---
title: "This Roundtable Shattered Three of My Assumptions About AI Short Dramas"
author: "Unique Research"
sourcePublication: "Unique Research Substack"
originalPublishedAt: "2025-12-18T08:01:09+00:00"
canonical: "https://ffcap.cn/en/research/src-20251218-01html"
source: "https://uniqueresearch.substack.com/p/src-20251218-01html"
language: "en"
---

# This Roundtable Shattered Three of My Assumptions About AI Short Dramas

_Original · Unique Research · 2025-12-18_

_Historical edition: This complete transcript preserves the speakers’ claims, commercial figures, opinions, and forecasts as reported in December 2025; they are not independently verified current findings. AIGC means AI-generated content, and “AI live-action” here describes AI-generated imagery resembling live action, not necessarily filmed performers. The source’s phrase “the upcoming November” is retained despite its December publication date; the event date and intended November are not independently established. Unspecified currencies in budgets, quoted fees, and revenue-sharing examples remain unspecified. The opening’s “more than ten million users” and the guest’s later distinctions between followers, users, and paying users are both retained as stated. Translated work titles are descriptive; the source names only “Wukong,” without identifying a specific title, and itself qualifies “Shengxia Fendela.”_

A short drama of less than twenty minutes, divided into dozens of vertical-screen segments consumed in one sitting, is quietly rewriting the content industry's underlying logic.

On the subway, on the sofa, or under the covers, viewers pay a few yuan for stimulation and stress relief. Behind the screen, however, creators and platforms are really competing over who can turn content production into an "industry" rather than a succession of isolated creative acts.

When AI entered short dramas, it magnified this question by several orders of magnitude.

If a production no longer requires dozens of people and more than ten consecutive days; if an IP owner's written world can be converted cheaply and at scale into an audiovisual world; and if one person can deliver the output that once required a small team, then are short dramas one step closer to "industrialized content"?

Who can genuinely absorb this explosion in productive capacity without being harmed by disorderly, excessive competition?

How should the line between "tools lowering the threshold" and "specialization through content tiers" be redrawn?

This time, we invited almost everyone working at the front line of AI short dramas onstage: a pioneer who explored the field through The Mirror of Mountains and Seas and The Nameless Special Operations Team and was among the first to benefit from "AI comic dramas";

A creative hotshop that produced a high-priced commercial short drama for Yili's Jinlingguan brand and is preparing to implement genuine "mass production" and "industrialization" in Wenzhou;

A product team that compares itself to "Bilibili for AI series," creating content while building a native community;

A director who switches among overseas productions, live action, animation, and custom brand work, using real-world data to calmly unpack the relationship between "blockbusters" and "high quality";

And a company entering film and television from education, using data and cases from more than ten million users to remind us that the adoption of tools never equals the adoption of capabilities.

They were moderator Zhang Yuanyuan, a partner at Yilun AI; Chen Kun, founder of AIpai; President Huang, Huang Chujie, founder of Shanghai Aimwise creative hotshop; Zhang Rui (Jerry), COO of StoReel; Yanchuan, founder & director of CoreX-AI Film Studio; and Huang Luxiang, Vice President of Business at Jinghai Shibei.

At the 2025 Beijing Unique Awards workshop for super individuals titled "AI Short Dramas—Driving a New Era of Industrialized Content Production," they discussed:

Why the comic dramas now enjoying both booming supply and sales sit precisely between users' minimum aesthetic expectations and the industry's maximum technical capabilities;

Why AI live-action short dramas are constrained not only by technology, but also by cost structures and talent structures;

Why some people are destined to pursue mass production in which overwhelming effort produces breakthroughs, while others can only defend premium, top-tier work;

And how, as the "technological sugar coating" gradually wears away, thinking only about computing costs and focusing only on the idea that "AI is cheap" can damage genuinely good content.

One participant drew a timeline for the industry: "Last year was the year of working for love, this year is the year of staking out territory, and next year may be the year when fruit can finally be harvested."

Another left those still watching with two extremely simple words—"Try it."

