How AI Can Change Editing in the Blink of an Eye: Walter Murch’s Prediction on the Future of Film Editing

Let’s first clarify that we’re talking about AI in a broad sense — not the current wave of “AI creative tools” churned out by gold-rushing, college drop-out startup founders who’ve never edited a real video in their whole life — and who consequently believe slapping TikTok subtitles on clips counts as “editing.”

We’re talking about AI as we understand its capacity today, particularly from its current use and impact in writing — and imagining the equivalent applied to film editing. I want to exclude AI-generated imagery and focus purely on editing: assembling logic from chaos, finding the thread of narrative lying beneath mountains of possibilities.

Linear vs. Random Access: Murch’s Revelation

There’s an extremely eloquent segment in In the Blink of an Eye where Walter Murch dissects the difference between linear access and random access systems.

In linear access, to find a shot, you have to scroll through every frame between where you are and what you’re looking for — effectively auditioning every shot in between. (“What about this shot?” “What if…?”)

When random access systems arrived, you could call up exactly what you needed instantly. Faster, stronger, better — except you have to know what you want.

For Murch, some of his greatest discoveries came from scrolling — looking for one thing but stumbling upon unexpected options.

It’s not the same as just watching the rushes at the beginning. When working through a film at different phases, your mind evolves. The same footage you saw before is examined completely differently as the edit progresses.

Brief, serendipitous encounters with different options can unlock possibilities you’d never have thought of otherwise.

Recognition vs. Articulation: Making a film is Learning a Second Language

Again, I paraphrase the master:

When you learn a second language, your ability to listen and understand will always outpace your ability to speak. You recognize more than you can articulate.

With non-linear random access, you have to articulate — you have to know exactly what you want.

But linear systems “talk back.” You don’t have to articulate everything upfront — you just recognize options as they appear in front of you. This lets you work not just at the level you can articulate, but at the level of your cognition.

Modern digital software — Premiere, FCP, Avid, Resolve — are all non-linear systems built on random access methodology.

So while today’s editors benefit tremendously in speed and cost, we are also locked into our level of articulation.

And make no mistake — the tools an editor uses have more of an effect on the final film than one typically believes.

How AI Can Bring Back the Best of Linear Editing

Creating a digital linear system today would be foolish. The digital world doesn’t want to be linear.

But AI — the next evolutionary phase of computing — actually brings back the benefits of linear systems by doing one key thing: presenting options.

… and that’s all AI should do.

Anyone expecting AI to take over the creative process simply doesn’t understand the iterative nature of editing — or they produce internet litter content, and have mistaken that for creativity.

There is no art without intention.

Once we stop expecting AI to make good creative choices, we can focus on its true strength: presenting options to speed up the iterative process.

What Happened to Writing May Happen to Video Editing

To start: I hate AI-generated writing. I feel disrespected when I see it in emails or social media posts.

But I do use LLMs in my own writing —as a writing partner — to give me options for phrasing, clarity, or just something different.

In true Murch fashion, LLMs are writing partners that talk back. They elevate your writing by allowing you to write at the level at which you can recognize something to be good.

But the choice and the responsibility in the end is mine. And treating it as such, I have become a better writer for it.

AI Video Editing Will Follow

Again, AI is the buzz. Bottom-feeders are crowding the space, making noise.

But soon, the landscape will shift. I believe AI can become a net-positive tool for video, film, documentary, and YouTube creatives.

And it won’t happen by attempting to replace editors but by doing what it does best: presenting options and helping us iterate.