In the future, people will be astonished that animation is drawn frame by frame.

A feast of audiovisual magic at the end of 2025 — the 7th Feinaki Animation Week arrives as scheduled, held at Beijing·Intermediate Theater from 11.14 (Fri) to 11.17 (Mon). This edition once again gathers a batch of the finest and most cutting-edge independent animation works from around the world.

The screenings are divided into 16 sessions, with 70 animated shorts, 15 music video animations, and 14 commissioned works receiving jury favor and entering the competition unit. Meanwhile, senior curators from Poland, Japan, Italy, and other regions bring several special themed units, along with various offline exhibitions, workshops, and academic forums.

As a media outlet that has long reported on and participated in the Feinaki Animation Week, we continue to track the highlights and trends of independent animation. However, the most surprising change for me this time is the first inclusion of AI-related content in Feinaki Animation Week: this year, a special AI Animated Shorts screening unit has been established, free of charge with advance reservation; additionally, in the traditional GIF Call for Submissions (this year’s theme is “Illusion”), a section exclusively for AI-generated works has been opened for the first time.

In mainstream commercial animation, AI is undoubtedly advancing, attracting traffic and money. But what role will AI play in the field of independent animation?

01​

AI vs Non-AI: Divergent Creative Orientations​

To answer this question, it is necessary to first understand the current state of “non-AI” “traditional independent animation” this year.

Independent animation / auteur animation, which Feinaki Animation Week focuses on, differs from collectively produced commercial animation in that it does not pursue large-scale public screenings or commercial ambitions. Instead, it appears more often at international film festivals, often only publicly released after several years of submission cycles. These works are frequently completed by a single person, with highly personalized creative themes and styles. Therefore, they represent alternative and avant-garde explorations in animation technique and narrative.

However, although Isao Takahata likens independent animation to “something like haiku,” the festival ecosystem is not so aloof: some animation styles seep into popular culture via forms like music videos or advertisements; some animators collaborate with capital to enter a niche of “stylization.” On the other hand, the personal expression and creative methods of independent animation also allow creators to capture reality and resonate with audiences more promptly and keenly.

This makes the works of this Feinaki Animation Week still gather a very diverse range of themes and expressions, from intergenerational conflict and identity to natural philosophy and sweet romance, from philosophical and minimalist to tragic and whimsical.

When I spoke with Luo Sijia, one of the preliminary jurors for the Feinaki Animation Week, she also explicitly mentioned the trends she felt and the emotions and issues she captured during this year’s screening process.

“You can tell that motifs/images of electronic devices (computers, phones) and social media appear frequently, at a higher proportion than I’ve seen at other animation festivals. I think that might just reflect the times…”

“…(In such works) there is often a person, with nothing else to do, lying on the sofa and naturally picking up their phone to scroll—this appears when depicting the daily barrenness of a person; or alternatively, it directly portrays the anxiety social media brings to a person.”

Among domestic student works, the theme of “involution” is particularly prominent. Many animations set stories in workspaces like factories and offices, tending to depict them as a dystopian scenario.

However, she particularly recommends two quite bizarre works: “The Taste of Beer” and “On Weary Wings Go By.” The former tells a common parent-child relationship in East Asian families with unique black humor and rhythm; the latter, set against Nordic natural scenery, is a winter poem using ceramic puppet animation.

Although not particularly fond of stop-motion form, in the current era of AI content proliferation, she indeed becomes more interested in animations made with real materials, because “AI-generated materials, though seemingly organic, are not truly organic.”

“Both works present a perspective on their surroundings and the world. Taking ‘On Weary Wings Go By’ as an example, perhaps more viewers see the ceramic puppets she made—but I see more of its background… the sun setting by the sea, the slowly melting snow. Incorporating natural phenomena into creation, I think that’s particularly good.”

“Perhaps from personal preference, I lean toward non-anthropocentric expression. So I’m attracted to pieces that feature natural objects, non-anthropomorphic or humanoid characters.”

Luo Sijia, along with four other preliminary jurors, watched 1,142 animated works from 69 countries and regions one by one, meticulously selected, and ultimately chose 108.

