Darren Aronofsky Partners with Google DeepMind on Generative AI Short Film Initiative

Darren Aronofsky, the Oscar-nominated director known for Black Swan and The Whale, has launched a new storytelling venture called Primordial Soup in collaboration with Google’s DeepMind Veo 3. The studio is focused on exploring the intersection of narrative film and emerging technologies, specifically generative AI.

Primordial Soup’s first public initiative is a partnership with Google DeepMind, the AI research division behind some of the most powerful machine learning models in development. Together, they’ve created a series of experimental short films using DeepMind’s new video generation model, Veo 3.

This collaboration was introduced in a short announcement film released on DeepMind’s official YouTube channel. The video features Aronofsky and other filmmakers discussing the creative potential and implications of using AI as a tool for visual storytelling.

The project’s centerpiece is a short film titled “Ancestra”, directed by Eliza McNitt. It blends live-action performances with visuals generated by Veo, and it will premiere at the 2025 Tribeca Festival. It is the first of several planned shorts intended to explore how AI might shape cinematic language.

The trailer includes surreal, dreamlike sequences that blur the line between traditional cinematography and AI output. One of the most visually striking moments mirrors the iconic Stargate sequence from 2001: A Space Odyssey. The original took months of meticulous optical work and hand-painted art to complete. In contrast, this AI-rendered version was generated in seconds. The similarity could be revealing.

What is DeepMind’s Veo 3?

Google DeepMind’s video model, Veo 3, is the core technology behind this collaboration. Although in limited early access, Veo 3 is designed to generate coherent, high-quality video based on text prompts or image inputs.

Key attributes of Veo 3:

• Text-to-video generation
• The ability to replicate specific visual styles
• Control over cinematic movement and framing
• The use of reference images for tone and mood consistency
• Longer form coherence, not just short clips

These features allow creators to shape complex visual ideas without the time, cost, or technical expertise typically required in traditional production workflows.

This AI-generated video called “Sailor and the sea” is extremely convincing. It also seems to be a copy of something from (or just similar to) “The Old Man and the Sea” by Ernest Hemingway.

In short, this tool gives users the power to create polished, stylized content almost instantly.

For more information about Veo, visit their official website.

Generative AI and originality

In the trailer for Ancestra, there’s a moment that feels like a clear nod to 2001: A Space Odyssey. It’s visually stunning, but also a little uncanny—like something you’ve seen before, just reassembled in a new way.

That’s the tension at the heart of using AI in filmmaking. These tools can generate beautiful images and open up exciting possibilities, but they also raise questions. Where’s the line between inspiration and imitation? And what happens to creative voice when the machine starts making the choices?

There’s a lot to explore here, and no clear answers yet. But as AI becomes a bigger part of the process, these are the kinds of conversations worth having.

For a deeper look at how generative AI is reshaping the film and video landscape, including concerns from within the industry, check out CineD’s in-depth analysis: “Generative AI for Video: Have We Already Lost the Battle, or Is There Room to Fight?”

What are your thoughts on using generative AI to help create films? It’s a big shift in how stories are made, and we’re curious how people feel about it.