Blomkamp's AI Short Film Receives Harsh Criticism for Poor Quality at Seedance 2.0
Neill Blomkamp's 13-minute AI-generated sci-fi short "Nightborne," made with ByteDance's Seedance 2.0, was criticized as "slop" by The Verge. An analysis of AI filmmaking's current state and challenges.
Neill Blomkamp, the filmmaker known for “District 9” and “Gran Turismo,” released a new short film titled “Nightborne” on July 20, 2026. The 13-minute sci-fi short is the debut production from his newly established AI startup and production company, Barley Studios, and was entirely generated using ByteDance’s text-to-video generation model Seedance 2.0.
Blomkamp described the project as a “test start” on his X (formerly Twitter) account, expressing his ambition to create a full-length feature using this format in the future. However, the reception from critics has been harsh. Charles Pulliam-Moore of The Verge labeled the film as “slop warmed over,” adding that “no amount of clever editing can make the chaos of Seedance-generated footage entertaining.”
Limitations of Seedance 2.0
Seedance 2.0 is a text-to-video generation model developed by ByteDance and launched in late 2025. It generates a few seconds of video clips based on prompts, leveraging the vast video data accumulated through TikTok for training. The model competes with offerings like Sora (OpenAI), Gen-3 Alpha (Runway), and VideoPoet (Google).
Currently, Seedance 2.0 can only produce short clips of a few seconds at a time. For “Nightborne,” these clips were pieced together, using a documentary-style interview format and frequent cuts to create an illusion of temporal continuity. The editing, overseen by Austyn Daines, could not overcome the model’s inherent limitations.
Despite Blomkamp’s claim on X that the film serves as a “strong proof of concept,” the project underscores the fundamental issues with text-to-video generation. According to The Verge, the film exhibits common shortcomings of generative AI, such as inconsistent character expressions and movements, as well as a lack of physical coherence in objects and environments.
The Story and Features of “Nightborne”
The short film loosely draws inspiration from Peter Watts’s 2014 novel “Echopraxia.” The story is set against the backdrop of a military conflict between the United States and Syria. The protagonist, a female pilot, is critically injured after her helicopter is shot down, leaving only her head intact. This makes her the ideal candidate for “Project Nightborne,” a secret government program designed to weaponize brain-dead soldiers through remote control. Utilizing an endless supply of corpses, the military trains the protagonist to become one of the undead soldiers sent to the battlefield.
While the characters’ voices and faces are based on real human actors, all visuals were generated using Seedance 2.0. However, elements such as audio-visual synchronization, facial expressions, and interactions between characters and their environments fail to surpass the usual standards of AI-generated content.
Why Was It Called “Slop”?
Pulliam-Moore of The Verge described “Nightborne” as essentially a collection of very short video clips, similar to other AI-generated films. Since Seedance 2.0 can only generate a few seconds of footage at a time, creating a 13-minute film required extensive editing to fill in the gaps.
This method mirrors the “workaround strategies” often seen in AI-generated fake videos or deepfake content. By increasing the frequency of cuts, changing camera angles often, and incorporating monologues or interview-style dialogue, the film attempts to mask the model’s inability to generate long, coherent scenes. However, the result is a disjointed viewing experience that lacks immersion.
Although Blomkamp has expressed interest in tackling full-length films in the future, the current capabilities of Seedance 2.0 make it technically infeasible to generate a coherent narrative lasting over 60 minutes. Extensive editing and post-production remain necessary, ultimately relying heavily on traditional filmmaking techniques.
The State of AI Filmmaking
AI video generation technologies have advanced rapidly in recent years. OpenAI’s Sora demonstrated the ability to generate realistic videos up to one minute long in 2024. Runway’s Gen-3 Alpha achieved more controllable video generation, while competitors like Pika and Stability AI’s Stable Video Diffusion also made strides. However, none of these models have yet achieved the consistency, storytelling depth, or emotional continuity required for feature-length films.
Instances of established filmmakers like Blomkamp experimenting with AI-generated filmmaking are increasing. In 2025, an AI-generated short film remake of “Terminator 2” gained attention, and by 2026, animation studios began incorporating AI into their workflows. However, most of these projects remain in the realm of “technical demonstrations” and have yet to be embraced by critics or audiences as fully-fledged cinematic experiences.
The founding of Barley Studios is notable as one of the first studios dedicated to AI-generated production. Blomkamp, known for his focus on special effects and VFX, demonstrated innovative visuals on a low budget with “District 9.” His commitment to AI video generation sends a strong signal to the industry.
Editorial Opinion
In the short term, the release of “Nightborne” highlights the current state of AI video generation technology. While models like Seedance 2.0 have practical applications in advertising, social media, and short-form content, they face fundamental limitations when it comes to long-form storytelling like feature films. Blomkamp’s experiment is an intriguing case study in how professional filmmakers might utilize AI tools, but the harsh reception—being labeled “slop”—may reinforce skepticism about the potential of AI-driven filmmaking.
From a long-term perspective, as generation capabilities improve and editing workflows advance, AI video generation could become a valuable tool in certain aspects of filmmaking. For instance, generating background visuals, creating crowd scenes, and streamlining post-production could become cost-effective alternatives to traditional VFX methods. However, as long as models like Seedance 2.0 remain confined to generating only a few seconds of footage at a time, AI will struggle to take center stage as a storytelling medium. The editorial team recognizes the ambition behind Blomkamp’s declaration to pursue full-length features as a noteworthy endeavor.
References
- “Neill Blomkamp’s new zombie AI ‘film’ is just slop warmed over”, by Charles Pulliam-Moore — The Verge, 2026-07-21T22:06:43.000Z (ARR)
- Source URL: https://www.theverge.com/entertainment/968703/neill-blomkamps-nightborne-barley-studios-seedance
Frequently Asked Questions
- What is Seedance 2.0?
- Seedance 2.0 is a text-to-video AI model developed by ByteDance. It generates a few seconds of video clips based on prompts and was trained using extensive video data from TikTok. It competes with models like Sora and Runway Gen-3. Currently, it struggles to produce long, consistent videos, requiring extensive editing to assemble cohesive content.
- Why is Neill Blomkamp exploring AI filmmaking?
- Known for his focus on special effects and innovative visuals, Blomkamp founded Barley Studios to experiment with AI-driven production methods. He sees potential in reducing costs and production time and released "Nightborne" as an experimental step toward full-length features. However, the reception has been largely critical.
- Can AI video generation become mainstream in filmmaking?
- Not yet. Current models face challenges in producing consistent, emotionally engaging, and physically coherent long-form narratives. However, AI is gaining traction in short-form content like advertising and music videos. Significant advancements in generation capabilities and editing techniques will be required for AI to play a major role in feature filmmaking.
Comments