Below is the complete transcript of this conversation about AI short dramas and the industrialization of content. We have lightly smoothed the language while preserving the pace and emotion of the live discussion as much as possible.

If you care about how AI will transform the content industry, or you are preparing to "stake out a position" in this field, it is worth reading from beginning to end.

Guest Introductions

Zhang Yuanyuan: Welcome to our roundtable. I am delighted that at this moment we can discuss a subject that may change our content industry: the industrialization of short-drama content.

Short dramas are in fact a very new phenomenon for the content industry and have achieved explosive growth over the past several years. Unexpectedly, the AI era has arrived, and AI plus short dramas may create new industrial opportunities.

Today we are fortunate to have invited several colleagues working in AI short dramas to discuss the industrialization of AI short dramas. First, I would like each guest to introduce themselves and their business briefly, and use one keyword to define what you see as AI's core value in today's short-drama ecosystem. Let us begin with Brother Kun.

Chen Kun: Hello, everyone. My name is Chen Kun, and online I am known as "Xianren Yikun." I should count as one of the earliest people to make AI short dramas. In July last year, we worked with Kuaishou on The Mirror of Mountains and Seas, which now appears to have been among the first several AI short dramas. At the end of the year, we then collaborated with Stephen Chow on The Nameless Special Operations Team, which was a genuine "AI comic drama"—what now seems like an ancient example of the form.

We currently operate a platform called AIpai. In addition to making comic dramas ourselves, we cooperate with institutions to explore the AI comic-drama category and have recently been conducting several experiments.

If I had to use one word to describe the relationship of AI to this field today, the word I most hope to achieve would be “turning AI into a practical tool,” or perhaps "industrialization." That is the direction in which I have greater confidence.

President Huang: Hello, everyone. My name is Huang Chujie, and I am known in the industry as "President Huang." Aimwise should be considered one of the teams relatively far ahead in commercializing AI. Short dramas are today's subject, and in commercial short dramas we also set a precedent by producing a highly commercial animated short drama for Yili's Jinlingguan brand this year.

Of course, everyone today also cares about how AI short dramas can be industrialized. Our next major move will be to turn "implemented industrialization" into reality, so I will probably implement this work in Wenzhou.

I have observed that platforms do not lack projects; they lack more professional, systematic teams capable of handling specialized needs. My goal is therefore to implement this "mass production" and "industrialization."

Zhang Rui: Hello, everyone. I am Zhang Rui, co-founder and COO of StoReel. Our company is a native community platform for AI series that entered through AI short dramas.

At the beginning of our entrepreneurial journey, we mainly made live-action short dramas. Of course, we also saw the major changes in narrative capabilities that the development of AI tools brought to the industry, and we transformed very early. We now actively embrace AI in both content and product. On the content side, we have an in-house AI content team and a large number of AI PUGC creators, embracing the AI ecosystem to produce content at greater scale.

On the product side, in addition to offering an app where users watch short dramas, we use various interactive mechanisms so users not only watch but also play on our platform. You can think of it as "Bilibili for AI series."

If asked for the most important keyword describing AI's effect on this industry, I would say "creativity at scale." What does that mean? In the past, shooting a traditional live-action short drama still required a crew of several dozen people. It might be filmed in eight or nine days, but if you personally had an idea to express, a large team was still required to complete it. With AI, many creative people—whether screenwriters, directors, or related roles—can produce the content they want very easily. Enabling everyone's idea to be replicated at scale is, in my view, AI's core variable and keyword.

Yanchuan: Hello, everyone. I am Yanchuan, founder of CoreX-AI Studio and a director. I am a director familiar to many people in the AI creative community. Brother Kun just said that in Naruto terminology, he is the "First Hokage." As for me, to put it unflatteringly...

Chen Kun: Just call yourself an "old-timer."

Yanchuan: You defined that yourself. We all know one another well in this circle. I have consistently worked on major AI films and other AI short dramas. In short dramas, I cover a relatively broad range. I may be one of the few people working on overseas short dramas, live-action short dramas, animated short dramas, and custom brand short dramas, so I have extensive experience.

