doing well in this space, I think the future is bright.

👦🏻 Koji

After you graduated from vocational school, you started writing web novels. You didn't have the halo of a prestigious university, nor did you have the kind of elite connections people typically talk about. Yet you've achieved what you have today — and you mentioned you were already financially free before AI anime series. What key things do you think you did right to get here?

🧑🏻‍💻 Jiangyou

The most correct thing was that in the era when I had nothing, I kept telling myself one thing: You only live once. If you don't try, if you don't hustle, then this life is over.

When I was very young, I understood one point: Working a job will never make you rich. When opportunity comes, grab the wave.

Writing web novels was hard too, because my family was ordinary — my parents only earned a few thousand yuan a month.

When we first encountered the web novel industry and saw others earning tens of thousands, hundreds of thousands, even a million, I felt this was the only opportunity. At that time I didn't have a computer, so I went to internet cafes every day to write. I'd wake up and go to the cafe, staying until early the next morning, because an overnight package only cost about a dozen yuan.

I'd write there all day, and when I finished, I'd publish it. My understanding at the time was that I was someone who could endure hardship.

I used to love playing games. But once I discovered web novels and started writing, I stopped gaming entirely. I set a goal for myself: no games until I succeeded.

That year, as a gaming-addicted teenager, I played games for less than 10 hours total in an entire year. I just forced myself to write. If a novel flopped, I'd switch to another one. Flopped again, switch again. At my peak, I could write 50,000 words a day, entirely by hand, without sleeping for 24 hours.

Why? Because I knew if I missed this opportunity, it would be gone forever. As you just said, I have a vocational school degree, no parental support, no one around me to lift me up. When we find an opportunity that's ours, that we can actually see, if we don't go all out ourselves, that opportunity disappears.

I believe ordinary people might only get one chance in their lifetime, so I had to seize it. Even if it wasn't some so-called hundred-billion or trillion-dollar market — that had nothing to do with me. I just needed to change my own life.

When we started writing novels, I earned over 10,000 yuan in my first month. Within half a year, I was making a million a year. Later I opened the country's first offline web novel incubation class. I thought, if I can write, can I teach others to write? I invited many people who wanted to write, taught them, they'd write and give it to me, and I'd submit to platforms. We'd split the earnings. I built a company like that.

I worked in that赛道 for 8 years. At that point, I hadn't fully achieved financial freedom yet.

Then came short dramas. As everyone knows, short dramas are a暴利 industry. When we first started, our highest-cost production was 150,000 yuan, and we earned over 7 million in a month. Then we made several more. I'm picky about content, so my costs stayed low — two or three hundred thousand to shoot one, and each earned several million. After about a year, we were financially free.

Right after financial freedom came the AI anime series era.

At the time we were kind of funny about it. Everyone thought, well, we've made so much money, investing in this thing isn't expensive. I said you guys try it, I'll pay. A few hundred thousand a month, let's experiment. After trying it, wow, this thing is incredible. So we kept going until around July or August.

The reason we went all in was also because short dramas had become the entertainment industry. Short dramas used to be low-cost, high-return — 200,000 or 300,000 to shoot, earn 5 million or 10 million. Now what happens? No matter how much you spend, you need a massive marketing team to promote it to make money.

That's not what I want. That's not making money from content, that's making money from celebrity效应 and fan circles. So I stopped, and quickly went all in on anime series. Because with anime series, I don't need marketing, I don't need actors, I don't have to please anyone. As long as the content is good and satisfies the audience, that's enough.

👦🏻 Koji

Could you share roughly how much you made in that year doing short dramas?

🧑🏻‍💻 Jiangyou

Not that much, a bit over 50 million.

👦🏻 Koji

That's already a lot.

🧑🏻‍💻 Jiangyou

I have a friend who used to write novels with me. He now runs a well-known short drama company called Shanhai Group. He's also incredibly inspiring. We were both losers, made a few million or tens of millions in the web novel circle. Then he spent 2 million in initial capital and earned several hundred million in a year through short dramas. Conservative estimates put it at 300 to 500 million. It took me over a year of grinding to make 50 million, and that was all from me frantically writing content. That 50 million was pre-tax too — after tax, I only took home a bit over 30 million. So in the short drama赛道, this isn't exaggerated.

👦🏻 Koji

Whether writing web novels or doing short dramas, you worked extremely hard from the start. In the content industry, what ratio would you say effort and so-called talent each account for?

🧑🏻‍💻 Jiangyou

To put it bluntly: without talent, no amount of effort matters. It's a waste of time.

