Recently, I attended a film festival focused on shorts created with generative AI, known as AIFF 2025. The event showcased a unique intersection of technology and creativity.
Runway, the company behind the festival, specializes in developing models and tools for generating images and videos. Throughout the event, a selected group of industry experts advocated for Hollywood to welcome AI technology. However, my private discussions with various professionals revealed a growing philosophical divide within the film and television sector regarding the future role of AI.
In an interview with Runway CEO Cristóbal Valenzuela, we explored the delicate position he occupies when promoting his products to a profession rife with mixed sentiments about AI’s implications for the industry.
To understand this dynamic, it is crucial to first examine the festival’s films, particularly the top prize-winning short, which encapsulates the festival’s central issues.
A Gathering of Unique Perspectives
This year marked the festival’s inaugural public opening, drawing a varied audience of AI enthusiasts, industry creators, and movie lovers, many of whom identifier with multiple labels.
The showcased films, predominantly shorts, were more akin to art film festivals than mainstream cinema. Their aesthetics ranged from animated styles, including anime influences, to live-action presentations. Some served as quasi-documentaries. These films either utilized Runway’s AI tools exclusively or incorporated them alongside traditional filmmaking techniques.
Many of the shorts embraced unconventionality, taking advantage of AI’s knack for bizarre and distorted visuals. Several creators leaned deliberately into this characteristic, transforming perceived limitations into unique strengths.
An illustrative example is Vallée Duhamel’s Fragments of Nowhere, which visually navigates the concept of interwoven dimensions, featuring cars transitioning into houses and humanoid figures taking on strange forms. While visually captivating, this film echoed elements seen in dreamcore and horror AI TikTok creators such as GLUMLOT and SinRostroz.
More impactful were shorts that leveraged oddity to convey deeper themes of human experience and identity. Notably, More Tears than Harm by Herinarivo Rakotomanana—a rotoscoped sensory collage exploring childhood memories from Madagascar—captured authenticity through its specificity and cohesive style, unlike Fragments of Nowhere. I also appreciated Riccardo Fusetti’s short, Editorial.
Notably, the two standout films in the festival garnered the Grand Prix and Gold prizes. The judging panel featured filmmakers and industry leaders, including Gaspar Noé, Harmony Korine, and several executives from major studios.
The runner-up, Jailbird, directed by Andrew Salter, was a brief documentary introducing an innovative UK program wherein chickens serve as companion animals in corrections facilities, yielding positive effects. AI’s role was essential for achieving shots that otherwise would have been unattainable on a small budget, much to the audience’s delight.
The Grand Prix winner, Jacob Adler’s Total Pixel Space, also presented a compelling case for the validity of AI art. Available for viewing on YouTube, this film resonated with me, prompting mixed feelings as I recognized its selection as the festival’s top entry with a hint of skepticism, despite its clear appeal.
Exploring Total Pixel Space
Though it risked self-congratulatory themes, Total Pixel Space effectively paired striking visuals with thought-provoking concepts, sometimes approaching profoundness in its exploration.
The film’s narrator posed the question, “How many images can possibly exist?” and elaborated on the idea of “total pixel space,” explaining how image generation functions:
Pixels, the fundamental components of digital images, are defined by coordinates of color and position. Thus, any digital image can be represented numerically. Much like we don’t need to enumerate every number to validate their existence, it’s unnecessary to generate every conceivable image; their existence is intrinsic to the underlying mathematics. Each potential frame symbolizes coordinates, and to deny their existence is akin to denying numerical reality.
The film posits that the total number of potential images surpasses the atom count in the universe, suggesting that artists discover images already present in possibility rather than creating anew. Within this framework, it follows that generative AI simply offers artists another avenue to “discover” rather than fabricate.
Navigating Complexities
“We’re all—and I include myself in that group—fascinated by technology, constantly discussing models, datasets, and capabilities,” Valenzuela remarked during our follow-up conversation. “However, looking back, the festival celebrated the creativity of filmmakers and artists.”
I shared my emotional connection to Total Pixel Space. “The winner never considered himself a filmmaker, yet he created a film that resonated,” Valenzuela reflected. “That is powerful, enabled by access to innovative tools that were unthinkable just months prior.”
The AIFF 2025 festival spotlighted first-time and independent filmmakers, yet Runway also collaborates with established studios, leading to inherent tensions.
Valenzuela detailed their operational approach: “We embed deeply within organizations, collaborating directly with teams. We host internal film festival sessions to help them navigate the creative process and recognize potential.” This unique strategy has distinguished Runway from competitors like OpenAI and contributed to its competitive edge.
Founded in 2018 by a trio of innovators at New York University’s Tisch School of the Arts, Runway has become a key player in the realm of video generation. It played a fundamental role in developing the widely used Stable Diffusion model.
