Who could have predicted that the short drama industry would be the first to feel the tremors of AI disruption?

As the lingering warmth of the Spring Festival holiday faded, an industry upheaval known as "the shake-up of short dramas" rippled through WeChat Moments and social media circles of industry insiders. The hashtag #HongguoSuspendedManyProjects# surged on Weibo's entertainment hot search list, with whispers swirling that Hongguo, a leading domestic short drama platform, was halting production of live-action short dramas en masse and redirecting its budget towards AI-generated comic dramas.

Amid the swirling public opinion, Li Liang, Vice President of Douyin Group, swiftly stepped in to clarify and refute the rumors, stating that "Hongguo will persist in increasing investment in live-action short dramas" and that incentives for high-quality short drama content would remain in place.

Yet, the official reassurance failed to fully dispel industry anxieties, as the meteoric rise of AI-generated content had become an undeniable reality. According to DataEye statistics, among the 8.67 billion total views during this year's Spring Festival short drama season, AI-generated comic dramas accounted for nearly 30%, a figure that was nearly negligible just six months prior.

The fierce competition between live-action teams and AI technology underscores the short drama industry's shift from a "manpower and speed" model to a "technology-driven computing power" era.

| AI "Prefabricated Content": Revolutionizing Production Capacity |

If 2024 marked the foundational year for AI large models and 2025 the inaugural year for AI-generated comic dramas, then by the 2026 Spring Festival season, AI-generated comic dramas had shed their experimental skin and evolved into "cash cows" capable of amassing billions of views and driving substantial commercial value.

Take the Spring Festival season's dark horse, Zhanxiantai AI Live-Action Version, as an example. The drama soared to the top of Hongguo's comic drama charts just two days after its release and surpassed 100 million views within six days. In the realm of 2D dynamic comics, Journey to the West: Mistaking the Jade Emperor for One's Father, released on the second day of the Lunar New Year, broke through 60 million daily heat (popularity) and rapidly amassed over 210 million views across platforms.

In DataEye's February top 100 comic drama rankings, the total view increment reached 7.587 billion, with AI-generated live-action short dramas surging at an astonishing rate, accounting for nearly 60% of the share. These figures demonstrate to the entire industry that, given sufficient audiovisual stimulation, AI-generated short dramas can rival live-action dramas in terms of traffic aggregation and user engagement.

Indeed, the rapid adoption of AI-generated short dramas by platforms and investors stems from their fundamental reshaping and simplification of traditional film and television production capacity. This reshaping pursues extreme standardization, drastically low costs, and high turnover rates that defy physical limitations—mirroring the "prefabricated meal" revolution in the catering industry.

Compared to the asset-heavy model of live-action dramas, lightweight production is the absolute strength of AI-generated short dramas. Conventional live-action short dramas require script polishing, cumbersome casting, location scouting, and are often constrained by weather, venue rentals, and actor schedules. As a result, the hard costs for ordinary projects typically hover around 500,000 yuan, with top-tier production companies like Tinghua Island and Majiu reaching millions.

In contrast, AI-generated short dramas, with next-generation video generation models like Seedance 2.0 and Sora addressing key shortcomings such as shot continuity and multi-angle character consistency, have broken free from physical-world constraints and established a streamlined, industrialized digital closed loop.

As long as the prompts are precise and the large model grasps the director's intent, AI can generate epic battle scenes—whether immortal-demon duels, post-apocalyptic wastelands, or thousand-strong armies—in just tens of seconds. Content that once required traveling to distant locations can now be perfectly produced in a single office with a few high-configuration computers.

When producing Journey to the West: Mistaking the Jade Emperor for One's Father, leading company Spark Animation had already established a mature "211 production standard," capable of creating a viral hit with just a 20-day production cycle, 100,000-level computing power costs, and a 10-person team, continuously pushing the efficiency limits of film and television industrial pipelines.

For short drama content platforms, a steady stream of new dramas to sustain user retention is the core fuel for their algorithmic recommendation mechanisms. The plummeting costs of AI-generated short dramas mean producers can generate more content with the same budget, maximizing the hit-making logic of "trial and error, rapid iteration."

| A "New Species" of Short Dramas Meets AI "Reshuffling" |

The rapid expansion of production capacity inevitably accompanies the collapse of old business models. This "short drama earthquake" is not merely the adoption of new tools but a major reshuffling of industry stakeholders' interests and power dynamics. For many mid-tier short drama production companies based in sink markets (lower-tier cities) like Xi'an and Zhengzhou, the past Spring Festival was particularly challenging.

The short drama industry's flourishing diversity in recent years was largely built on platforms' "guaranteed mechanism" crutch. To secure absolute production capacity, platforms provided guaranteed production fees to numerous producers. For small and medium-sized teams, regardless of whether a project became a hit or how revenue was shared, this fee was sufficient to cover the hard costs of live shooting.

However, if content platforms tighten or even suspend guaranteed fees for conventional live-action projects, the financial lifelines of these small and medium-sized producers will snap. Teams accustomed to being platform "contract manufacturers" will be pushed to the front lines of self-financing and bearing significant market risks, leading to a surge of "production crew disbandment" and "company-wide AI transformation" announcements on social media.

