Original · Unique Research · 2026-05-19
Editor's note: The first-person report and its judgments belong to the original Chinese author. This English rendition retains the opening essay, the full Q&A (Q1–Q10), all named authors, works, companies and the "source code" metaphor. Company, personal and work names are transliterated where official English forms remain unverified. Industry claims about compliance, monetization and market structure are source or speaker attributions, not independently verified findings.
Unique Awards · Guest Interview
What the Short-Drama Industry Needs Most Is Not Technology,
But Good Stories
Unique Research · Industry Observation
"
In the 2026 AI short-drama race, strange things keep piling up.
Tools keep getting better, content keeps getting more homogeneous.
The production-capacity anxiety is solved, and the creative anxiety arrives right after.
Everyone can make content, but what gets made increasingly looks like the same story told over and over—just with a different backdrop and a different character setup.
Guomai Culture enters at this juncture. After talking with Li Deheng, co-founder of Guomai Zhiyu & OpenReels, I think this is more interesting than it appears on the surface.
What They Have
Guomai Culture has been deeply rooted in the book industry for over a decade, producing many acclaimed works and gathering renowned writers such as Yi Zhongtian, Han Han, Dai Jianye, Luo Xiang and Cai Chongda. It has accumulated a large reserve of high-quality works and built a rich, diverse and highly competitive core IP matrix.
At the same time, in film and television, it has co-produced films and theatrical animations including The Continent, Pegasus and The Starry Sky of the Three Kingdoms,accumulated rich hands-on experience in IP visual adaptation.
Now, together with OpenReels, it has established Guomai Zhiyu, dedicated to AI short dramas and comic dramas, based in Shenzhen Qianhai.
On the surface this looks like crossing industries. But Li Deheng says it is the natural extension of the content industry chain, and the inevitable landing of IP value in the AI era.
"The core essence of the publishing industry has never been paper carriers and book distribution—it is the continuous discovery of quality stories, the shaping of classic characters, theaccumulated of emotional value, and the transmission of spiritual power."
He used a metaphor that I think clarifies a great deal: text is the source code of IP, and video is merely the compiled output. A book that sells hundreds of thousands of copies proves that its characters hold up, its plot withstands scrutiny, and its emotions have penetrating power. These things do not become outdated in the short-video era—on the contrary, they are the scarcest thing around. Because the biggest pain point in the short-drama industry right now is not "not knowing how to make video," but "not having good stories."
This is the core of the problem.
What publishing companies hold in their hands is exactly what the short-drama industry lacks most. It is just that in the past these two things could not be connected, because the cost of adapting text IP into video was too high—only one or two projects could be done per year, and a large number of market-validated mid-tier IPs just lay there, untouched.
AI changed this.
What AI Actually Does
It is not making video look better—it is driving down the "compilation cost."
In the past, visual adaptation of literary IP required huge crews, long cycles and massive capital, so resources concentrated only on top-tier IP. Only works with an enormous built-in fan base could support this cost structure. Books in categories such as psychological healing, workplace growth, traditional culture and popular history—categories with stable sales and real readerships in the publishing market—were always seen as "too small in scale" by traditional film and television, and nobody touched them.
Now they are being touched.
After AI dramatically lowers the barrier to content adaptation, these mid-tier IPs are entering entirely new exploratory space, and their commercial value now has a chance to land. They do not need to chase mass traffic; their audience circles are precise, their content has strong emotional resonance, and their reader base has stable stickiness. In terms of fit, they are likewise quality content choices worth deep exploration for AI comic dramas.
Li Deheng said something that captures it: a good cultural ecosystem should be one where tall trees can reach the sky, shrubs can grow in thickets, and moss can cover the earth. In the past only the tall trees were seen. AI gives the shrubs and moss a chance too.
The Industry's Real Dilemma
But there is a problem here.
More and more people are entering, more and more content is in the race, but the number of memorable works has not increased proportionally. Li Deheng put it bluntly: the problem is not that too little is being made, but that everything looks too much alike.
