Original · Unique Research · 2026-04-12
Editor's note: This translation preserves the original narrative and the supplied discussion, including disagreements and source-only claims. “IP is dead” is Yubo's business thesis, not a statement that intellectual-property rights have disappeared. Company financing, output, box-office estimates, profitability, zero-cost production and future earnings have not been independently verified here. The introduction says five guests; the roster lists four guests and a moderator. The transcript refers to a recent end-of-April surge despite the article's April 12 date, and describes Canvas as launching “next month”; that unresolved chronology is retained, not presented as current availability. The quoted offer of special API access is reproduced as part of the discussion, not endorsed or used. Names follow the source and may not be official English renderings.
Unique Awards
“IP Is Dead”: How Many AI Startup Illusions Does That Puncture?
When tools are free, content is abundant, and user attention grows scarcer, the real question for AI content startups is no longer “Can we make it?” but “Why would anyone pay?”
"
IP is dead.
I don't think IP matters at all—even when it comes to clothing.
Why should I pay for a big-name brand and then advertise it?
It should be paying me.
At a Unique Awards · Hangzhou AI WEEK roundtable, a provocative view made quite a few people shift uneasily in their seats.
“IP is dead. I don't think IP matters at all—even when it comes to clothing. Why should I pay for a big-name brand and then advertise it? It should be paying me.” The speaker was Yubo, a former Alibaba and Feishu executive and now the founder of YouMind.
Moderator Huang Lin was briefly taken aback as well. His company, StoReel, had just raised funding in the tens of millions of US dollars in North America. It makes short dramas featuring AI-generated lifelike humans, with accumulating IP assets at the heart of its business. With one sentence, Yubo seemed ready to overturn the content industry's entire table.
The discussion was originally about “Text and IP: The AI-Driven Revolution in Content Creation.” But the five guests came from different backgrounds—digital art, taking web fiction overseas, twenty years of advertising and marketing, and AI products—and turned a routine panel into an argument about how AI content actually makes money and whether it will even need to be consumed in the future.
Technology is moving so quickly that it has pushed everyone to a crossroads. Looking at all the free “productivity” scattered around them, people suddenly realize that tools are everywhere, but paying users are nowhere to be found.
Flooded with Cheap Content, How Much More Digital Junk Must We Watch?
An Shi's first question was awkward, but intensely practical.
“Many people stay up all night queuing for Sora and other video models. But what I most want to know is: How many are profitable?”
As a digital artist, An Shi is deeply skeptical of the AI-generated live-action-style short dramas appearing in bulk. His logic is blunt: Without monetization, spending money and time waiting in line is a joke. Worse, when everyone uses AI to make identical “real people,” an abyss of sameness becomes inevitable.
“Those ultra-short, gratification-driven dramas aren't expensive to shoot with real people in the first place. Now AI still means queuing and waiting. Why would everyone prefer low-priced, cheap, or even poor-quality work to something good?”
The question captures the anxiety of everyone working in content. AI does reduce costs and improve efficiency—but it reduces the cost of copying and improves the efficiency of padding things out. The pool of attention is only so large. Even Lunar New Year box-office receipts are shrinking. Why should digital junk capture users' attention?
Sandra, founder of Xuyuan Yanjin, picked up the thread. A veteran from China Literature, she had lived through the entire great age of Chinese web fiction's overseas expansion.
She pinpointed the current mismatch: “AI has so far changed only part of the production process. It has revolutionized supply, but demand-side monetization has not responded quickly.”
Put simply, AI letting you write ten comic dramas in a day does not mean readers can consume ten comic dramas today.
Sandra saw the situation clearly. In today's content jungle, she believed there were only two ways to make money:
One is to take industrialization to its limit, reproducing content in bulk from a template.
The other is to take premium quality to its limit.
She offered a particularly incisive line: “Emotion is temporary. The spirit is enduring.”
We used to consume paternalistic content. Later, short video fed us maternal-style content—whatever we liked, it kept feeding us until we were stuffed. But after watching short dramas driven purely by dopamine spikes, viewers cannot even remember the title the next day. Tools only lower the cost of trial and error. If the content itself conveys no deeper resonance, however impressive the AI used to generate it, nobody will watch.
A Fork in Monetization: From “One Experience per Person” to “Eliminating the Audience”
While Sandra and An Shi were discussing content's quality and direction, two other guests were looking toward completely different end states.
