The Wealth Creation Illusion and Content Reconstruction of AI Short Dramas
When an AI short drama *Armed into the Song Dynasty*, which was "collaboratively created online" by online ride-hailing drivers, stay-at-home moms, freight drivers and ordinary office workers, has reached 10 seasons with about 1,000 episodes, each less than 2 minutes long, claiming over 12 billion total views across all platforms and being constantly urged for new episodes by netizens, it is far more than a simple cultural and entertainment news story.
This year, the successive hits of the two blockbuster AI short dramas *Wife in the Wasteland* and *Armed into the Song Dynasty* have landed exactly at the industry's turning point: AI short dramas have gone through the "tool verification phase" and entered a period of structured industrial game. Most analyses on the market only stay at the level of "whether it is good to watch, whether it is profitable, and whether it will lead to homogenization", ignoring that the biggest impact of AI short dramas is not the addition of a new type of short video content, but the rewriting of the bargaining power, cost structure and value distribution logic of every link in the content industry chain.
But almost at the same time, the hashtag #Hongguo Short Dramas Reduce Intelligence went viral on hot search lists, with controversies pointing directly at whether AI short dramas are "computing power garbage". In the same market, while some hail it as a miracle, others criticize it for "reducing intelligence" — this is the most real cross-section of AI short dramas: it is no longer an "emotional fast-moving consumer good" in the sinking market, but it is far from entering an inclusive era where everyone can make a fortune with low cost easily.
A netizen who is currently learning AI comic dramas told New Business Research Institute that the production of AI short dramas generally includes the following processes: generating stories, creating scripts and storyboards, extracting character information, generating 3-view drawings of characters and scene images, creating prompts for video generation, and finally generating videos. The threshold for using AI tools themselves is not high, and ordinary people can master them after one or two days of learning. But to produce high-quality works, every process needs to be refined. Script polishing, historical research, character shaping, narrative logic and prompt setting are the real tests that cannot be mastered in a short period of time.
From text and images, audio, long videos to short videos, live streaming and AI short dramas, every content carrier is evolving and splitting. The richer the content ecosystem is, the fiercer the competition for audience attention will be, and the traffic and commercial value will also transform accordingly.
From Overbearing President Melodramas to *Armed into the Song Dynasty*, Short Dramas Enter the 2.0 Era
If we review the development of short dramas in recent years, we can see a relatively clear watershed.
Before 2026, short dramas were mainly live-action works with themes of counterattack, revenge, overbearing presidents, mother-in-law and daughter-in-law conflicts, and war gods, targeting the sinking market and middle-aged and elderly groups. The content logic was "fast stimulation, fast forgetting, fast monetization", with face-slapping plots in 3 episodes and plot reversals in 5 episodes. They took amplifying human nature and conflicts as hooks, and completed the full viewing rate by investment promotion and refreshing pleasure points, giving birth to phenomenal works such as *The Son-in-law* and *I Became a Stepmother in the 1980s*. We can call this the 1.0 stage of short dramas.
Compared with long dramas, the biggest advantage of this type of live-action short drama is cost. A short drama of about 100 minutes usually only has a production cycle of about one month, does not need star actors, and costs an average of only hundreds of thousands of yuan. Relying on huge cost advantages, live-action short dramas have shown a surging trend, and the richness of themes has also attracted a larger audience. However, in the past two years, industry competition has been fierce. In order to pursue higher certainty, the investment in high-quality short dramas has generally reached millions or even higher levels. The cost has risen sharply but the revenue has not increased, leading to an awkward situation.
In recent years, with the maturity and diversification of AI technology, AI short dramas have been launched one after another. The first representative work to go viral is *Wife in the Wasteland*, a female-oriented farming and time-travel AI comic drama created by a post-95s couple, which has exceeded 4.1 billion views across all platforms. Until recently, the male-oriented historical time-travel drama *Armed into the Song Dynasty* broke through the circle again with its rigorous historical background and well-researched costumes and props, attracting history and culture enthusiasts, young male users, knowledge-based users and long video users, directly expanding the audience boundary of short dramas.
