Friday, November 20, 2026

Understanding China. Seeing what’s next.

Understanding China. Seeing what’s next.

Writing AI dramas in China used to pay off. Now a veteran says delivering food is safer

By November 2026, Dandan, an AI short-drama scriptwriter based in Hangzhou, had survived three years of relentless shifts in China’s booming micro-drama industry by jumping across three or four different companies. Today, as hit rates for AI-generated dramas on platforms like ByteDance‘s Hongguo plunge to roughly one in a thousand, his advice for anyone thinking of quitting their day job to write AI dramas full-time is blunt: you would be better off delivering food.

Producing an AI short drama from initial pitching to online release typically takes just one to two weeks, leaving Dandan a tight window of five to seven days to finish a script. If quality standards are relaxed, he can churn out a manuscript in three to five days.

These days, his role is strictly confined to content creation and writing. He no longer touches storyboarding, asset generation, or video editing. While he admits he has “lost touch with the technical side of AI tools,” he believes his core storytelling skills translate seamlessly from traditional human-acted short dramas.

Just three years ago, he was writing web novels during the sector’s golden age, when his former employer could rake in more than 50,000 yuan ($6,900) in daily profit. Since then, he has transitioned from online literature to live-action short dramas, and finally into AI-generated animated series. He has lived through virtually every trend in the ecosystem: traffic-driven distribution, original content creation, expansion into China’s lower-tier cities, and low-cost production.

Though he describes himself as someone who successfully rode these market waves, he speaks of the journey without excitement. Over the past three years, he has witnessed two major industry upheavals and watched one employer completely vanish. Each time the market shifted, he was forced to reassess his position. His only regret, he says, is that he did not jump ship to new tracks fast enough.

Writing AI dramas in China used to pay off. Now a veteran says delivering food is safer

Short Dramas Started as Little More Than Novel Ads

Dandan entered the short-drama industry in mid-2023, just as bite-sized vertical videos were morphing into a standalone commercial empire. Before that, his career was rooted in online literature, spanning the eras of mobile reading, 3G, 4G, and paid-traffic promotion.

He views novels and short dramas as essentially cut from the same cloth: the same narrative presented through text versus audiovisual media. Short dramas were not born out of a vacuum; they evolved as a monetization offshoot from the web novel business. The emotional hooks, pacing control, and paywall designs perfected by novel companies could be directly transplanted into short video scripts. Consequently, a large cohort of the earliest entrants into the short-drama industry came from online literature firms, and Dandan was among them.

Initially, he stayed out of direct creation, focusing instead on distribution. Armed with traffic-buying permissions on Jiuzhou’s short-drama mini-programs, he promoted other people’s content to turn a profit before finally deciding to produce his own shows.

At the time, Jiuzhou and Dianzhong Technology were the dominant heavyweights in the sector. Jiuzhou’s early productions were far from polished, yet they proved lucrative. Many established web-novel companies, however, doubted the model’s longevity, dismissing rough, smartphone-shot content as a short-lived cash grab that would survive for only two or three months.

By the time those conservative companies finally entered the market in 2024, they were already six months to a year behind the pioneers.

Writing AI dramas in China used to pay off. Now a veteran says delivering food is safer

Every Time the Industry Changed, He Changed Companies

Dandan’s career trajectory mirrors the structural evolution of China’s short-drama market: starting with traffic distribution, shifting to original content, pivoting to low-cost productions tailored for lower-tier markets, and finally embracing AI-generated content when it surged in late 2025.

“Every single transformation in the industry meant I had to switch companies,” he says.

His departure from the online literature sector was driven by a fundamental disagreement with his employer over the viability of short dramas. While many in the novel industry believed short dramas would vanish within months, Dandan saw a golden opportunity and resigned.

This was not a choice born of artistic idealism, but a survival calculation common among practitioners facing rapid market disruption: when new mechanics emerge, clinging to obsolete models means missing the next wave. An employee’s only viable move is to seek out companies willing to embrace change.

His experience at a subsequent employer further validated this philosophy.

Between late 2025 and mid-2026, the volume of AI-generated animated series exploded while live-action short drama output shrank to 20% or 30% of its previous peak. Lighting technicians, cameramen, and actors across studios in Hengdian—China’s massive film and television production hub—suddenly found themselves out of work. This slump, Dandan notes, served as a clear early-warning sign.