David is an advertising veteran with twenty years in the industry. To him, debating whether AI output is art is meaningless. The iron rule of commercial content is that it must serve brand growth.
What excites him about video models such as Sora is not “artistic creation,” but two things: efficiency and realism. Virtual things do not sell products; commercial expression must be grounded in a sense of reality.
“The technology I look forward to moves from ‘different experiences for different audiences’ to ‘one experience for each individual.’” David's eyes lit up at the prospect of extreme industrialization.
He described a scenario: As the same product advertisement reaches different people, AI analyzes each person's data in real time and generates a dedicated video addressing their particular needs, at zero cost and at the click of a button. This fully automated, dynamically changing perspective would be the ultimate weapon for small and medium-sized businesses to challenge larger rivals. No need to hire a film crew in Hengdian or reshoot. Just remix a product image and watch sales take off.
But Yubo clearly disagreed. He was not satisfied with AI as a more efficient salesperson. He wanted to uproot the existing model of content consumption.
“The first thing we want to do with YouMind is see whether we can get people to spend a little less time on Douyin—even if we take away only a tenth of its traffic.”
His logic runs counter to intuition: Today's AI creation tools are excruciatingly difficult to use. But that does not matter, because the future is not fundamentally about “using AI to produce better content for the masses.”
The end state is this: Stop consuming content and create directly.
Yubo's remarks exposed a blind spot. “In the future, people won't need to scroll Douyin watching others make things. If you can create for yourself with AI, the dopamine, endorphins, and even economic returns will definitely exceed what you gain as a passive spectator.”
That is why he dared to proclaim the somewhat audacious “IP is dead.” In this new world, individuals' desire to create is the gold mine. Every old order that tries to contain users within established IP will be broken.
What Are We Arguing About?
When Yubo said that bags should pay him advertising fees and that IP did not matter, the atmosphere onstage became delicate.
Huang Lin immediately picked up the conversation. Representing a platform, he did not go along with the argument; instead, he firmly defended his position.
“Entrepreneurship requires a non-consensus view. When nobody agrees that IP is useful, I'm delighted. If everyone starts applauding ‘IP is dead,’ we really won't have an opportunity.”
Huang Lin's confidence came from StoReel's experience on the market's front line. However widely technology becomes accessible, platform operations and accumulated IP still turn content into assets. Whether in North America or Southeast Asia, human genes mean we will still pay a high admission price for a digital character with a compelling persona and a good story—even if it is entirely AI-generated.
The discussion did not manufacture a harmonious grand conclusion. It cut open the unspoken thoughts of today's AI content startup world: People's underlying beliefs are fundamentally different.
The clear-eyed digital artist worries about bad work driving out good. The seasoned web-fiction operator tries to counter cheap manufacturing algorithms with an enduring spiritual core. The advertiser is excited by a tireless salesperson. The product obsessive wants to smash the board altogether and draw all of humanity into the ranks of creators.
Who will have the last laugh? Perhaps each has a chance.
But this is what the eve of a technological explosion really looks like. AI has never arrived bearing a ready-made answer for getting rich. It is more like a mirror reflecting human intentions: What you believe determines what you make with it.
The guests shared just one unspoken conviction: Stop navigating the seas of the new world with maps of the old.
More from the Conversation
Unique Awards · Hangzhou AI WEEK Trend Roundtable Panel
“Text and IP: The AI-Driven Revolution in Content Creation and the Reconstruction of IP Value”
Guests:
Digital Artist — An Shi
Founder of Xuyuan Yanjin — Sandra
Founder and CEO of ADGo — Ju Dawei, David
Founder and CEO of YouMind — Yubo
Moderator: Head of Business at StoReel — Huang Lin
Huang Lin: Hello, everyone. I'm Huang Lin from StoReel, and I'm pleased to moderate this roundtable. StoReel is doing two things. First, we use text-driven creation to produce high-quality short dramas featuring AI-generated lifelike humans. Second, we will soon launch our own creator tool, StoReel Canvas, which lets creators benefit from AI's efficiency while retaining control. Our four guests will share industry insights from the perspectives of art, narrative, technology, and products. Please introduce yourselves briefly in one sentence and name an AI trend you have been following recently. Let's begin with An Shi.