One strong feeling I have about short dramas is that everyone is making tremendous noise, but the real test is who can commercialize this model well—what we often call achieving better monetization.

I began making commercial overseas short dramas in October last year. When I shared the work then, I was asked why I did not make them in China. My view was that domestic competition was excessive. That is also the point I want to define today: everyone discusses and makes AI short dramas, and the field looks highly active. But after entering it, we need to think more about how the market, business conditions, and rules can form a healthy cycle without descending into destructive competition.

As technology iterates, I think thresholds for tools and other aspects may fall. But that lower threshold will bring more people into the field. From one perspective that seems positive; from another, it may not be so wonderful.

Huang Luxiang: Hello, everyone. I am Huang Luxiang, head of marketing and business at Jinghai Shibei. Sitting today among veterans and experts specializing in film and television, I admit that being the one person from education—AI education—seems slightly out of place.

By coincidence, before being invited to participate, we entered a strong collaboration with Google and the TapNow team. We will subsequently produce a series of Google's advertisements on YouTube for them. This should be our education company's first preliminary step into AI film and television.

Although we have just entered film and television, in education we are already China's leading comprehensive operation. Our founder, Li Genji, has an account on Bilibili called "Genji Really Wants to Teach You." I do not know how many people have seen it; those who have studied PS, AI, or PR may know it. We now have more than 10 million followers across the internet. Our main business is AI courses, through which we have served more than 2 million users, including more than 80,000 deeply engaged paying users.

In the upcoming November, we will use the user data accumulated in the past to launch our flagship AGI education application. We therefore very much hope to speak with fellow practitioners after the event.

Returning to AI's use in film and television, some time ago there was an event you may have seen: By-Health and an Antarctic expedition team jointly launched a million-level AIGC television commercial (currency unspecified in that description). Our partner showed me the film and asked me to guess what it had cost. Although I do not work in film and television, I studied film production at both undergraduate and master's levels. I said that sending all those people to Antarctica must have cost several million (currency unspecified in the source). He said no—the "computing cost" was only 30,000 yuan.

I said, my goodness. Anyone interested should watch that film; I found it astonishing. If I had to use one word, I would say AI truly "reshaped" the production relationships of traditional film and television. It genuinely broke what we call the "impossible triangle"—low cost, high quality, and a short cycle. We now appear to be seeing the beginning of that trend.

Topic One: The Ranking and Current State of Factors of Production in AI Short Dramas

Zhang Yuanyuan: You are all frontline explorers deeply engaged in AI short dramas. In your view, what stage have AI short dramas reached? If we begin from factors of production, what will be the most important factors next—computing power, models, high-quality data, production processes, or something else? Let us again begin with Brother Kun.

Chen Kun: Let me share some data from our current work. We now produce more than 600 minutes of comic dramas each month. That is only this month; next month we hope it will double, followed by several consecutive months of doubled growth.

What supports this? We now see not only Hongguo but also China Literature, Qimao, and other major IP owners opening comic-drama access to their IP. Why? It is simple. Written IP of course has value and may already earn strong returns on a platform. But if AIGC converts written IP into video IP, its value expands by several orders of magnitude. Judging from the industry's average level, it will certainly increase by orders of magnitude.

Why, then, are comic dramas so popular today? We call it "booming supply and sales." I have one modest opinion: today's comic dramas sit at a crucial point where the minimum level users will appreciate overlaps with the maximum level of the AI technology industry.

When each of us watches different content formats, our minimum acceptable standard differs. A few days ago, for example, I watched a terrible film that everyone should avoid. I "ate excrement" for 2 hours in the cinema and came out cursing it constantly. Its screenplay, story, dialogue, acting, and visual language were all terrible. But think about it: you paid 30 yuan to watch it in a cinema. If the same thing appeared on a phone as a short drama, or what was called a "major production," you might think, "Oh? This is not bad."