It can be said that the real significance of the popularity of AI short dramas such as *Wife in the Wasteland* and *Armed into the Song Dynasty* is not that AI is better at "generating images", but that AI enables short dramas to have a composite form of "serialized literature + genre drama + fan operation" for the first time, marking that the short drama market has entered the 2.0 era. This is reflected in the fact that the creation quality has shifted from "routine density" to "self-consistent setting, knowledge density, and character growth"; the audience has expanded from sinking middle-aged and elderly groups to Generation Z, hardcore male content fans, female farming novel readers and overseas users; the commercial ecosystem has changed from single drama sales to a combination of platform revenue sharing, series IP, advertising, e-commerce, overseas expansion and copyright licensing.
Data shows that the number of new AI short dramas launched this year continues to rise. Nearly 550,000 micro short dramas have been launched in the first 10 months, nearly 17 times the total number of 2025, of which AI dramas account for more than 94%.
The Illusion of "AI Wealth Creation", The Lively Account Is Actually Very Cold
Judging from current media reports and hot search topics, the success of AI short dramas such as *Wife in the Wasteland* and *Armed into the Song Dynasty* is mostly packaged as a story that "ordinary people can also create wealth with AI". The labels of this story are mainly amateur teams, low cost, zero traffic investment, tens of billions of views, and tens of millions of returns.
However, if we further break it down, *Armed into the Song Dynasty* presents another version of the story. This story includes: in terms of IP selection, the original work is the novel of the same name on Fanqie Novel, and the reason why the lead creator Zhang Hansheng chose it is that he "can't stand the face-slapping routines of overbearing president dramas"; in terms of asset library construction, 100 episodes a week are made manually by "card drawing", ordinary shots can be passed after an average of 3 generations, and grand war scenes need more than 10 generations; in terms of capital investment, although there is no traffic investment, the computing power cost of the whole series exceeds 1 million yuan.
The team behind *Armed into the Song Dynasty* is not a complete novice. Zhang Hansheng admitted that in order to survive, the team had up to 5 crews working at the same time at its peak, producing about 20 "assembly line dramas" a month, accumulating more than one year of running-in and practical experience in undertaking a large number of commercial orders, and forming the current collaboration foundation after continuous trial and error. Now the monthly output has dropped to single digits, and one highlight scene of an episode takes one to two days to polish, with most of the time spent on the script.
"In the second half of this year, a large number of AI short drama companies across the country have closed down. Many people and many companies we know have gone bankrupt, and there are also large-scale layoffs," Zhang Hansheng reminded. Many people only see the result, but not the efforts behind it. "We stay up all night, polishing every plot little by little. It is not as easy as imagined."
A set of data also confirms the difficulties of the AI short drama industry. Statistics from Zhongtai Securities show that the production cost of AI short dramas has dropped by 80% to 90% year-on-year, only 2% to 10% of that of traditional short dramas. With the cost reduced, the supply has exploded. In the first half of 2026, 221,900 AI dramas were launched on Douyin. Taking 50 million views as the break-even baseline, only 1.3% of the new dramas did not lose money, and only 1 out of about 77 dramas could recover the cost. The hit rate of AI comic dramas is even less than 0.1%.
Objectively speaking, AI has indeed reduced the production threshold of short dramas to an unprecedented low level, but cost reduction does not mean lower standards. The more you use AI, the more it tests people's subjective abilities. The choice of stories, the understanding of characters, the grasp of emotional propriety, the selection of editing points and the understanding of narrative logic are still in the hands of human beings, which are the core values that machines are difficult to replace.
In addition, AI has also reduced the cost of film and television production, but the saved money has not stayed in the pockets of producers. It flows to computing power providers (model companies), traffic distributors (platforms), and traffic investment intermediaries (advertising systems).
While AI short dramas are booming, doubts are also surging, mainly focusing on two levels: content quality and cognitive impact. Some netizens believe that the templates of revenge, counterattack and overbearing president are repeatedly used in many short dramas, the motivation of characters' behaviors is vague, plot conflicts are promoted by coincidences, and there is a binary opposition between good and evil, with almost no gray area. Some netizens also think that AI-generated characters have stiff performances and mechanical dubbing, with too strong "AI flavor" and poor viewing experience. The deeper concern lies in the anxiety of "intelligence reduction". Some netizens reported that after watching dozens of episodes in a row, their "mind goes blank" and "they can't remember what they watched specifically", believing that long-term exposure to strong stimulating content that does not require thinking will weaken the willingness of active logical deduction.