Despite being outspent by rivals like OpenAI, Runway maintains a proactive, hands-on strategy, focusing on addressing real issues faced by creative professionals rather than lofty aspirations regarding artificial general intelligence.
Yet, the company is not immune to conflict. Ongoing legal disputes over intellectual property rights highlight the complex aspects of its operations. While Valenzuela has defended Runway’s practices, recent evidence indicates questions surrounding the training of its models, particularly regarding copyrighted content.
According to Valenzuela, studios prioritize liability concerns over ethical dilemmas, stating:
“Most copyright anxieties are related to outputs—how do we ensure models don’t produce pre-existing material? We ensure our tools promote creative guidance while respecting their context. We collaborate with every major studio, along with offering indemnification.”
He also differentiated Runway’s output from traditional asset re-creations, asserting that the generative process stands apart legally and creatively from merely extracting content from a database.
“Many view AI as a magical system capable of conjuring images without user input. That’s misleading. User engagement and responsibility remain paramount.”
Although his belief in AI as a legitimate creative tool seems steadfast, Valenzuela acknowledges a significant divide in the sentiment among industry professionals regarding its merits and implications.
Hollywood’s Tenuous Divide
While in Los Angeles for the festival, I engaged with two long-time friends employed in film and television. Upon sharing my attendance at an AI-focused festival, one expressed disgust, while the other brimmed with enthusiasm, keen to discuss how he utilizes AI in editing processes at his company, lamenting the sluggish pace of adoption there.
This dichotomy reflects wider industry sentiments. Hollywood’s historical adaptability to technological change is not a new phenomenon.
Prior transitions—from silent to sound cinema and film to digital projection—resulted in significant upheaval, with some careers evolving or collapsing in the face of new methods. Opinions were divided: some heralded each change as revolutionary, while others resisted, asserting artistry was at stake. Bruce Markoe from IMAX shared insights into these transformations at a festival mixer, acknowledging the dread and skepticism that often accompany such societal shifts.
As Markoe explained, “The motion picture and television industry has consistently remained industry leaders, leveraging advancements to enhance their craft and operational efficiency. I believe some filmmakers will embrace AI tools sooner than others, especially for pre-visualization, although widespread comfort will take time.”
Despite these positive outlooks, skepticism remains. Director Mike Rianda articulated a common fear: “AI holds promise, but in corporate hands, it poses substantial risks.” Others express concerns over the rapid integration of technologies into production environments without adequate forethought or safeguards.
Alegre Rodriguez, an editor from the Motion Picture Editors Guild, added a note of caution, stating, “AI lacks the nuanced understanding of human emotion and storytelling that defines great narratives.”
A Moment of Reflection
Flight back from Los Angeles provided an opportunity to distill key differences between the current generative AI disruption in Hollywood and previous technology transitions.
Firstly, neither the silent/talkie nor film/digital transitions raised existential concerns over intellectual property rights. While studio heads prioritize liability protection, creatives worry about the integrity of their contributions, advocating for acknowledgment and compensation for their influence on training datasets. Ongoing legal battles underscore this pressing issue.
Secondly, the cultural and philosophical dimensions of AI’s impact on creativity pose additional challenges. Valenzuela noted this collective concern: “AI acts as a mirror. It reflects our fears, hopes, and expectations regarding the future.”
In online discussions, you often encounter extremes: some portray technological advancements as miraculous solutions to society’s ills, while others depict AI as a major ethical threat. Amidst such polarized narratives, the film festival illustrated realities on the ground. New filmmakers are now gaining visibility and recognition in the industry thanks to accessible technology that allows for high-quality output without extensive resources.
“People overlook the substantial positive potential of this technology,” Valenzuela remarked, pointing out the efficient pathways it opens for creatives.
Yet, there’s genuine apprehension among many creatives who feel their foundational work fuels these technologies without due credit. As the industry grapples with this disruption, the push for legal frameworks designed to safeguard human interests continues to gain momentum, exemplified by legislative initiatives like the Generative AI Copyright Disclosure Act.
Settling Philosophical Questions Legally
The festival-winning film posited that “total pixel space represents both ultimate determinism and ultimate freedom—every possibility coexists, awaiting meaning through consciousness.” This philosophical stance implies that true creativity is fundamentally about curation, suggesting originality may be a myth, with human expression distilled into mathematical constructs.
This notion resonates with many who believe that all possibilities exist, positioning artists as entities revealing previously latent images. Contrarily, others cherish the belief that unique creations gain value precisely because they emerge from the individual’s creative process.
While these discussions delve into profound philosophical realms, the resolution ultimately hinges on legal frameworks, contracts, and compensation mechanisms—elements that have always played a crucial role in Hollywood’s evolution alongside artistic innovation.