👦🏻 Koji

How do you judge whether you have talent?

🧑🏻‍💻 Jiangyou

It's actually simple. We've been in content for so long — you can basically tell from the first time someone writes whether a story works. If your first attempt is passable, decent, with strengths, if you can identify your strengths — maybe you tell stories well, or you follow trends well, or your writing style is good — as long as you find one strong point, you have talent. If your writing is bad, your story structure is bad, your imagination is limited, and you have none of these three most important strengths, then it's basically very hard to make it in this industry.

Everyone says 99% effort plus 1% talent. Actually that 1% talent is more important than the 99% effort. The more successful you become, the more you realize talent really matters.

What can effort do? You'll earn the appreciation of leaders and bosses, they'll give you things to do, you'll make money. But it won't make you successful. It'll only let you get by in life, moving bosses or the market through your effort and character, and they'll reward you with money.

But if you want to succeed, effort alone won't do it.

Because any successful entrepreneur or successful person in the world didn't rise through effort alone.

For example, no matter how hard I studied, I couldn't get into university. I had no talent for studying. Besides Chinese, where my grades were especially good, I failed everything else — math, English, chemistry, physics, geography. You could say I had a bit of口才, very down-to-earth口才.

Compared to those who stand on stage and clearly explain professional terms, I'm not as good as them. But when it comes to metaphors, especially down-to-earth metaphors, they're not as good as me.

👦🏻 Koji

How would you use a down-to-earth metaphor to describe the AI anime series industry?

🧑🏻‍💻 Jiangyou

It's turning the old landline into a mobile phone.

👦🏻 Koji

Turning the old landline into a mobile phone?

🧑🏻‍💻 Jiangyou

You can think of it like our parents' generation using landlines to make calls. No one could have predicted that when Nokia phones came out, the first company to go bankrupt would be the flashlight company, because Nokia had a flashlight feature. When it came out, everyone thought landlines would disappear, but landlines survived — flashlights disappeared.

Now let me ask you: is a mobile phone a device for making calls? A phone can send WeChat messages, go online, watch videos.

That's the logic of AI. You think AI can only do anime series? No, it can do games, code, PowerPoints, anything you want it to do. That's the difference.

Managing a Thousand People

👦🏻 Koji

Let's return to AI anime series. Your team expanded from a few dozen to a thousand people in just a few months. That must have been a massive management challenge. How did you maintain stability and order during this expansion?

🧑🏻‍💻 Jiangyou

I need to explain something here. When we imagine managing 1,000 people, we might think it's about talent management. But it's more like factory management logic.

Employees, team leaders, supervisors — everyone distributes tasks downward, finish your work and you can leave. So the management cost isn't as high as we imagine. We don't need countless departments like many big companies, with different departments handling different clients.

We only have three departments total: an administrative department, a general staff department, and what we call the senior and middle management department.

It's fundamentally factory logic. Of course, I'm also simultaneously hiring many professional talents to assist me and help manage the company better. But in the early stages, we didn't need them to keep the company running.

👦🏻 Koji

How would you describe the work atmosphere and organizational culture? For example, do people work a lot of overtime?

🧑🏻‍💻 Jiangyou

Actually, not much overtime. Our logic is: all regular employees don't work overtime. Everyone at team leader level and below doesn't work overtime. Only those above team leader work overtime.

👦🏻 Koji

None of the regular employees work overtime?

👻‍💻 Jiangyou

First, their income isn't particularly high — maybe 5,000 yuan a month including social insurance, so 50,000 or 60,000 yuan a year.

In this situation, I don't think we should encroach on their nighttime lives. Let them feel that from the moment they get off work, they're free. Want to go play? Go play.

They have their lives, their friends, their families — not just work. But conversely, why do supervisors and above need to work overtime? Because they earn high salaries. When you're earning a high salary, you don't have so-called friends, family, relatives. At that point, you need to give back to the company. That's our logic.

👦🏻 Koji

During our pre-interview, you also mentioned that having people work only 8 hours a day is because we're doing content, doing creative work, so people's brains can't work too long either. Is that also a reason?

🧑🏻‍💻 Jiangyou

We work 8 hours a day, with 2 hours for lunch break, so actually only 6 hours of work.

Because I came from the bottom. I've done every ordinary job — waiter, mover, delivery driver, real estate agent. You could say I've done all the ordinary work there is. I understand what every person is thinking.

I hope that in my company, they either earn money, or they feel this company is pretty good — I only work a few hours a day and I can leave, this company is quite humane.