An Shi: Hello, everyone. I'm An Shi. I mainly create AIGC work inspired by Eastern aesthetics. You can find some of my work by searching for “An Shi” on Xiaohongshu. Recently, I have been following AI animation, film, and television. But I keep returning to one question: As the technology matures, I think we will enter an age of content overload. There will be more and more content, but each of us has only one share of attention. Fundamentally, this is a struggle for attention. Look at this year's Lunar New Year film market: Box-office receipts seem to be only about half of last year's. The real blockbusters are still made by directors and shown in cinemas, but I think people's attention is drifting away and they are no longer going. If AI's “cost reduction and efficiency improvement” generates enormous amounts of content that dilute our attention, and that content is inherently cheap and low-priced and may include more poor work, why would people prefer to watch it? I have never understood the commercial logic or what value it brings consumers. If people are not even watching the good work, why would we watch the cheap, low-quality material? That is an answer I hope to find here today.
Huang Lin: Sandra, please briefly introduce the trends you have been observing.
Sandra: Hello, everyone. I'm Sandra, one of the first generation of people taking web novels overseas. Recently, we have been building a creation tool for AI-oriented novel writers—what we see as AIGV. The entry barrier is a little higher because it is a desktop client. We want more people to use AI tools well to tell genuinely good stories, rather than, as An Shi said, using stronger tools to mass-produce shoddy work. Our tool is called SO Entertainment, or Xuyuan Yanjin. Our vision is that, with AI's help and support, an ordinary individual creator will be able to tell a story alone, moving from text to images, sound, and finally video, becoming a persona-based IP and a one-person entertainment company across that chain. One trend I have noticed, through talking with many writers while building the tool, is that writing is an engineering process—not simply putting chapters on a page. Reasoning backward from an engineering perspective, many things in AI should not be reduced to isolated, flat tasks. Tasks should be systematic and engineered, especially serious creation, writing, and delivery, because many people we serve aim to monetize commercially. Once creation becomes a system and an engineering process, I spend more time thinking about how to help people use AI tools through that lens—for example, making video, audio, novels, and code within an engineering framework. I am here to discuss and learn from everyone. An Shi's observations are indeed quite objective.
Huang Lin: Let's hear from David.
David: Hello, everyone. I'm David. Much of my experience has been in advertising and marketing, including executive roles at major groups such as Ogilvy and Dentsu. Over roughly 20 years in advertising, I have lived through the whole cycle—from traditional advertising to the internet, digital mobile, and social-platform content. What we do is based on a purely commercial purpose: Content serves businesses and brands, helping them grow their brands or earn profits. We are therefore more purpose-driven, unlike the previous two speakers' greater emphasis on art and creativity. Of course, essentially all content creation must serve commerce. Starting in 2023, I began converting the entire advertising planning and production process, especially the production system, to AI. That means rebuilding the full brand-content planning and production workflow—from data and user insights to social influence—with AI agents. As video technology such as Sora evolves, we keep evolving too. What we do best is systematically empower small and medium-sized businesses, which generally lack major platforms' marketing and planning capabilities. We give them those capabilities directly. Services that previously required millions or even tens of millions in advertising spend can now be obtained through these tools for very little, helping many companies grow. My observation is that video production used to take us a long time. When tools such as Sora emerge, their greatest commercial value is “realism.” The biggest difference between realism and a virtual world is that every commercial brand needs content with a sense of reality to express itself and monetize. The opportunity I see is therefore to truly empower smaller businesses in advertising production. Of course, many people still spend a great deal of time queuing for generation. If you are interested in API services without queues, you can speak with me privately. We have some special channels for obtaining them.
Huang Lin: Yubo, please share your thoughts as well.
Yubo: I'm Yubo. Since starting YouMind, the trend I have focused on most is whether we can get people to use Douyin a little less. That is the one thing we want to do. The idea behind YouMind is that Douyin and many short-video platforms have benefited society but also brought many downsides. Can we remove those downsides? Through YouMind's efforts, perhaps we can take away a tenth of Douyin's traffic. That is all we want to do, and we have kept working on it.
Huang Lin: Listening to everyone, it is clear that you care deeply about AI-generated content—its realism, artistry, or whether it is shoddily made in bulk. These questions all come back to AI tools having become more capable. What remains is people's demand for control over the tools and who creates the work. David just discussed the evolution of Sora and other technologies, whose capabilities keep growing. How can generated films better match an artist's or creator's intentions? That is what Sandra's tools address, and StoReel will release related tools too. I would now like to go deeper into how AI can truly serve creators. An Shi, as a digital artist, how do you assess the maturity of AI-generated lifelike human production? You have already shared part of your view, but the recent explosion of vertical short dramas at the end of April, with technologies such as Sora producing strikingly realistic characters, raises questions about character realism and performance. Could you discuss those from your perspective?
