Aronofsky’s newly founded studio, named Primordial Soup, aims to be at the forefront of this next creative wave. Its debut project, a short film titled Ancestra, directed by Eliza McNitt, is set to premiere at New York’s Tribeca Film Festival. The film blends live-action footage with surreal, likely AI-generated, visuals, showcasing how artificial intelligence can merge with traditional filmmaking in new ways. Although currently limited to a short format, the project hints at broader ambitions to scale up production.
Google’s newly introduced platforms—Veo 3 and Flow—are central to this shift. Veo 3 enables the generation of video content directly from text or image prompts, while Flow provides creators with advanced control over aspects like camera motion, editing, transitions, and scene structure. Flow integrates capabilities from Google’s chatbot Gemini, the Imagen image generator, and Veo, offering a comprehensive creative environment. This toolkit is designed to empower filmmakers by expanding their options beyond the constraints of conventional production methods.
Veo 3, in particular, sets a new benchmark in the AI video space. It produces more coherent sequences, with improved physics and spatial consistency, areas where earlier models often struggled. A major leap forward is its ability to produce synchronized sound, including ambient noise, object-based effects, and spoken dialogue. Examples already circulating online include AI-generated interview-style videos and short clips featuring characters speaking in varied accents or even singing—sometimes nonsensically, but often convincingly enough to fool a casual viewer.
Flow also includes a feature called Flow TV, a public library showcasing user-created films alongside the specific instructions used to make them. This transparency aims to foster community learning and inspiration. Among the more notable examples is a short film called Freelancers, which highlights both the capabilities and limitations of current AI tools—smooth visual continuity and mood on one hand, and unstable facial features or spatial inconsistencies on the other. Similar flaws have been observed in other systems like OpenAI’s Sora.
Currently, Veo 3 and Flow are only available in the United States through Google’s AI Pro and AI Ultra subscription plans. In other regions, including the Czech Republic, users can experiment with a more basic version, Veo 2, which generates silent, 8-second clips based on text or image input. After an initial free period, subscription fees apply.
While these technologies are not yet mature enough to replace traditional filmmaking—except in niche projects like Ancestra—rapid progress suggests that this might soon change. Hollywood unions and guilds have already included AI provisions in their contracts, signaling the seriousness with which the industry views this transition. Aronofsky, among others, sees AI not as a substitute for human creativity, but as a tool to expand artistic freedom and streamline complex workflows.
This perspective is shared by other prominent figures in the film industry, including Avatar director James Cameron, who is now part of the advisory board at Stability AI, the company behind the image generator Stable Diffusion. He sees AI as essential for reducing production costs on CGI-heavy films without compromising quality. Rather than eliminating jobs, he believes AI will allow visual effects professionals to focus more on artistic and creative decisions—areas where human input will remain indispensable for the foreseeable future.