Our consumption mindset therefore differs across content formats. How do people consume short dramas? I cycled to the top of Dayang Mountain, where the owner of a small shop had placed a phone playing short dramas loudly while telling customers, "That will be three yuan, five yuan." The consumption mindset is simply different.

When I refer to the AI industry's overall technical level today, I do not mean the capability of one person. A top team may produce a film approaching a million-level audiovisual effect (currency unspecified in the source), as the previous speaker described, but one person's capability does not represent the industry level. By industry level, we mean a standard that at least 1000 individuals or institutions can achieve. Today, comic dramas are the one and only category capable of reaching that standard.

Top teams such as Yanchuan's may be able to make live-action dramas, but today's general industry level cannot, so a healthy market with both strong supply and demand has not yet formed.

What will the future depend on? Growth in the industry's overall level. That will come partly from increasing model and Agent capabilities, and partly from more talent entering from other industries, improving both production standards and familiarity with AI tools. I believe these developments will bring an explosion in AI-like live-action dramas next year—and I do believe it will be next year.

Zhang Yuanyuan: So you firmly believe AI short dramas will still move toward live-action imagery rather than first pursuing an illusory, animation-like form?

Chen Kun: Today we can only produce this form; its explosion today occurred because this is all we can do. From the audience's perspective, they do not care whether something was made by traditional means or AI. Consider today's deliberately crude comic animations: little AI is used in their production, though more is used in screenwriting and dubbing. Audiences do not care; they simply think, "I enjoy this—give me more."

Zhang Yuanyuan: All right, thank you, Brother Kun. President Huang, how do you view the current stage and the ranking of key factors of production?

President Huang: This resembles going out to eat. There is private-kitchen cuisine and there is fast food, such as KFC. Each is a commodity, and different commodities correspond to different markets.

For products distributed at scale, we must pursue the ultimate SOP (standard operating procedure). Only after an excellent SOP is clarified can even a 10% improvement in efficiency become a 10% net profit margin for a company. That is crucial.

Private-kitchen cuisine, meanwhile, depends heavily on the chef's craft. An exceptional chef may be especially skilled at Cantonese cuisine or Sichuan and Hunan cuisine; that is private-kitchen food.

Different product structures therefore imply different priorities among factors of production. Private-kitchen cuisine must be "people-centered"; fast food must emphasize "efficiency and process." I do not need every dish to taste absolutely perfect, only the overall production efficiency to be stable, reliable, and high.

Zhang Yuanyuan: Good. Jerry, how do you view this question?

Zhang Rui: Because our company focuses more on AI live-action short dramas, I believe the category is currently in a "semi-automated" stage of scaling. Producing a genuinely high-quality short drama still requires extensive refinement and human participation throughout the process. It has certainly not reached the scale that today's AI comic dramas or motion comics may already have achieved. AI live-action short dramas remain at an intermediate stage.

If you ask me to rank factors of production, I place standardized SOP first. Once you define the entire creative process—from script and storyboard to image generation, video, Lip-sync, and editing—whoever solves this complete SOP best and implements it most efficiently at the greatest scale can truly bring scaling into reality.

The second-most-important factor is data and assets. In the creative process for AI live-action dramas, I interviewed many directors and found considerable difficulty in maintaining consistency and ensuring consistency across asset libraries. If we accumulate the relevant data and turn consistency—for example, casting actors—into reusable assets that everyone can preserve, it will certainly improve efficiency substantially.

Next would be the evolution of model capabilities, which directly determines substantial productivity gains. Where I may originally have needed 20 generation attempts, perhaps I can now achieve a strong result in 5 attempts. Computing power may come last, although token prices are continually declining as model capabilities improve.

Zhang Yuanyuan: I am also curious: if models evolve to a certain level, will the preceding process need to be reconstructed? How do you balance what comes first?

Zhang Rui: I understand. Even as model capabilities evolve, I will always place SOP first because it is a production workflow. My asset library may perhaps move lower in priority: models may already be able to solve consistency, multi-character performance within one scene, or unnatural lip movement very well. If all those problems are solved, the value of that component will continue to decline.