As an animator and the only former award winner among the jurors, Luo Sijia says she has no opinion on the trend of “shorts getting longer” among submissions, and she is very skeptical of measuring an animation’s value by “labor intensiveness.”

Luo Sijia cares more about the “unity” of animation expression—besides the completeness of the work, it’s also because she doesn’t want to see superficial aesthetic patchwork (especially in student works). When screening, she constantly asks herself: “Does the technique suit the theme? Why use this technique?”

When we apply the same standard to examine “AI animation within independent animation,” and watch the 16 works in the AI animation screening unit—6 from open call, 10 invited by the curatorial team—I roughly categorize them into “four types” and “two types of creators”:

Type 1: Artistic reflection. For example, “Sun Yiyang: Elephant Image,” depicting an elephant’s funeral and a ritual of image rebirth. Leaning toward philosophy, treating AI animation as an art installation, reflexively exploring the ontology of animation as concept-first.

Type 2: Music animation. For example, “Patterns of Life,” using an AI-synthesized perspective to showcase the exquisite organic forms of plants and the rhythm of life. The visuals pulse with the music, letting the images develop freely, with changes and transitions generating rhythm, emotion, and metaphor.

Type 3: Spectacle collage. For example, “Adult Swim ID: Organism,” displaying generated bizarre scenes, but with very weak coherence and spatial order, like a dream.

Type 4: Allegorical stories. For example, “City of Numbers and Floating Sea,” derived from Pu Songling’s “Strange Tales from a Chinese Studio” story “Rakshasa’s Sea Market,” telling a story set in the “Rakshasa Kingdom” where appearance judges people, beauty and ugliness reversed, and ugliness is prized. Story-driven, often with narration, mostly featuring dark Chinese-style legends or depictions of modern urban loneliness and emptiness, often using allegories like “the banality of evil,” “the trolley problem,” “The Truman Show,” or “brain in a vat.”

02​

What Is the “Authorship” of Animation?​

Luo Sijia herself has used AI to assist animation creation. In 2021, she produced the music video “Be Gone” for the band Blade Rain, telling the story of a small phone learning what humanity is between life and death, using VQGAN x CLIP + Google Colab + AI image upscaling. It later won international animation awards like Animest.

In 2025, she was invited to create the key visual animation for the 12th New Chitose Airport Animation Festival in Japan, using Lake, ripples, and snowflakes as core imagery, creating about 10 seconds of animation with Runway AI.

In these works, all core content—whether composition, color palette, character design, or character animation—is personally decided by Luo Sijia. She first completes the draft of shots and animation herself, then uses “image reference” to let AI generate materials and scenes according to instructions.

Taking the “small phone” in “Be Gone” as an example, the character design starts stroke by stroke from a paper sketch, with actions like walking and swinging drawn frame by frame by hand. After completing the storyboard draft, AI is introduced to generate assets. Through repeated “drawing cards,” the AI content is finally composited with hand-drawn content in animation software, with various details processed manually throughout.

Not averse to using AI for creation, her attitude is neutral, but she does not fully trust AI’s directorial ability and does not generate AI animations starting directly from prompts. In her view, the key to using AI assistance in creation lies in “reduction”: “AI can generate a single frame with extremely rich information in a second. So if you go along with it and greedily keep adding various things, you’ll be ‘led astray,’ and the so-called authorship or personal choice gets completely diluted.”

“In animation, authorship isn’t just about art style; I think motion itself is also part of authorship.”

She mentions: “A while ago, a junior contacted me, wanting my opinion… She said she wanted to use AI to generate ‘unseen landscapes,’ then trace over the generated animation with watercolor herself. I felt the logic was a bit off… The moving landscape generated by AI reflects AI’s understanding of motion. (I said at the time) that you could actually incorporate your own understanding of motion and collaborate with AI through a different workflow.”

Luo Sijia frankly says that most of the AI animations she saw during the preliminary round were not satisfying to her; the works screened at this Feinaki and the selected GIFs are already relatively high-quality samples.

She has also seen many popular AI animations online—and why do some, despite extremely high resolution, still feel lacking in “human flavor,” with a sense of “fragmentation” and “uncanny valley”? Why does the most cutting-edge technology still feel outdated?