Machine production efficiency has gone up, but creative direction has become more blurred. Without quality, licensed IP as a foundation, teams can only follow trends and formulas—strong conflict, strong gimmicks, hitting traffic checkpoints—mass-producing things that are AI short dramas in form but have no story at their core.
And that is not all.
Compliance is another hurdle. AI face-swapping, voice cloning, and unauthorized adaptation and infringement are not uncommon in the industry, and a large number of violating works have already beenconcentrated removed by platforms. Most mid-tier teams lack professional legal support and risk-control systems; the old low-cost,coarse creative path is increasingly unworkable, and shifting to compliant operations is hard for them to afford.
Monetization is also an issue. Most teams' revenue depends heavily on platform revenue sharing, their bargaining power is weak, profit margins are continuously compressed, and it is hard toaccumulated content assets of their own.
Li Deheng's judgment: the deepest capability gap in the industry is not technology, but commercial-operating thinking and long-term IP incubation capability. Many small and medium teams have strong execution and are good at short-term project operations, and have produced decent results, but in the current rapidly changing market environment they often cannot balance long-term positioning, andaccumulated their own core assets remains extremely difficult.
What They Want to Do
This is what I find most noteworthy about this interview.
Guomai Zhiyu's positioning is not merely an AI content production and distribution company. Li Deheng hopes it will become an ecosystem builder and enabler that drives the standardization and premiumization of the AI audio-visual industry.
Relying on Guomai Culture's decade-plus of content accumulation, combined with OpenReels' investment-incubation and internet-operations DNA, Guomai Zhiyu has built a complete system from IP discovery, incubation and adaptation to AI industrialized production and full-chain operations. At the same time, they want to give teams in the industry that have strong creative and practical ability but lack long-term content foundations, capital, commercial operations and compliance support a chance to stand on a more solid starting point.
Specifically, this means several things: opening licensed IP resources to more quality creative teams, while investing capital and resources to support more emerging content creators in the AI era—solving the industry's pain point of lacking good stories at the source; sharing mature compliance and risk-control systems and industrialized production experience, helping peers reduce costs and avoid risks; building industry exchange platforms to promote efficient resource flow and break down industry barriers.
Li Deheng says the play of purely chasing short-term traffic is already cooling, and the next mainstream direction for the industry is resource interconnection and抱团 development. Guomai Zhiyu hopes to become a core force driving collaborative industry upgrading with its own strength.
As for why Shenzhen Qianhai—beyond the technology-industrymatch套, the most critical factor is going global. AI short-drama globalization is a clear trend, and content going global is not just translation—it involves a whole system of cultural adaptation, overseas compliance and traffic distribution. With Qianhai's Shenzhen-Hong Kong linkage advantage, this is a high-quality strategic high ground forintegrate the full chain of going global and telling Chinese cultural stories well to the world.
One Final Word
What impressed me most in this interview was not their business model, but that "source code" metaphor.
What the publishing industry has accumulated over the past decade or more is, in essence, a narrative database repeatedly validated by the market. Before AI, most of this database was locked up. Because conversion was too expensive. After AI, the liquidity of this content has been released.
The impact of this may be greater than many people realize.
Selected Interview Q&A
Q1. Without using the official introduction, what exactly are you doing?
We are like building a 24-hour online cinema for text IP sleeping on bookshelves. The readers are still the same people craving good content—it is just that everyone has more or less shifted from reading quietly to being more accustomed to watching content on screens. We, as publishers deeply rooted in book publishing, have naturally融入 this new audio-visual field—our identity and form are iterating, but what never changes is our commitment to good stories, quality IP and original content.
Q2. Why enter AI short dramas at this point in time?
The core industrial value of AI is to dramatically lower the cost of converting text into video, turning high-barrier, scarce adaptation into low-cost, routine production. In the past, visual adaptation of literary IP had high production costs and long development cycles, and a large amount of quality mid-tier content could never achieve visual presentation. Now, AI has kicked down that barrier.