An Shi: The first thing I care about is this: So many people have made so much, but how many are profitable? That is my primary concern. Without profit, it is not sustainable. You make something good today, fail to monetize tomorrow, and stop the day after. Queuing, spending money, and spending time is not sustainable. It is like a self-media account: However many followers you gain, without monetization the account will eventually be abandoned. We make commercial art too. Second, making “real people” with AI does not make sense to me. When everyone pursues real people, the result is sameness—not only within AI, but also compared with live-action filming. I do not think AI-generated real people will ever outperform filming real people. Those ultra-short, gratification-driven live-action dramas are not expensive to shoot in the first place. AI is not necessarily cheaper now; it involves queues, money, and waiting. I do not think that uses AI's strengths. In commercial art, we want to compete through differentiation, to be unlike everyone else. I also want to ask everyone: When you give customers something homogeneous, how do you maximize its commercial value? If we want to charge ten times our peers' prices, but everyone uses the same templates and the same realistic people, does being faster let you charge more? That does not hold up. So I never create realistic people in my work, because that means competing in a homogeneous field. I try to stylize the work and make it artistic, developing a style people recognize as mine. Compete in another field with fewer people, and you gain the power to define your price. Making realistic people is like making comic dramas: I have not seen people around me make substantial money from it, and a very small number of exceptions cannot represent the whole. That is my understanding of AI-generated realistic humans. Perhaps a very good business model will emerge later, and I would like to hear everyone's views, but it does not yet work in my logic.
Huang Lin: As you can see, we are not being strongly adversarial. We are simply looking from artistic and commercial perspectives at how to combine them and grow together through disagreement. This is a valuable topic. Sandra, you also mentioned technology, and we have been thinking that technology must serve narrative, rather than replace what a director does. As AI technology becomes more powerful, what do you think is the most valuable and irreplaceable part of IP?
Sandra: An Shi mentioned comic dramas. Yesterday in Hangzhou, I happened to meet what is currently the industry's most profitable comic-drama company. Since joining China Literature in '11, I have worked in content operations and platforms. Over the years, I have observed that commercial narrative content—novels, live-action dramas, short dramas, films, and so on—has moved distinctly in two directions as traffic-driven operations have grown stronger. One is industrialization: mass production from a mold. The other is IP development and premium quality. I believe only two kinds of people make money in this market: those who take premium quality to the limit and those who take industrialization to the limit.
My second point is that AI has changed part of the production process and improved efficiency, producing a revolutionary effect on supply. But demand-side monetization has not changed quickly in response. Many friends, for example, make comic dramas, a product between comics and animation. In the past, this product did not receive strong demand validation in China because the audience was not large enough. AI can now mass-produce comic dramas or animation, but that does not mean everyone will immediately watch them. A revolution in production cannot quickly change consumption. Some friends in animation, however, feel that many good stories previously could not be adapted because costs were too high. With AI now sufficiently inexpensive and efficient, previously A- or S-grade stories can be released, giving people more things to watch. That is a possibility, but I have not yet seen the market truly move in that direction. Many people are actively entering the productivity revolution and learning new tools, shifting from traditional industrial production to AI-generated live-action-style work or AI dramas. Both the experience and the efficiency represent a new way of working. We cannot say everyone is making money, but at least many teams are evolving.
Third, what content is competitive? Those of us from novels often say, “Emotion is temporary. The spirit is enduring.” The shift from long-form work to short dramas reflects people's increasing emotional demands. We used to run content in a “paternal” way, telling people what was good. Today's viewers do not want paternal content; they want “maternal” content fed directly into their mouths. If they like fries, feed them fries until they are stuffed. In that maternal environment, things people vote for with their feet can enjoy enormous demand within a particular cycle. But emotion is temporary, while the spirit endures. A good story always conveys the deeper resonance needed by an era's main consumer population. After many short dramas, you cannot even remember the title the next day; the gratification ends and it is over. Good work stays with you, and if it is adapted for television, you will watch it again. Creators should therefore think more about conveying that deeper meaning. Tools only reduce the cost and accelerate the pace of trial and error.