The fundamental reason capital is ruthlessly "overturning the table" and embracing AI at this moment lies in the increasingly fierce competition in the traditional short drama red ocean over the past year.

After fandom culture infiltrated the once lower-tier market-dominated sector, actor salaries for short dramas began to irrationally surge. Some top short drama actors now command daily fees approaching 50,000 to 80,000 yuan, with a single short drama burning through tens of millions just on the salaries of the male and female leads. Coupled with soaring costs for costumes, props, location rentals, and expensive user acquisition spending, the hard costs of short dramas continue to climb.

According to relevant media reports, the overall loss rate in the short drama industry now exceeds 90%, reducing it to a red ocean gambling den where "nine out of ten dramas fail."

Under immense pressure to recoup costs, producers, afraid to innovate and take risks, can only fall into path dependency, frantically replicating market-tested hit formulas. The same hit IP is remade by multiple teams, with the novel Jingzhi Yuye being adapted into over a dozen versions like Zhizhi Fuzhizhi, fatiguing audiences and causing hit rates to plummet.

When platforms and capital realize that live-action dramas' viewership, user retention, and final revenue cannot match those of AI-generated immortal cultivation comic dramas produced by small teams with just 100,000 yuan in computing power costs, capital's shift becomes an absolute commercial inevitability.

Under the new system, a new profession of "all-rounder AI producers" is emerging. They may not be formally trained directors but understand audiovisual language, shot logic, and can efficiently interact with AI tools to "draw cards" (generate content), amplifying their personal productivity tenfold or even a hundredfold through AI empowerment.

In this new production context where "one person is an army," the barriers of traditional short dramas relying on manpower and resource accumulation are shattered. This reshuffling not only eliminates outdated production capacity and business models but also transfers creative rights and era dividends of film and television content to "super individuals" wielding next-generation productivity tools.

| Stripping Away the Bubble, Returning to the Emotional Core of Good Stories |

Faced with the invasion of efficient and low-cost AI-generated short dramas, are live-action short dramas truly headed for oblivion? The answer is no. When we look past this wave of technological euphoria and strip away the conceptual hype, it becomes clear that AI still has fundamental weaknesses at this stage and for the foreseeable future.

Zhou Hongyi, founder of 360 Group and de facto controller of Nanomi Comic Dramas, acknowledged while embracing AI that producing short dramas with AI still faces challenges. Ensuring consistency in characters, props, scenes, and storylines requires further technological breakthroughs.

Although models like Seedance 2.0 have solved issues of visual consistency and grand scene rendering, they inevitably struggle with conveying the subtlest and most complex human emotions.

Human performances consist of countless subconscious details. Current AI still cannot stably and naturally generate the micro-expressions that construct a character's deep inner world. AI can create a flawless image of a protagonist shedding tears but cannot grasp the complex emotions behind those tears.

In Zhanxiantai Live-Action AI Version, the nuanced mindset of protagonist Lu Fan—experiencing both pain and joy after avenging his family—could not be generated by AI. Although the voice acting expressed Lu Fan's multi-layered emotions, the character's stiffness still jolted audiences out of immersion, severely weakening the intended emotional tension and resonance.

Given AI's strengths and weaknesses, a future where "AI single-handedly slaughters live-action" is unlikely in the short term. As Li Liang stated in his clarification post, both have advantages in different genres. Thus, a clear industrial divergence and a healthy dual-track ecosystem may be the ultimate destination for industry players and platforms.

On one hand, AI can dominate the "spectacle and thrill" genre. Themes like immortal cultivation, primordial chaos, sci-fi, post-apocalyptic wastelands, infinite-flow Ghost Stories (weird tales), and psychic awakenings require grand world-building and rely on high-density visual effects. Audiences in these genres prioritize imaginative and visual thrills over nuanced interpersonal emotional tug-of-wars.

In Teaching for Five Years, I Cultivated a School of Demons, released during the Spring Festival season, the plot was filled with grand scenes of divine beasts wielding flying swords and myriad demons battling. Traditional live-action filming with CGI would have required a budget in the tens of millions, but with AI generation, creators could present a mythological epic texture (texture) at low cost simply by typing and adjusting models.

On the other hand, live-action teams can deep cultivation (cultivate deeply) the "real-life emotion" genre. Rather than competing head-on with AI in visual spectacle, they should invest in script polishing to make genres requiring strong empathy—such as urban realism and family portraits—an impenetrable fortress for live-action short dramas.

Take this year's Spring Festival season hits A Family of Three in the Same Class and Northbound as examples. The former tapped into contemporary audiences' real desire to break traditional family authority and build warm households, while the latter, through Lunar New Year homesickness, delicately portrayed selfless kindness among strangers during migrant workers' Spring Festival travel rush, striking a chord with countless viewers.

The genuine resonance that cannot be precisely calculated by code and computing power once again proves that no matter how democratized or advanced technology becomes, the film and television industry will ultimately return to the timeless essence of "telling good human stories." In this AI-driven industry upheaval, computing power and models determine the lower limits of short drama production capacity and the breadth of visual effects, while insight into audience needs and sincere emotional delivery still define the upper limits and soul of works.

And when the technological tide recedes, creators who embrace new productivity tools while steadfastly upholding content standards will emerge as the true winners at the table.

Image sourced from the internet. Rights reserved for the original creator.