Q3. Why can the text assets accumulated by the publishing industry be reactivated today?
Text is like the source code of IP, and video is merely its compiled output. In the past, compiling once was too expensive; now AI has dramatically压低 compilation costs, and the value of the underlying source code naturally迎来 a comprehensive revaluation. A book that can sell hundreds of thousands of copies proves that its characters hold up, its plot withstands scrutiny, and its emotions have penetrating power—these elements are not outdated in the short-video era; on the contrary, they are the scarcest thing.
Q4. What is your view on the judgment that mid-tier IP has a chance to be developed?
This is one of the most easily underestimated contributions of AI to the content industry. A good cultural ecosystem should be one where tall trees can reach the sky, shrubs can grow in thickets, and moss can cover the earth. In the past only the tall trees were seen—only top-tier IP could support the cost structure of traditional film and television. After AI, mid-tier IP does not need to compete with nationwide mega-IP for total internet traffic; as long as they deeply cultivate their circles, with precise audiences and high fan stickiness, they can run a stable commercial path.
Q5. Among your IP reserves, which type of content is most suitable for AI comic dramas?
Psychological healing, workplace growth, traditional culture, popular history and similar categories—these have stable sales and fixed audiences, but are often seen as too small in scale from the perspective of traditional film and television. Placed in the AI comic-drama race, these IPs are instead quality content choices worth deep exploration: precise audiences, clear emotions, high fan loyalty. They may not be so-called mass "blockbusters," but they have deep reputation among vertical reader groups, and this kind of recognition is very hard to replicate.
Q6. What is the most real dilemma in the AI short-drama industry right now?
The problem is not that too little is being made, but that everything looks too much alike. AI has solved the efficiency problem of how to produce and how to mass-produce, but what content to make and what good stories to tell remains the corechallenge for the entire industry. Many teams, starting from production-capacity anxiety, have gradually fallen into creative exhaustion—the higher the machine production efficiency, the less everyone can find a clear creative direction.
Q7. How serious is the compliance problem?
AI face-swapping, voice cloning, unauthorized adaptation and infringement are occurring frequently, and a large number of violating works have already beenconcentrated removed and rectified by platforms. Most mid-tier teams lack professional legal support and complete risk-control systems; the old low-cost,coarse creative path is gradually becoming unworkable, and shifting to compliant operations is hard for them to afford, leaving them in a very被动 position. This is not a low-probability risk—it is already happening.
Q8. What is the deepest capability gap for mid-tier teams?
Many teams are deeply rooted in frontline production and have strong practical execution ability, but they generally lack systematic commercial-operating thinking and full-process content risk-control capability, and it is also hard for them to balance the mindset of long-term IP incubation. Daily development is mostly based on short-term project cooperation, and they can neveraccumulated their own core content assets. The more they do, the more it feels like working for someone else.
Q9. How does Guomai Zhiyu plan to cooperate with industry partners?
We maintain an open mindset and, relying on our own content and incubation advantages, collaborate with and empower industry partners. First, opening licensed IP resources to more quality creative teams; second, sharing mature compliance governance and risk-control experience to help peers operate steadily; third, continuously building industry exchange platforms so that resources flow more effectively between different teams. The play of purely chasing short-term traffic is cooling, and resource interconnection and抱团 development is the next direction—we want to be the ones who make this happen.
Q10. Where do you think the entire race is headed?
The industry is at a critical节点 of transition fromcoarse growth to standardization and premiumization, and this is an inevitable trend. But during the transition, teams without content foundations and without risk-control systems will have a very hard time. The deeper problem is structural—inconsensus on creative value, low efficiency in industry-chain collaboration, and missing industry standards. These are not things any single company can solve alone; they require platforms with resources and strength in the industry to take the initiative to integrate and collaborate to break through.
This article is based on an interview by Unique Research with Li Deheng, co-founder of Guomai Zhiyu & OpenReels.