Zhang Yuanyuan: Understood. Thank you, Jerry. Director Yanchuan, how do you view the question?

Yanchuan: Why do we divide AI short dramas into live action and motion comics? Why are AI live-action dramas unable to gain traction today? Several dimensions are involved.

One is technology; another is cost. The technical problem is easy to understand: people compare live-action performance and movement with the texture of actual filming, immediately exposing advantages, disadvantages, and shortcomings.

Why do animation and motion comics perform better? People already understand traditional animation as animation, and it contains what we often call a "jerky feeling." Current AI technology can often satisfy that standard and therefore integrates well with it. When watching, we know these are animated characters, so people tolerate movement that lacks fluidity or performances that fall short.

That does not mean AI live-action dramas cannot improve through subsequent technical iteration. Some good live-action short dramas already achieve strong results in capable hands, though output may remain limited to a very small group at the pyramid's apex.

As thresholds fall and SOP processes become standardized, large volumes of content will enter, and the market will begin differentiating user profiles. One group will certainly defend premium, top-tier work, while another will mass-produce content on the principle that overwhelming effort creates breakthroughs.

Yesterday, President Huang and I discussed which identity is most awkward for an AI content producer today. The middle position is the most awkward. You may be unable to reach the standard of top-tier premium work, which may remain limited to certain super-IP creators. But some people are also unwilling to make cheap, high-volume market content assembled for a few yuan, because anyone creating content carries at least some idealism.

I therefore believe the market will move toward two poles. You must choose your direction: pursue volume in the hope that one of ten works becomes a hit, or make only several works a month and turn them into premium productions.

Zhang Yuanyuan: That is a profound insight. I want to follow up on the cost issue you mentioned earlier. Do you have any intuitive data? Is producing with AI now more expensive than live filming, or what is the approximate level?

Yanchuan: I cannot give general data, but I can describe my own evolving feelings.

At the beginning of this year, many investors approached me about making short dramas with AI. Creator-run media accounts were filled with dramatic claims such as "how much can be produced in one day," evoking a Great Leap Forward. Investors thought AI was already powerful and cheap. At the time, if someone offered half the cost of a live-action production to make it with AI, I considered that relatively conscientious.

Conversely, the cost of AI creation is only the cost of the steel and cement used to renovate a house. If the work is content, consider why film and advertising cost so much. Is it the camera expense? No—it is the value of the content produced.

If people want to pursue high-volume, street-stall-style output, that is fine; they can run that model. Establish one SOP, and platform-tool providers like Brother Kun can run it very quickly and produce a great deal. That is a market segment with recognized demand.

But if you want something good while paying a price based on some perceived "computing cost," I think the approach must change. You are buying good content.

Some people say during discussions, "My budget is only 100,000 or 200,000." I ask why they do not look at the subject they want to produce. A modern urban story can be filmed live very quickly, with dozens of episodes shot on set in one day. But if you are making fantasy or science fiction requiring extensive production design, why assume that a little more than 100,000 or 200,000 can solve it? With AI, however, the price is fair to young and old alike; every subject costs the same. Is that not cost-effectiveness?

Zhang Yuanyuan: Understood; fascinating. Mr. Huang, do you have a different insight from this perspective?

Huang Luxiang: Regarding the ranking of factors of production, because we work in education, I think it must be divided by stage and industry.

We have accumulated nearly 4000 students and users. Sometimes when enterprises send us requests, we give the work to students. We have discovered one major problem: brands do not clearly understand the boundaries of AI capabilities.

Last month, a brand told me it had a budget of only 200,000 and described the scenario it wanted. I told them bluntly that using AI for this scenario might look somewhat fake, and live filming would produce a better result. They said, "How could that be? Are we not told that everything is possible now?"

On the other hand, when we quote a customer 200,000 for a three- or four-minute AIGC film, they say, "My goodness, do you not just type two lines into a prompt? You are charging 200,000?"

That is the misalignment in understanding between the two sides.

Viewed by industry, the people at the forefront of AI production rank models first. They have already established their SOP, and algorithmic models are now what limit their performance.