Yet his employer failed to take action—a fatal mistake in the fast-moving short-drama sector. Dandan concluded that the company was “almost certainly doomed to be eliminated” because the market would no longer tolerate players relying on outdated playbooks.

True to his prediction, the company eventually exited the business entirely. Reflecting on the episode, he adds a generous note about his former boss, saying he “was a decent guy” who managed to pay out all back wages until the very end.

Writing AI dramas in China used to pay off. Now a veteran says delivering food is safer

Three Months Is the Limit

Following the collapse of his employer, Dandan spent two to three months unemployed at home.

Initially, he looked for operational and distribution roles, but his job applications yielded no results. Operations and distribution for AI animated series differ vastly from live-action short dramas, having evolved into an entirely distinct profession where “past work experience doesn’t carry much weight anymore.”

Because his previous employer had been slow to adopt AI animation, Dandan lacked direct hands-on exposure, meaning his traditional short-drama operational expertise could not easily be leveraged. He soon pivoted his job hunt back to content creation and quickly landed a new position.

Economic pressure, he admits, is unavoidable, and China’s high-pressure socioeconomic environment “leaves very little margin for error,” making extended career pauses impossible. Those two or three months were less of a relaxing break and more of a self-imposed survival deadline.

During his unemployment, he also tried producing an AI short drama on his own. He spent half a month writing a script and another week uploading and testing videos, sinking between 1,000 and 2,000 yuan ($140 to $280) of his own money into the experiment.

The test was short-lived after he crunched the numbers: according to industry data he cited from Hongguo’s first-half ecosystem report, the platform’s overall hit rate hovered around 1‰ (one in a thousand), while more than 400,000 animated series had been launched on the platform in just the first six months of the year.

He ran the math: producing an 80-episode AI animated series at his current company requires a production budget of 40,000 to 60,000 yuan. For a corporate balance sheet, that is standard project overhead; for an individual creator, it is a formidable financial gamble.

Risking 40,000 yuan of his own savings to chase a 1‰ chance of breaking even makes little financial sense, he argues. “I might as well buy a lottery ticket. It’s much easier and saves me the trouble,” he says.

Even if a show miraculously strikes gold, the payouts may not justify the gamble. Dandan believes Hongguo’s per-ten-thousand-views revenue has shrunk to the point where gambling tens of thousands of yuan for a 0.1% chance of a return offers terrible risk-reward symmetry. Consequently, he pulled the plug on his solo venture.

While he knows how to write scripts and operate the necessary AI tools, individual creators cannot absorb the cost of continuous trial and error. “If you give me an infinite budget, I can make any show you want,” he says, noting that his solo experiment was ultimately halted by financial constraints.

Lower Technical Barriers Mean Higher Financial Hurdles

When outsiders discuss AI short dramas, they frequently fixate on technical bottlenecks such as maintaining character consistency, lip-syncing accuracy, voice acting, and visual stability. To Dandan, those challenges “were hurdles of late last year.” Today, anyone willing to learn can master them within a day or two.

Now that technical capabilities are democratized, the real bottleneck crushing producers is the grim math of costs and profit margins.

He points out that AI-generated dramas mimicking live-action styles have grown remarkably refined, with some productions rivaling real-world footage. While critics often lambast AI characters for stiff facial expressions, Dandan notes that such flaws have largely vanished on premium minidrama portals like “Green Bubble Short Dramas” (a colloquial reference to WeChat-based mini-programs). Shows on Hongguo still feature stiff animations largely because the platform hosts a higher proportion of anime-style content compared to live-action aesthetics.

As for the concept of “one-click drama generation,” he acknowledges its technical feasibility but doubts it can reliably produce quality art. Whether a one-click show is any good is debatable, he says, but “anything generated with a single click is bound to be trash. If you could generate boutique dramas with a single click, this industry would cease to exist.”

In his view, while AI can automate an increasing number of production steps, determining whether a story actually works still requires human intuition to guide rhythm, character dynamics, emotional resonance, and narrative arcs. Creativity remains the hardest asset to replace, though he concedes that even this boundary is shrinking: “Now, AI is beginning to compete with ordinary people even in the realm of creativity.”

