AI PRODUCT
No Animation Work After Graduation: Graduates Who Embrace AI Motion Comics
“Our generation has nothing left to lose”
The craze for "AI comic dramas" is staggering. From the earliest meme-style collages, to dynamic comic-style 2D, then 3D cartoon and even hyper-realistic; big companies entering the scene, data soaring, expectations inflating, and state-backed entities like CCTV signaling support—AI comic dramas are being produced, expanded, and iterated at an astonishing speed.
In stark contrast, traditional animation and film face high barriers, high costs, long cycles, and uncertain market returns. Many creators remain cautious or wary of AI.
In this context, current or recent animation students may find themselves trapped in anxiety: How to adapt to the current job market? How to affirm the value of their own creations? What is the meaning of studying animation?
Some animation graduates have made a rather "alternative" choice: they embrace AI comic dramas and join the front lines as directors. Here are their accounts:
01Xiao Y(pseudonym) works in the AI production department of a large enterprise. Before producing AI comic dramas, the team also explored AIGC in areas such as artistic short films, music videos, and pre-production concept designs for films.
During her school years, Xiao Y focused on traditional frame-by-frame hand-drawn animation, having almost no exposure to AI animation. Her undergraduate graduation short film had a fresh art style and a relatable story, garnering over 100,000 views online and receiving comments praising it as “hopefully turned into an anime series.”
But faced with the post-graduation job market, Xiao Y chose to accept a job in the AI field.
Although most of her classmates oppose AI, she never had strong feelings about it, describing herself as “highly tolerant of new things” and someone who “tries to do well in whatever field she’s in.”
Initially hired as an artist, Xiao Y now primarily serves as the director and writer for AI comic dramas within the department, handling tasks such as storyboarding, video generation, and editing—covering almost the entire AI comic drama production pipeline.
She recalls the difficulties of learning AI when she first started. Early on, using Midjourney to generate images required learning specific prompts to adjust image proportions and character designs. “Back then, learning SD(Stable Diffusion) was also agonizing—just installing and configuring it on your computer already put you ahead of 99% of people.”
However, as models improve and products iterate, using AI becomes easier and easier.
Both SD and Midjourney have long since faded from her workflow. The team has now turned to online AI video generation products like Runway, Sora, Jimeng, Keling, and Vidu, increasingly using domestic Chinese AI tools.
The open-source/local AIGC ecosystem is still developing, but in the efficiency-focused, price-squeezed environment of AI comic drama production, she prefers ready-made products: “just generate directly.”
When she first entered the industry, Xiao Y felt AI was “dumb.” When she needed to modify an image, she thought typing a long description was more trouble than just editing it herself; but now, to save time, she prioritizes letting AI regenerate. When editing, she no longer uses Premiere Pro but directly uses CapCut, “because it’s lighter and has a lot of internet-savvy built-in assets.”
Xiao Y remembers that before NanoBanana and the new version of Jimeng launched, all image generators felt like “little idiots”—even the most detailed prompts couldn’t achieve the desired effect. But in the past six months, AI’s ability to use reference images and edit has improved dramatically. “It can now fully follow the settings I provide to replace and generate images—basically, it has what you might call Photoshop capabilities.”
In fact, the three major limitations often criticized when AI animation first emerged—line jitter, character consistency, and camera continuity—Xiao Y believes have largely been resolved in AI comic drama production. Because AI comic dramas avoid the troublesome rotoscoping workflow and instead directly generate video from images; existing AI products can generally maintain consistent character appearances; and camera continuity depends on the operator’s own directing and editing skills.
02 Using a legendary sword to chop vegetables
If you only looked at her college assignments and short films, you might think Xiao E(pseudonym) would thrive in the 2D animation industry, becoming a “master animator” whose name appears in anime credits.
The animation shorts she made in college featured intense action and stunning transformations, with impressive spatial perspective awareness. Her undergraduate graduation short also captivated hundreds of thousands of viewers online with its emotional release(also praised in comments as 'hopefully turned into an anime series'), earning industry recognition.
Near graduation, Xiao E began researching mainstream 2D animation companies in China—the results were discouraging. Among the companies he looked into, about 60% were making game advertising animations, and 40% were either working as contract studios for platforms or sticking to original creations. After learning more details, he backed off.
“At that time, I kept wondering how I could stay true to my original creative vision,” Xiao E recalled. “I didn’t have a clear answer. I couldn’t figure anything out.”
This confusion became his entry point into AI. Although he had dismissed AI during university and even believed it would never affect animation production, the latest AI animation results changed his mind: “This thing might actually be able to tell stories.”
So he packed his bags and resolutely joined an AI comic drama production company as a director. “Anyway, I’m young and have no burdens. One person full, the whole family is full. Since I didn’t get a real chance in animation, I might as well seize the current trend. Our generation has nothing left to lose—what’s the harm in trying AI?”
Once he started producing AI comic dramas, Xiao E was also amazed by AIGC’s efficiency. The work can be simply described as a loop using products like Sora, Jimeng, and Vidu: “Generate storyboards → Manual review/regenerate → Generate animation → Manual review/regenerate → Output.”
A team of about 20 people can produce nearly 120 minutes of comic drama in two weeks—a volume unimaginable in traditional animation.
But he emphasizes this is a very rough estimate; in practice, multiple projects run in parallel at different stages. And often, “they just put effort into the first three to ten episodes, then let a bunch of interns churn out the rest. As long as the basic story and character design aren’t completely ruined, it doesn’t matter.”
After working for a while, he also formed some impressions and evaluations of the major AI tools: for example, Sora 2 is better at Japanese anime style, Vidu Q3 excels at action and dynamic shots, and Jimeng’s images have higher saturation. Nowadays, he can tell at a glance which AI was used for any given comic drama by the lines, color temperature, and clarity.