Some industry insiders said bluntly that AI has reduced the cost of a single drama from about 300,000 yuan to about 50,000 yuan, shortened the production cycle to 7 to 15 days, and expanded the output to dozens of times that of live-action dramas. The efficiency dividend is real, but the accompanying controversies are also real: homogenization, "intelligence reduction", minor protection, copyright and personality rights infringement, and the most "universal" criticism — it takes up too much time.
In response, responses from regulators and platforms are being intensively implemented. The National Radio and Television Administration has recently deployed a new round of special governance against vulgarization and homogenization of micro short dramas, which will last for two months from November 1 to the end of December. This governance not only targets the dramas themselves, but also includes titles, promotion materials, bullet screen comments, and even platform home page recommendations and popular lists into the remediation scope.
Reconstruction of Content Industry Ecosystem Under Co-opetition
The superposition of short dramas and AI effects is also reconstructing new traffic entrances and content ecosystems.
According to the latest data released by QuestMobile: as of September this year, the daily active users of Hongguo exceeded 200 million for the first time, reaching 203.2348 million; the average daily usage duration reached 165.98 minutes.
With the expansion of user scale, Hongguo has gradually formed a closed loop from IP supply to traffic distribution: Fanqie Novel provides online literary IP, Douyin is responsible for diversion, Hongguo undertakes viewing, and then realizes monetization through advertising.
At the same time, Tencent's Green Bubble Short Drama Mini Program had 89.68 million weekly active users in the first week of October, which is close to the 100 million mark, an increase of 44.1% compared with the first week of September. At present, WeChat short dramas have completed the entrance replacement, and searching for "short drama" will default to entering the Green Bubble Mini Program. The advantages of Green Bubble are light entrance and convenient access, while the shortcomings are that content supply and user habits still need to be established, and the content options provided to users are not rich, which involves content construction, copyright authorization, cooperation between content parties and producers, and revenue sharing systems.
At the same time, long video platforms are no longer waiting and watching, but accelerating the integration of AI into the content system and commercial system of their main sites.
This year, Mango TV established the AIGC Innovative Content Center and launched *The New Journey to the West*, China's first AIGC long drama broadcast on satellite TV; Tencent Video distributes through both the Huolong Comic Drama channel and the main site, with the exclusive pure revenue sharing copyright coefficient up to 240%. Some comic dramas have achieved revenue sharing of over 20 million yuan per work with a production cost of about 400,000 yuan; iQIYI has also upgraded Nado Pro from a tool to an ecosystem, with unified revenue sharing for live-action short dramas, AI short dramas and comic dramas; Youku has recently officially launched new revenue sharing rules for AI short films.
From the perspective of brokerage research reports, the industrial logic of AI short dramas is extending from simple cost reduction and efficiency improvement to content supply expansion, IP value release and overseas market expansion. At the same time, the content form of AI film and television is also expanding from short dramas and comic dramas to medium and long dramas, movies and interactive video games.
The explosion of short dramas brings both challenges and opportunities to major video platforms, and also gives an opportunity to rearrange the pattern of the content industry. In particular, AI short dramas can be regarded as the "low-cost IP verification layer" in the entire content ecosystem. Compared with the extremely high IP trial and error cost of the long content industry, AI short dramas can quickly turn an IP into a series of episodes with a cost of tens of thousands or hundreds of thousands of yuan to test user retention, comment feedback and payment willingness. If the market feedback data is good, we can further develop live-action short dramas, long dramas and animations to build a more extensive full-link IP development; otherwise, stop the loss in time.
Conclusion
The viral popularity of *Armed into the Song Dynasty* has provided the industry with a precious sample: AI short dramas can achieve 12 billion views without relying on melodramatic plots, traffic investment or professional teams, relying on a good story and respect for history. But the scarcity of this sample exactly exposes the truth of the industry. Only 1.3% of the 220,000 dramas do not lose money, and the vast majority of AI short dramas are still algorithm feed, discarded after use.
Policies are tightening, supervision is stepping in, and major manufacturers are seizing the market. AI short dramas in 2026 are going through the stage that short videos experienced ten years ago: the barbaric growth is coming to an end, and regular troops begin to take the market. This also means that the curtain of a major restructuring of the content industry's power structure has been officially opened.
This article is from the WeChat official account "New Business Research Institute", written by Xiao Yi, and published with authorization from 36Kr.