Directing the future: craft and creativity in the age of AI

Key Concepts

• AI-Driven Filmmaking: The use of generative AI to create, augment, and accelerate the production of feature-length films and interactive experiences.
• Agentic Characters: AI-powered characters capable of natural language interaction, maintaining context, and participating in dynamic, non-linear storytelling.
• Immersive/XR Storytelling: The integration of Extended Reality (XR) headsets and 180-degree stereoscopic video to place the audience inside the narrative.
• Agile Production Pipeline: A methodology borrowed from software development where the traditional, rigid stages of film production (pre-production, production, post-production) are blurred to allow for constant iteration and technological adaptation.
• Performance Capture: The process of recording the movements and performances of actors to be processed and integrated into AI-generated environments.

---

1. Main Topics and Key Points

• The Evolution of Filmmaking: Doug Liman emphasizes that art has always reflected the technology of its time. From using high-speed film stock in Swingers to utilizing AI for his latest film, Bitcoin, Liman views technology as a tool to empower storytelling rather than replace it.
• 30 Ninjas Studio: Founded by Doug Liman, Jed Weintrob, and Julina Tatlock, this studio focuses on two main pillars:
  • Interactive Story Extensions: Using generative AI to allow audiences to interact with characters from linear films.
  • AI Film Production: Utilizing AI to handle complex visual effects and environment generation, significantly reducing post-production timelines (e.g., from a projected 40 years to 6 months).
• The "Bitcoin" Project: A feature film starring Casey Affleck, Gal Gadot, and Pete Davidson. The film explores the world of cryptocurrency and uses AI to enable the director to shoot in 150 global locations—a feat impossible under traditional budget and logistical constraints.

2. Important Examples and Real-World Applications

• "ASTEROID": An immersive film for the Galaxy XR headset. After the linear film ends, the viewer assumes a role to rescue a character (played by DK Metcalf) using Gemini Live API for real-time, natural language interaction.
• "Roadhouse" vs. "ASTEROID": Liman discusses the creative tension of "killing" characters. In Roadhouse, the death was essential to the character arc; in ASTEROID, the audience's desire to save the character led to the creation of an interactive, AI-driven "sequel" chapter.

3. Methodologies and Frameworks

• The "Artist-First" Approach: The team emphasizes that AI is used to "supercharge" human creativity. They maintain a human-centric workflow where writers, editors, and actors provide the core creative vision, while AI handles the technical heavy lifting.
• Agile Filmmaking: By treating film production like software development, the team iterates on scenes in real-time. If a scene isn't working, they rewrite and reshoot, blurring the lines between production and post-production.
• Guardrails for AI Characters: To ensure AI characters remain consistent with their scripted personas, the team uses extensive training and "guardrails" to prevent the AI from breaking character or providing inappropriate responses.

4. Key Arguments and Perspectives

• Humanity in AI: Liman argues that AI should not be used to "cheat" or replace the writer, but to amplify the filmmaker's unique voice. He warns against "AI slop" and advocates for using tools to bring out human nuance.
• Job Creation vs. Displacement: While acknowledging industry anxiety, the founders argue that their startup model creates new roles (e.g., AI artists, interactive narrative designers) and provides work for talented professionals in a shrinking traditional market.

5. Notable Quotes

• Doug Liman: "I don't know how to do this, but God knows I'm going to try because I'm going to have to face an audience when it's all done." (On the value of embracing the unknown in filmmaking).
• Doug Liman: "Figure out a way how to use these tools to bring out your humanity, but not to cheat and just let it do it for you."

6. Data and Research Findings

• Production Efficiency: The team successfully reduced the post-production timeline for an AI-heavy feature film from a projected 40 years (at the start of the project) to 6 months.
• Scale: The Bitcoin project utilized 150 locations and a cast of 125–140 actors, significantly exceeding the scale of traditional films like The Bourne Identity (which had 50 actors).

7. Synthesis and Conclusion

The presentation highlights a paradigm shift in cinema where the "box" of traditional filmmaking—defined by budget, location, and linear structure—is being dismantled by AI. By combining the expertise of traditional Hollywood storytellers with the agility of software developers, 30 Ninjas is pioneering a new form of "living" media. The main takeaway is that the future of film lies in the synthesis of human-driven narrative arcs and AI-powered interactivity, provided that creators use these tools to enhance, rather than automate, their humanity.