AI-Generated Videos: The Future of Entertainment or a Threat to Hollywood?
In today’s digital age, encountering AI-generated videos has become almost unavoidable. Scroll through any social media platform, and you’re likely to come across uncanny clips depicting fake natural disasters or animals performing impossible feats. While many of these videos are rough around the edges, they often garner thousands of likes and comments, with many enthusiasts claiming that AI-driven content represents a revolutionary art form poised to transform the entertainment industry.
The Rise of Realistic AI Clips and Industry Concerns
Artificial intelligence has advanced to produce increasingly realistic images and videos. Despite their imperfections, these clips are often hailed by supporters as a glimpse into the future of filmmaking, with some even warning Hollywood to be wary. The idea that AI-generated video could fundamentally alter or even threaten traditional cinema has gained significant traction among tech advocates. However, skepticism remains—mainly because many AI outputs simply aren’t suitable for professional-grade movies or TV series.
Changing the Narrative with Ethical AI
Filmmaker Bryn Mooser aims to challenge this narrative with Asteria, a new production studio he launched last year. Asteria is developing a groundbreaking AI-generated feature film in collaboration with Natasha Lyonne, Mooser’s partner and an advisor at Late Night Labs—a studio focused on generative AI that Mooser’s company, XTR, acquired last year. What sets Asteria apart? Its commitment to “ethical” AI. Unlike other generative models trained on unlicensed or questionable material, Asteria’s model, created in partnership with Moonvalley, is built exclusively on properly licensed content.
This emphasis on ethical AI is especially timely, following high-profile lawsuits from Disney and Universal against Midjourney over copyright infringement. As the industry navigates these legal challenges, ethical AI could become a key factor in how generative technology is adopted more broadly in entertainment.
The Challenges of AI in Filmmaking
Mooser emphasizes that understanding what generative AI can and cannot do is crucial. “When we thought about building Asteria, it was clear that Hollywood’s AI tools weren’t designed by people with filmmaking experience,” he explains. “The idea of simply typing ‘Make me a new Star Wars’ and getting a complete film is a fantasy. The tools available are not yet capable of delivering the control filmmakers need at the pixel level.”
To address this, Asteria uses its core model, Marey, to develop project-specific models trained on original visual assets. For example, in producing musician Cuco’s animated short “A Love Letter to LA,” Asteria trained Marey on 60 illustrations by artist Paul Flores. This allowed the studio to generate diverse assets in a consistent style and build a stylized fictional world. While impressive, the project highlights current limitations—the technology still struggles with fine control over every pixel, especially for highly detailed work.
Ownership and Cost Savings
Mooser notes that depending on contractual arrangements, filmmakers could retain partial ownership of AI models after a project’s completion. Asteria also pays licensing fees for training material and is exploring revenue-sharing models. The primary appeal, however, lies in reducing production costs. “If you’re making a Pixar film, creators often don’t own what they produce or share in profits,” Mooser says. “Using AI to cut costs could enable independent filmmakers to finance projects that were previously out of reach.”
Upcoming Projects: Testing AI’s Creative Limits
Asteria’s ambitions include testing AI’s potential with “Uncanny Valley,” a feature film co-written and directed by Natasha Lyonne. The story follows a teenage girl whose perception of reality becomes increasingly surreal, with many visual effects created through Asteria’s models. Mooser envisions these stylized, sometimes hallucinatory visuals as a deliberate aesthetic choice, embracing AI’s quirks rather than hiding them. His goal is for audiences to focus on the human touch—“nobody wants to just watch a computer’s dreamscape,” he affirms.
Democratizing Art and the Future of Filmmaking
Mooser believes generative AI can democratize art creation, making filmmaking more accessible. He envisions a future where movies could be produced for $10–20 million instead of hundreds of millions, opening doors for emerging filmmakers. Smaller teams could work more efficiently, collaborating closely with key creative leads without the extensive revisions typical of larger productions. However, this shift also raises concerns about job displacement.
Industry Transformation and Workforce Implications
While AI promises efficiency, Mooser acknowledges the potential impact on employment. The recent collapse of VFX studios like Technicolor underscores ongoing industry upheaval, with many workers facing layoffs even before AI hype took hold. He stresses that many professionals—art directors, cinematographers, writers, actors—can pivot to new roles centered around AI if they adapt to the changing landscape.
A Cautionary Future
What’s most concerning isn’t necessarily AI replacing the entire studio system but its ability to streamline production with fewer people. If AI workflows become standard, it could challenge the industry’s employment model and the livelihoods of countless creatives. Nonetheless, Mooser remains optimistic, emphasizing the importance of understanding both the opportunities and risks of AI in entertainment.
In the end, the future of AI in filmmaking hinges on how well creators and studios can harness this technology responsibly—balancing innovation with ethical considerations—to shape a new era of storytelling.