Huang Lin: Sandra has worked deeply in content for a long time and has many ideas to share, as that detailed answer shows. Anyone interested can continue the conversation afterward in a livestream or offline. Restraint in expression makes for the best creation, whether in comic dramas or anything else. David, you have been relatively optimistic about technology, so let's discuss that. You can also respond to An Shi's question. Please try to keep it to around two minutes.
David: Let me answer An Shi first. Advertising and artistic creation share something: When serving a brand, we must ask what makes it unique and why users should buy it rather than someone else's. That forms the originality of the work. If the commercial content we produce is the same as everyone else's, we cannot help a customer build a brand or sell products. We must create distinctive communication and tone. That is consistent with original creators' goals. Advertising does, of course, have some shortcut tactics, such as “product placement through derivative IP creation.” I think technology can do that quickly. Previously, placing a product in an IP property meant bringing back the entire original team to reshoot. Now AI can take my product image and create a derivative work with the same story and characters at no cost. Technological iteration makes some commercial content much quicker and easier to monetize.
Returning to technical bottlenecks, I think video generation has reached an industrial level. There are almost no major barriers left to improving visual polish and repeatable production; those can be handled very well. For commercial advertising, I look forward to AI moving from “different experiences for different audiences” to “one experience per person,” or even “multiple experiences for one person.” The technology I especially want would automatically generate different outreach content based on user needs in real time. A product may appeal to two or three of my needs. If an advertisement reaches me three times, the system could analyze my data in real time and automatically produce three video presentations from different perspectives as each ad is delivered. Fully automated technology like that would be commercially powerful. Future technology may therefore be less about production alone and more about a fully automated workflow connecting data to creation.
Huang Lin: Yubo, you have worked on relatively successful products such as Yuque. From a product perspective, do existing AI creation tools—including the Canvas that StoReel plans to release—present a high or low barrier to ordinary creators? What core capabilities should they have to maximize their usefulness?
Yubo: First, I think YouMind is still far from successful. It is at a very early stage. Second, I do not know whether you have tried existing creation tools. Whether for video or animation—apart from article writing being slightly better—even making a PPT may still be difficult. Today's creation tools are extremely early and extremely hard to use, so difficult that they make me want to cough up blood. That is still where we are. At the same time, my third point is that AI has brought a huge change: In many future situations, perhaps we will not need to consume content but will create directly. I am fairly convinced that AI has the potential to let everyone create. People will no longer need to scroll Douyin or watch what others make. The monetization logic is also changing. It need not mean making a drama and immediately earning hundreds of millions, as An Shi discussed. Perhaps everyone will have a chance to create with AI, gain more dopamine and endorphins than consumption provides, and also earn hundreds of thousands or millions. Society is changing; everything is changing. Do not look at the new world through the old world's thinking.
Huang Lin: From a product perspective, what elements should a healthy IP trading platform generally have? You also just suggested that the future will involve more than platforms of that kind.
Yubo: IP is dead. I do not think IP matters at all, including bags and clothing. This may come from my own experience: If I buy a big-name bag, why should I pay and also advertise it? It should be paying me. Many IP logics have therefore changed. Things will become more personalized. If you can provide what I want, that is best for me. As for your IP, I do not want it at all.
Huang Lin: That is a very striking view. Let me repeat it: Yubo says, “IP is dead.” But I suspect many people here disagree. Time is limited today, so we may have to stop here for now. The collision of perspectives has been clear. Whether viewed through art, products, or monetization, the changes AI brings remain exploratory. Building on our discussion, I would like to take two minutes to explain what StoReel is doing. We began as a platform for overseas short dramas, combining AI products with content operations. Last year, we released more than a dozen dramas featuring AI-generated lifelike humans, and our AI-drama production capabilities are strong. We also have strong platform-content monetization capabilities—which partly answers An Shi's question—and the ability to operate and monetize IP assets. We turn your content into an IP asset. That is why I disagree with Yubo's “IP is dead.” Entrepreneurship requires a non-consensus position. When nobody agrees, I am especially pleased. If everyone begins applauding that IP is dead, there is no opportunity. So I am delighted. Our own tool, StoReel Canvas, will also launch next month. It is an all-model tool platform, and we welcome you to try it then. Many thanks to our four guests for sharing their views. We can continue more in-depth conversations afterward. Thank you, everyone!