At the beginning of the year, when someone told me they wanted to build an all-AI film and television project, I suggested waiting. Only five or six months later, we can discuss it this actively today because we have genuinely seen breakthroughs in algorithmic models.

I believe that in the near future, perhaps next year, AI-generated content may become genuinely indistinguishable from filmed work. At that point, using a low-cost, highly efficient, stable SOP for production will make SOP more important.

Topic Two: "Blockbusters" and Commercialization in AI Short Dramas

Zhang Yuanyuan: Although everyone is watching AI short dramas, so far no phenomenal blockbuster—especially an AI-produced short drama—has entered mainstream public view. What are the reasons? If one appears in the future, what small spark might set it off? Let us begin with Jerry.

Zhang Rui: In both China and overseas, we have already seen some commercially validated AI live-action short dramas. In China, small examples may have generated more than five or six hundred thousand yuan in payments within a week; overseas, some have validated monetization at hundreds of thousands or just under a million US dollars. I therefore think the model is beyond doubt.

But a true blockbuster earning tens of millions of US dollars like a traditional live-action short drama may not yet have arrived. I see several reasons:

First is still model capability. The long workflow requires extensive human intervention, preventing genuinely scaled, high-quality production.

Second is the team. Integrated teams that understand scripts, hooks, and emotional reversals while also producing content with high-quality AI tools remain relatively rare.

Third, every high-quality blockbuster must rest on volume. Producing one AI series still requires one or two months, so the overall scale is insufficient.

Another point is that I look forward to diverse forms for AI live-action short dramas, including interactive and participatory content, which may help produce a blockbuster.

Zhang Yuanyuan: President Huang, what do you think?

President Huang: Fundamentally, I think there is still no hot money in this vertical industry. In the first half of this year, we made a custom commercial series lasting 20 minutes and charged 1.3 million, or 65,000 per minute. I think that is already the pricing ceiling for AI short dramas, but it remains a custom product.

This results from the fact that more professional teams have not yet entered; most current practitioners remain individual workers or hobbyists. Most directors who have produced blockbuster short dramas currently look down on this work.

Overall, however, the tipping point will arrive. People like Director Chuan and Director Kun still need to go ahead and encounter the hazards.

Zhang Yuanyuan: Since you have been mentioned, Brother Kun?

Chen Kun: I have one feeling that may be relatively pessimistic: the "technological sugar coating" AI gives everyone is becoming thinner.

When we premiered The Mirror of Mountains and Seas in July last year, more than a hundred media organizations filled a cinema despite torrential rain. At that time, AI content was wrapped in technological sugar coating. Today, AI comic dramas no longer have that coating. When users consume AI content, they do not care whether it is AI or live action.

I therefore believe that before "content formats only AI can create and traditional film and television cannot" appear, an explosion on the level of Ne Zha or Wukong is unlikely.

For business owners, investment scale comes down to four Chinese characters: reduce costs and increase efficiency. Under this premise, no one can spend 100 million making a 90-minute AI film. Without an investment of 100 million, how could such a thing appear?

I therefore think a universally known, purely AI blockbuster may have to wait for the next era.

President Huang: No, I think a small hit is enough. A series earning 10 million in revenue sharing would already be satisfactory.

Chen Kun: Then no one would know it. Some series now receive more than 100 million views but may earn only an ordinary amount. What counts as a hit? Consider the recent Shengxia Fendela—the source adds that this may refer to a particular work or popular short drama— which flooded feeds. That was a hit. But it resulted from changes in genre and production polish and had little to do with the technology itself. Viewed separately, I do not think AI technology alone will produce a major hit.

Zhang Yuanyuan: Director Yanchuan, what do you think?

Yanchuan: High quality and blockbusters are two dimensions. High quality does not guarantee a blockbuster, and a blockbuster is not necessarily high quality. Hits depend more on story, creativity, and marketing.

AI's lower costs and greater efficiency make it an excellent mechanism for testing potential hits through data and volume.

Yesterday, President Huang and I discussed a model called the "short-drama mystery box." Once AI production capacity catches up, overwhelming effort may allow some works to strike an audience need.

