Natasha Lyonne Ventures into AI Filmmaking with 'Uncanny Valley'

The actress and director Natasha Lyonne collaborates with virtual reality pioneer Jaron Lanier and co-writer Brit Marling on a new sci-fi film utilizing generative AI, aiming to explore the intersection of technology and storytelling.

Peabody Awards, CC BY 2.0 <https://creativecommons.org/licenses/by/2.0>, via Wikimedia Commons

Natasha Lyonne has always marched to the beat of her own creative drum, and her latest endeavor only reinforces that reputation. The actor, writer, and director known for roles in Russian Doll and Poker Face is now diving into the frontier of artificial intelligence with a new sci-fi project called Uncanny Valley. The film, still in its early stages, will reportedly use generative AI not just thematically, but as part of its actual production process—an increasingly hot topic in entertainment circles.

Co-written with Brit Marling (The OA), Uncanny Valley follows a teenage girl who becomes deeply enmeshed in a virtual reality game that begins to bleed into her real-world experiences. Lyonne, who will also direct and star in the film, has partnered with VR pioneer Jaron Lanier to ensure the project isn’t just visually compelling, but also philosophically grounded. The story aims to explore questions of digital identity, consciousness, and what it means to be human in an age where AI and immersive tech are redefining boundaries.

The film is being produced under the banner of Asteria, a new AI-conscious production company co-founded by Lyonne and Bryn Mooser, the Emmy-nominated producer behind Body Team 12. Asteria is positioning itself at the intersection of entertainment and ethics, promising to use tools that are trained on properly licensed datasets—an increasingly important distinction in a legal landscape still catching up to rapid advancements in generative tech.

One of those tools is Marey, an AI system developed by Moonvalley, which is helping to craft visuals for the film. Marey is designed to work collaboratively with creatives rather than supplant them. According to Moonvalley, the tool was trained only on content that was licensed and paid for—an effort to address criticisms that many AI companies use scraped, copyrighted data without consent. This approach stands in contrast to many commercial AI products that have come under scrutiny from writers, artists, and unions over unauthorized training data.

Lyonne, never one to shy away from unconventional narratives, says the film is inspired by everything from The Matrix to the existential character studies of Charlie Kaufman. “We’re using AI to tell a story about AI, but with all the weirdness and human soul that movies are supposed to have,” she explained in a recent interview. “It’s not about showing off tech—it’s about using it to dig into something deeper.”

While Uncanny Valley doesn’t yet have a release date or distributor, industry insiders are watching closely. A project with this kind of talent—Lyonne, Marling, Lanier—could signal a new phase in AI cinema, one that isn’t defined by novelty or gimmick but by creative control and intention. It also arrives at a moment when debates over AI’s role in filmmaking are reaching a fever pitch. From deepfake casting concerns to the use of synthetic voices in animation, creators are wrestling with both the promise and the pitfalls of new tools.

Critically, the team behind Uncanny Valley isn’t just using AI for the sake of it. They’re trying to model what responsible integration can look like—an ethos Lyonne and her collaborators hope will resonate beyond this single project. Whether the film lands as a genre hit, a cult oddity, or a spark for broader change remains to be seen. But in a landscape often polarized between hype and alarm, Uncanny Valley feels like it’s carving out a third path: one where artists remain at the center of the conversation, even when the tools are algorithmic.