I want them to feel good about this company. I don't want them to say they have to be happy and joyful at work — who the hell is happy working? I don't think work is a happy thing. But I want them to feel, compared to other places, that this company is actually pretty decent.

👦🏻 Koji

We're recording this podcast on a Sunday. Before we started, when I saw you guys setting up the equipment and environment, quite a few colleagues were helping out. So the management team doesn't rest on weekends either?

🧑🏻‍💻 Jiangyou

No rest.

👦🏻 Koji

During our pre-interview, you said something I found really interesting. You said, "Our management team doesn't rest all 365 days because we don't know when the wind will stop blowing. When the opportunity's gone, you won't be able to earn this money even if you want to."

🧑🏻‍💻 Jiangyou

Right. We have to maintain this mindset, though it doesn't mean it will necessarily happen. Treat every day like it's your last, rather than thinking, I made money today, so I can take a break.

Building the Pixar of the AI Era

👦🏻 Koji

You've actually already made a lot of money, yet you still treat every day like it's your last chance to earn. How do you maintain this mentality?

🧑🏻‍💻 Jiangyou

For me, it's probably not about making money. Money doesn't really mean anything to me.

Why? Because I don't spend much myself. The money's with my wife. I don't buy watches. My drinks are the 1.5 RMB kind.

I want to seize this opportunity to build a man's dream. Is it possible to go public, or create a completely new AI concept stock?

I want to build a brand, so I keep pushing myself.

But everyone below me, apart from me, just wants to make money.

👦🏻 Koji

A team of 1,000 people is actually quite large. Are you worried about any risks that might emerge with such a big team?

🧑🏻‍💻 Jiangyou

I don't think there are any risks. The core is content driving production, not production driving content.

Once our content is solid, it doesn't matter who does the production. I'm not worried about any so-called risks.

The only risk might be employees fighting — that's something every company in the world worries about, and it's unavoidable. But it's hard for that kind of thing to happen these days.

👦🏻 Koji

What about competition? Earlier we mentioned that out of 22 manhua dramas that broke 100 million, 5 were yours, but 17 came from competitors. How do you view this competition? Are there real differences, or is it actually quite homogeneous?

🧑🏻‍💻 Jiangyou

I want to make one point: I never think of this track as having peers.

👦🏻 Koji

You never think this track has peers?

🧑🏻‍💻 Jiangyou

Right. Because I think everyone is still story-driven. The story I want to tell is different from the story you want to tell, so we're not peers in the same track. Regardless of whether the art style is similar or not, if our stories are different.

For example, are Under One Person and Fox Spirit Matchmaker in the same track? Would their companies be hostile to each other? I don't think so. Do you think the companies behind Crayon Shin-chan and Detective Conan are rivals? Bitter competitors?

👦🏻 Koji

Because the content industry is incredibly diverse.

🧑🏻‍💻 Jiangyou

Right. They can never steal what's in my head, and I can never steal what's in theirs.

Maybe if I earn 1 million, he can only earn 500,000, because the market is only 1.5 million total. But that's fine too, I don't see it as competition. Because good content earns 1 million, bad content earns nothing even without him.

👦🏻 Koji

But in terms of talent competition, or having to buy traffic on platforms, wouldn't you still encounter peers in those areas? How do you view that kind of competition?

🧑🏻‍💻 Jiangyou

On the technology side, I don't think it's a big deal either. Because in the future, technology will be democratized. The AI technology used worldwide will eventually become equal.

It's just that I'm a bit faster. If I'm the first to play this game, I've already cleared level one and I'm on level two, while you might still be in the middle of level one. But you'll eventually reach level two — it's just a matter of time. So I don't care about so-called competition. My people going to you, your people coming to me — it's meaningless, because everyone is doing the same thing, just changing seats to do repetitive work.

As for buying traffic, that's even more meaningless. Because if the content is bad, even if you want to lose money buying traffic, you can't. Douyin won't let you lose money. You can only buy traffic if your content is good; bad content can't buy traffic.

👦🏻 Koji

Earlier you mentioned you're also considering fundraising, but my understanding is that you're actually quite profitable. Why raise money? What's the thinking there?

🧑🏻‍💻 Jiangyou

First, I want to go public. I want to become a global AI application concept stock. Everyone's hyping AI stocks now, but nobody's doing applications. Everyone says selling chips, selling technology, selling GPUs is profitable, but nobody says they're making money — big money — with AI models. I want to become that kind of company.

Second, the reason I'm fundraising isn't because I need money today, but because I want to do so many things in the future. Our plan for next year is to first disrupt the entire domestic animation market, stir things up.

