# Film as Executable Text: AI as Director, Not Painter

**The prevailing trend today is to ask AI to “paint” an entire video from scratch out of statistical imagination. In a hands-on engineering experiment, my team took the opposite road: an AI director that writes every editorial decision as inspectable text, an execution agent that assembles real footage, and an uncompromising automated verifier—with a human signing off at the end.**

Emad Jumaah

---

Whenever video and artificial intelligence are mentioned together today, people immediately think of generative video: type a two-line prompt, and a diffusion model conjures a moving scene out of thin air. The result is dazzling for fleeting social clips, but anyone who has tried to use generative video in serious documentary or commercial production knows its steep price. You have no control over fine details: ask to tweak one spoken word or adjust a camera angle, and the model regenerates the entire shot from scratch, handing you a completely different scene. Worse, the output arrives as a flat, baked-in block of pixels with no editable layers—take it with its flaws, or roll the dice again.

Above that engineering limitation looms a deeper crisis that strikes at the heart of journalism and documentary work: **a generated scene, no matter how photorealistic, is a picture of something that never happened.**

Over the past several weeks, my team and I have been building an experimental AI video production studio grounded in the exact opposite philosophy. I share what we learned here not because video production is my sole focus, but because the architectural principle behind it extends far beyond media—offering a blueprint for software builders, newsrooms, and serious content creators alike.

---

> **Publication Notice:** This is the opening preview and abstract. The full article/monograph is scheduled for upcoming publication and will be released soon.