One reason is that some creators are obsessed with using AI to imitate existing styles of other artists. Additionally, while AI-generated animation is convenient, it is not yet coherent enough, and some creators lack sensitivity to image quality (e.g., focal length choice, light distribution, camera rhythm).

On the other hand, Luo Sijia believes what truly makes animation unique are the subtle traces that run through an animator’s creation and even their growth:

“Maybe they have a hand habit, some small pen habits that cause everyone’s lines to differ; maybe during their growth, they prefer certain color combinations; or when drawing sketches or studies, drawing others’ faces ends up looking more like their own facial features.

Some conscious or unconscious habits, even small flaws—these are things that AI-generated content doesn’t embody.”

03​

Independent Animation in the AI Era​

While AI’s powerful capabilities are obvious, its impact on the existing order of the independent animation circle may not be immediately visible. After all, independent animation creators and festival events do not operate with commercial goals like “efficiency” or “output.”

Luo Sijia says that many independent animation creators like herself do not use AI because they lack technical skill. The group of independent animators experimenting with AI treats it more as exploring new artistic languages, rather than an everyday tool readily adopted.

Additionally, she has stopped expecting AI animation to develop its own performance style beyond imitation, nor does she feel the urge to do “reverse identification” (learning the specific production method), “because I think AI tools change rapidly, and (AI creation) always gives me a sense of ‘staking a claim.’ I feel why should I spend my energy to distinguish who has staked which claim?”

The deeper impact might occur within the creators’ psychology.

Recently, Feinaki Animation Week conducted an AI-related survey among past selected and invited creators, with 78 animators participating and sharing their views.

Regarding “attitude towards AI,” about 60% of respondents hold an embracing or curious attitude, while 40% choose to wait and see, feel conflicted, or resist.

On whether they have used AI-assisted creation, nearly half said “tried but not proficient,” about 20% said “have completed works with AI.”

And on the question “Have you seen impressive AI works?” nearly half gave negative or lukewarm answers.

An anonymous animator expressed concern in the survey: “What I fear most is the impact of lowered creative thresholds and content oversaturation—ethical dilemmas, copyright issues, and an existential crisis: is our creation just ‘content’ in this internet era… Making animation isn’t just about producing a valuable work; the process itself is a form of healing.”

Animator Mo also has similar thoughts: “I enjoy the process of making and creation. Every wrong line, every moment of frustration and pain during creative block is an indispensable human part of creation. I don’t want to skip this step to chase results… Not to mention that (generative AI) is essentially plagiarism and theft of human artists’ creative output, and its generation results are infused with systemic bias and stereotypes.”

In animator gnitiz’s view, the AI vs. human win-loss is meaningless. “Works without any ‘human’ background, especially those that can be mass-produced quickly, are actually meaningless… For creation and appreciating creation, people expect things from ‘humans.’ The problem AI brings to animation (and other humanities fields) will no longer be ‘will it replace humans,’ but ‘what is more moving.’”

Animator Li Zehao believes: “AI is a chaotic ocean, floating with fragments left by human production. So, like playing mahjong, the more professional you are, the better you can see order within chaos and integrate it, appreciate it… It will become increasingly difficult to tell from appearance whether something is a photo, hand-drawn, or AI-generated. We need to look beyond surface to extract and appreciate the author’s expression and personality.”

But the creators’ feedback still cannot fully answer a lingering question Luo Sijia posed to me during our conversation: “Is it important to many people whether they know they’re watching AI-generated animation?”

Looking at the broader social context, AI animation is rapidly penetrating fields like advertising, short videos, and film, driving leaps in productivity. On the other hand, it is also lightly skipping over part of the copyright system and artistic ethics established since modern times. It brings more people access to the fresh integration of technology and art; it also forces people to re-examine the meaning of creation, accelerating the divergence of artistic concepts.

But at the very least, we should not be numb and slumped, controlled by feelings of being catered to or fear—whether we are animators, festival organizers, media, or audiences. In the face of this wave, we need to respond to our inner true feelings and reaffirm the value of animation through repeated discussion and viewing, selection and exclusion.