The other route is to produce high-quality premium work. Someone must identify hazards and encounter pitfalls for the industry.

Returning to short dramas, when we make overseas short dramas, investors are unwilling to invest large sums mainly because they do not know whether the money can be recovered. If a hit were certain, people would even take loans to produce it.

We therefore use one model for filmed live-action dramas. For a 50-episode series, we do not complete all of it; we make the first 5 to 10 episodes and immediately release them to the market for testing. We even design paid unlocks into those 5 to 10 episodes, not to make money, but to use data such as completion and unlock rates to analyze whether the series has hit potential. This is a way to balance costs.

Zhang Yuanyuan: That is quite pragmatic. Mr. Huang, since you chose education, how do you view talent?

Huang Luxiang: The film industry has developed for a century, yet China has produced few blockbuster works in recent years. AI entered only two or three years ago, and while neither volume nor scale has formed a system, expecting a major hit is genuinely difficult.

The upper limit of tools does not represent the upper limit of content creation, and widespread tools do not mean widespread capability.

When handling YouTube advertising work, we discovered that AI film talent must first be film talent and then know some AI for the work to succeed.

We must first teach them film aesthetics and storyboards. Only after they possess basic aesthetic decision-making ability in film can they begin making AI films. We therefore now work with several well-known Chinese artists to lead students through production.

Topic Three: Tools and the Future Outlook

Zhang Yuanyuan: Brother Kun, is your product intended for ordinary consumers or professional producers?

Chen Kun: We want both. We created two modes:

Agent mode: serves consumers. You do not need to understand audiovisual language; the Agent helps you write prompts, choose voice timbres, and dub.

Standard mode: like a car's "manual transmission." You can make precise revisions to every shot and prompt, meeting commercial-delivery standards.

Some time ago, I wrote a dedicated SOP for short dramas. Anyone who needs it can ask me.

Zhang Yuanyuan: President Huang, are standards for commercial advertising clearer?

President Huang: The standards are unclear—even less clear than theirs. It is much harder. The largest cost in commercial advertising is not the tool cost but the communication cost. You must satisfy too many requirements: placement expectations, selling points, logo size, and more. Every piece of content is "haute couture."

Zhang Rui, adding a point about global expansion: there is no fundamental difference between overseas and domestic markets. We strongly value what creators provide—aesthetics, creativity, and understanding of shots. We are willing to pay high-quality top creators more rather than price their work according to computing cost.

Conclusion: One-Sentence Questions and Advice

Zhang Yuanyuan:

Finally, please offer practitioners and interested people one question or recommendation worth considering.

Yanchuan: AI is an inevitable future trend. My experience is that last year was "the year of working for love," this year is "the year of staking out territory," and next year may be "the year of harvesting fruit." If you cannot stake out a position this year, next year's fruit will not belong to you.

Do not try to turn yourself into a six-sided all-round warrior; it is too exhausting. Doing well in the field you know is enough. As I say, there are only a few places at the summit, so I will simply collect garbage at the foot of the mountain. This garbage is mine to collect, and I can become the king of recyclers.

Huang Luxiang: Many people feel anxious that AI has arrived while their lives have not changed. There is no need for excessive anxiety. First board this train and learn more. Get on the ship before deciding anything else.

Chen Kun: I have only two Chinese characters: “try it.” If you are interested in AI, try it—and of course, first try our “aipai.ai” platform.

President Huang: Find your position. As the industry matures, professional roles will become increasingly specialized. I simply want to be the most profitable man among AI creators.

Zhang Rui: AI tools are a good knife, but only a great chef can truly prepare a good dish. I encourage young creators to use AI tools to make more high-quality content.

Zhang Yuanyuan: Good. Thank you all. We look forward to everyone finding their differentiated advantage in the AIGC wave. Thank you to all our guests!

---

Original publication: https://uniqueresearch.substack.com/p/src-20251218-01html
On-site reading page: https://ffcap.cn/en/research/src-20251218-01html