Not to mention that AI itself is evolving rapidly, with new features appearing every week and workflows updating monthly. For instance, the brand-new Seedance model that emerged just when we were interviewing: “We used to be especially afraid of running shots, but now with Seedance’s update, after testing, running shots are basically okay—they won’t suddenly break into a dance or have character consistency issues like before.”
This job made Xiao E realize the powerful capabilities of AI animation, and he is fully optimistic about its prospects. However, in terms of the current content format and market environment of AI comic dramas, he believes they are far from mature. The industry is full of low-price competition, homogenized copying, and chaotic “shoddy crews.”
In his daily work, he finds the scripts annoying and complains about his boss, but he has no complaints about AI.
“AI animation has a lot of potential, but using these models to make comic dramas is like using a legendary sword to chop vegetables.”
03 A transitional form of AIGC, or a replacement for traditional animation?
In Xiao Y’s view, at least before the Spring Festival, mainstream AI comic dramas were still an extension of dynamic comics. AI technology enables low-cost, high-output production, quickly learning and iterating popular market elements. Combined with the paid promotion tactics of short-video platforms, they achieved astonishing traffic/revenue numbers in a short time.
She believes the quality of AI comic dramas already far exceeds mainstream dynamic comics, but many clients still assume AI comic dramas are a relatively cheap content format. In the future, as AI generation becomes more refined, “AI comic dramas” will gradually evolve toward “AI animation.”
But at least for now, the production philosophy of AI comic dramas is fundamentally different from traditional animation.
Xiao E believes there is still a “wall” between the comic drama and animation industries; people like him, who studied animation and now make comic dramas, are a tiny minority. The “regular army” of the animation industry hasn’t yet entered the AI fray. Truly skilled animation directors and teams are still guarding their own turf.
He also describes his team’s division of labor, including a director, image generators, video generators, and editors—superficially corresponding to pre-production, production, and post-production in animation. But compared to traditional animation, where pre-production is done thoroughly before meticulous execution, the weight of AI comic drama is almost reversed.
In his team, the director handles overall control and key filtering. The workflow for generating images and videos is “mindless.” But the editor is crucial, closer to an “executive director.”
“AI comic dramas now resemble the early film industry: the director shoots a ton of footage with the crew, then hands it to the editor to piece together a story.” You throw in reference images and prompts, never sure what shots will come out, but like actors improvising, it might yield surprisingly stunning results.
Xiao Y also mentions that when using AI, she adopts a “big client mentality”: giving short, vague requirements and forcing AI to repeatedly redo them. If the prompts are too specific, AI responds poorly.
“Take Jimeng for example: I think whatever you give it should be no more than 100 characters, or even less than 50, for the best response.”
But that doesn’t mean human operation is unimportant. On the contrary, both interviewees note that “AI animation doesn’t require effort” is a common misconception among clients and bosses.
“Bosses think productivity has skyrocketed thanks to AI, but that’s only because there are people directing it,” Xiao Y said. “It’s not because I know how to use AI, but because I already have images in my mind. AI is an ‘extreme brush’—it replaces the ‘hands’ and ‘feet’ of animation production, but the ‘brain’ is something AI cannot replace for now.”
“Having images in your mind” refers precisely to professional animation directing and writing skills.
In the teams of Xiao Y and Xiao E, they are almost the only employees with formal animation education. They no longer draw by hand in their daily work. Yet both still believe their animation education is extremely important, especially the training in aesthetic sensitivity and audiovisual language that continues to benefit them greatly.
Xiao Y said, “You’ll find that people who can independently produce a finished film must have basic storyboarding knowledge. As a director, when you read an image or text, images should already be ‘playing out’ in your mind—but most people don’t have that mindset.”
In her view, limiting animation expertise to “drawing great keyframes and inbetweens” is too narrow. The long-term training in aesthetic judgment and systematic theoretical knowledge allows her to better judge what constitutes good performance, shots, and pacing in her work.
As for the rapid development of AI technology, both interviewees agree that while professionals need to stay updated on AI advances, AI tools themselves are always evolving toward “easier to use.” For those with an animation background, they can be mastered quickly without much of a learning curve.
Most anxiety stems from the fear of “working in isolation” and falling behind the times.
04 Conclusion
Although they work full-time on AI comic dramas, neither interviewee has developed a habit of consuming AI comic dramas for leisure. They only watch the top-ranked works to understand market needs.
In their spare time, Xiao Y prefers slow-paced artistic works. She maintains an open and tolerant attitude toward AI comic dramas, but most current works, which emphasize fast-paced, high-intensity thrills, make her feel anxious and restless. She is also not the target audience for mainstream genres like cultivation, rebirth, or weird stories. Moreover, the current narratives in AI comic dramas are highly homogenized, with many scripts lacking a “human touch,” which she finds inferior to even some “abstractly bad” live-action short dramas.
Xiao E notes that current AI comic dramas have not really entered the viewership of “animation audiences”; they primarily attract the lower-tier market of short drama viewers. He believes the true charm of animation lies in its ability to create “stickiness” through narrative, character design, and world-building—exactly what AI comic dramas currently lack.
However, when AI allows most people to generate animation imagery above the standard line, lowering the barriers for both capital-funded projects and personal creations, Xiao E thinks AI might bring Chinese animation to a stage where it truly competes on creativity and storytelling, similar to how early manga artists like Osamu Tezuka formed a “one-man army” in the Japanese manga industry.
Thus, while Xiao E continues his work as a contracted director for AI comic dramas, he is also planning how to use AI to create his own works in the future. “As a creator, you should seriously think about the path ahead, rather than diving headfirst into comic dramas and crazily chasing trends—trends are endless; you need to let the audience adapt to your rhythm.”