Ranked by vulnerability to displacement, he believes production staff and directors will bear the brunt first. As AI model capabilities improve, studios will no longer need labor-intensive prompt generation, and directors will no longer need to micro-manage every visual frame. Scriptwriting, however, will continue to rely on human experience for the foreseeable future.

This reality is what drove him back to a content role—partly due to his writing proficiency, and partly for professional self-preservation. “Content creation is a relatively scarce skill across the entire industry; other skills simply aren’t,” he says.

Platforms Hold the Novels, the Traffic, and the Exits

When discussing financial distribution in the sector, Dandan pulls no punches: “Everyone is working for ByteDance,” he says, because ByteDance is the only entity guaranteed to turn a profit.

His reasoning points to ByteDance’s fully integrated ecosystem: Fanqie Novel supplies the source stories and intellectual property, Hongguo handles short-drama distribution, and both TikTok’s Chinese counterpart Douyin and Hongguo control the traffic spigot. Production companies launching dramas on Hongguo must continuously purchase traffic to promote their own titles. “They take all our money, and we are left earning meager wages for our sweat,” he says.

Revenue data shared by other industry insiders confirm this margin squeeze. The head of a production studio with dozens of employees reported that the company’s total monthly revenue across all channels sits between 500,000 and 600,000 yuan, with only about 40% of titles breaking even. At a major top-tier studio capable of pushing out 120 dramas a month, the company president estimates that only 10% of projects ultimately yield a positive return.

While mega-hits still deliver outsized payoffs—generating seven, eight, or even tenfold returns—they function more like lottery jackpots covering the losses of other projects rather than representing the norm for the broader industry.

Dandan argues that the most chaotic segment of the market isn’t driven by established heavyweights, but by “a massive influx of dabblers.” Viral internet stories trumpeting low-cost, high-yield short dramas continually lure outsiders into sinking hundreds of thousands of yuan into a handful of productions. Lacking familiarity with platform guidelines, content standards, and production workflows, these amateurs flood the market with low-quality, boundary-pushing garbage.

He notes that with over 400,000 animated series launched in just the first half of the year, they cannot all be masterpieces. “Probably 90% of them are garbage churned out by people playing around for fun.”

He remains deeply skeptical of viral folklore about stay-at-home mothers or ride-hailing drivers striking gold after a few days of work. Regarding popular independent hits like Entering Song with Weapons, which have been marketed as self-made marvels, Dandan says that while concrete details are scarce and he prefers not to speculate, he personally finds it hard to believe: “Only those who have actually done the work know how brutally difficult it is to engineer a hit.”

He readily admits that whether a show goes viral is not purely a function of quality; mediocre productions occasionally strike a sudden chord due to luck and serendipity.

Even so, if someone close to him announced plans to quit their job and plunge into AI short dramas full-time, he wouldn’t encourage them: “I’d tell them their brain is fried. Of all the things you could do, you’d be better off delivering food.”

Only Those Who Stay Can Catch the Next Wave

Asked whether he still remains optimistic about AI short dramas, Dandan splits his answer into two halves: the industry perspective and the individual perspective.

For the industry, he concedes that while market size is expanding and competition is intensifying, the exodus of small, amateur teams and the reduction of low-tier content could ultimately improve survival conditions for studios with stable production capabilities. With 400,000 shows, the hit rate is 1‰; if volume shrinks to 40,000 shows, the hit rate rises to 1%, simply because audience attention is finite.

He describes this filtering process as “purification”: the departure of small and medium-sized players cleanses a content ecosystem that had been muddled by low-tier noise competing for traffic. With fewer rivals standing in the way, those who remain will find it easier to operate.

For the individual worker, however, he notes that he has few alternatives: “I don’t know how to do anything else; this is all I know.” He remains willing to press on, offering neither blind optimism nor despair, framing his entire existence around a single objective: “How to become part of the surviving cohort instead of being phased out.”
Most people in the industry will lose money, he recognizes. His job is not to figure out how to win the 1‰ lottery, but to ensure he isn’t kicked out of the game when the next industry shift arrives.

As the interview concludes, Dandan must return to his keyboard. An 80-episode script awaits, and his seven-day window is ticking down.

(At the interviewee’s request, Dandan is a pseudonym.)

Leave a Reply

Your email address will not be published. Required fields are marked *