Third, we want to go global. I want to make something truly big, not just earn this hard-earned money today doing things without much substance, things that could never capture the world's attention.

Let's be realistic: the manhua dramas we're making now are to satisfy viewers' need for decompression — you could call them decompression videos. But I want to make something like Coco. But doing that requires massive human resources, material resources, energy, and financial resources.

Our company wants to do more IP ecosystem development, including merch culture, figurines and designer toys, building our own IP. We want to make games through AI — that's also on my list.

I want to do too many things. The ideas are good, but they need to be executed one by one. We want to grow the company bigger and bigger through fundraising and going public. Not just making manhua dramas.

I hope next time we talk, we'll be discussing the global layout of AI, how AI is entering every industry.

👦🏻 Koji

Always believe that something wonderful is about to happen.

Actually, today many people want to rush into AI manhua dramas to make some quick money. What qualities do you think someone needs to first survive in this industry, and second, actually make money?

🧑🏻‍💻 Jiangyou

First, I think they need content. Content is the driver, the engine.

Second, they need technology — you could call it the car's exterior. If the exterior doesn't look good, who will buy your car?

Third is the boss's execution. If the boss is a hesitant person, thinking here and there, afraid to make decisions here and there, that won't work either.

But the core requirements are two things: content plus technology.

👦🏻 Koji

I believe 2026 will be an explosive year for AI manhua dramas. What do you think are the most likely changes to happen that year?

🧑🏻‍💻 Jiangyou

Next year, AI will produce one product after another that颠覆s everyone's imagination.

Including something we're working on that's top secret, extremely secret. We can only announce it after the New Year. And we've already succeeded — it will definitely cause a huge shock to the market.

👦🏻 Koji

You mean to the AI manhua drama market, or the entire AI market?

🧑🏻‍💻 Jiangyou

The AI market. You can't imagine AI can do this.

👦🏻 Koji

Wow, so curious. Looking forward to Crossing being the first to break this news with us.

🧑🏻‍💻 Jiangyou

Sure, definitely.

👦🏻 Koji

We talked about 2026. If we look three years out, what changes do you think are most likely in this industry?

🧑🏻‍💻 Jiangyou

I think in three years, most visual content you see will really be replaced by AI, including in movie theaters. I won't say this year, I won't say next year, but the year after, you might watch a movie and afterwards someone tells you that ten-plus minutes of it was entirely AI-generated, and you might think, "How is that possible?" That kind of thing will definitely happen.

AI will move to bigger stages, not just circulating on platforms like Douyin and Kuaishou. It will enter bigger markets, bigger stages, and let more people know how powerful AI is.

👦🏻 Koji

Currently there are many other startups building products and tools for AI manhua dramas or AI short dramas. Do you use any of these products or tools?

🧑🏻‍💻 Jiangyou

No. None of them.

👦🏻 Koji

Then what kind of studios or creators are using them?

🧑🏻‍💻 Jiangyou

Teams without technology use them. Directly using tools for one-click generation, quickly and efficiently getting things done. Mostly startup teams, maybe teams of 10 or 8 people use them more.

👦🏻 Koji

If you were an investor, would you think these companies building tools and products have investment value?

🧑🏻‍💻 Jiangyou

Not at all. Zero value. I think they have two major core risks.

From an investor's perspective: first, they rely on other companies' open-source large models to get in. If one day all the large models cut you off, wouldn't your company be worth zero? The probability may be low, but that doesn't mean it doesn't exist.

Second, and most critically: if one day everyone thinks this tool works great, and I — as Shengshu, Keling AI, Dreamina, Vidu, Hailuo AI, Nano Banana — can upgrade my features to be identical in just 5 days? Then what's your advantage? What's your core competitiveness?

They're not the ones selling shovels; they're middlemen reselling shovels. Simply put, the large models are selling shovels, the tools are selling services. I teach you how to use the shovel, I write you a manual, then I sell you the shovel at a markup.

This is somewhat helpful for beginners. But for people who've been in this industry for a month or two, they'll study it themselves and find it's not that useful. So I don't think any tools currently on the market have investment value, because they have no core competitiveness. They don't even know how to compete for users.

👦🏻 Koji

Well, I'm very happy Jiangyou could join us at Crossing today. I think many of your perspectives, including your story, can be quite inspiring and energizing for people.

Thank you, Jiangyou, and best wishes for smooth sailing in your career!

🧑🏻‍💻 Jiangyou

Thank you.

👦🏻 Koji

Hope you'll have the chance to come back to Crossing.

🧑🏻‍💻 Jiangyou

Sure, thank you Koji.

🚥